Division 44 Environment and Infrastructure Sector project "Transport Policy Advice"

Sustainable Transport: A Sourcebook for Policy-makers in Developing Module 3a Mass Transit Options – revised July 2005 – OVERVIEW OF THE SOURCEBOOK Sustainable Transport: A Sourcebook for Policy-Makers in Developing Cities What is the Sourcebook? Modules and contributors ThisSourcebook on Sustainable Urban Transport Sourcebook Overview and Cross-cutting Issues of addresses the key areas of a sustainable transport Urban Transport (GTZ) policy framework for a developing . The Institutional and policy orientation Sourcebook consists of more than 20 modules. 1a. The Role of Transport in Urban Development Who is it for? Policy (Enrique Peñalosa) The Sourcebook is intended for policy-makers in developing cities, and their advisors. This target 1b. Urban Transport Institutions (Richard Meakin) audience is reflected in the content, which 1c. Private Sector Participation in Transport Infra- provides policy tools appropriate for application structure Provision (Christopher Zegras, MIT) in a range of developing cities. 1d. Economic Instruments How is it supposed to be used? (Manfred Breithaupt, GTZ) The Sourcebook can be used in a number of 1e. Raising Public Awareness about Sustainable ways. It should be kept in one location, and the Urban Transport (Karl Fjellstrom, GTZ) different modules provided to officials involved Land use planning and demand management in urban transport. The Sourcebook can be easily adapted to fit a formal short course training 2a. Land Use Planning and Urban Transport event, or can serve as a guide for developing a (Rudolf Petersen, Wuppertal Institute) curriculum or other training program in the 2b. Mobility Management (Todd Litman, VTPI) area of urban transport. GTZ is elaborating Transit, walking and cycling training packages for selected modules, being 3a. Mass Transit Options available since October 2004. (Lloyd Wright, University College London; What are some of the key features? Karl Fjellstrom, GTZ) The key features of the Sourcebook include: 3b. A practical orientation, focusing on best (Lloyd Wright, University College London) practices in planning and regulation and, 3c. Bus Regulation & Planning (Richard Meakin) where possible, successful experience in 3d. Preserving and Expanding the Role of Non- developing cities. motorised Transport (Walter Hook, ITDP) Contributors are leading experts in their fields. 3e. -Free Development An attractive and easy-to-read, color layout. (Lloyd Wright, University College London) Non-technical language (to the extent Vehicles and fuels possible), with technical terms explained. 4a. Cleaner Fuels and Vehicle Technologies Updates via the Internet. (Michael Walsh; Reinhard Kolke, How do I get a copy? Umweltbundesamt – UBA) Please visit http://www.sutp.org or http://www. 4b. Inspection & Maintenance and Roadworthiness gtz.de/transport for details on how to order a (Reinhard Kolke, UBA) copy. The Sourcebook is not sold for profit. Any charges imposed are only to cover the cost of 4c. Two- and Three-Wheelers(Jitendra Shah, printing and distribution. You may also order World Bank; N.V. Iyer, Bajaj Auto) via [email protected]. 4d. Natural Gas Vehicles (MVV InnoTec) Comments or feedback? 4e. Intelligent Transport Systems (Phil Sayeg, TRA; We would welcome any of your comments or Phil Charles, University of Queensland) suggestions, on any aspect of the Sourcebook, by 4f. EcoDriving (VTL; Manfred Breithaupt, e-mail to [email protected], or by surface mail to: Oliver Eberz, GTZ) Manfred Breithaupt Environmental and health impacts GTZ, Division 44 5a. Air Quality Management (Dietrich Schwela, P. O. Box 5180 World Health Organisation) D - 65726 Eschborn / Germany 5b. Urban Road Safety (Jacqueline Lacroix, DVR; Further modules and resources David Silcock, GRSP) Further modules are anticipated in the areas of Financing Urban Transport; Benchmarking; and 5c. Noise and its Abatement Car Free Development. Additional resources are (Civic Exchange Hong Kong; GTZ; UBA) being developed, and an Urban Transport Photo Resources CD-ROM is available. 6. Resources for Policy-makers (GTZ) Module 3a Mass Transit Options

Findings, interpretations, and conclusions expressed in this document are based on infor- mation gathered by GTZ and its consultants, partners, and contributors from reliable sources. GTZ does not, however, guarantee the accuracy or completeness of information in this document, and cannot be held responsible for any errors, omissions or losses which emerge from its use.

Authors: Lloyd Wright (Institute for Transportation and Development Policy) and Karl Fjellstrom (GTZ) About the contributors With additional contributions from Armin Wagner and helpful review comments on the Bangkok TheInstitute for Transportation and Develop- Skytrain system by Phil Sayeg of Policy Appraisal ment Policy (ITDP) was established in 1985 Services Pty Ltd. to promote transport options that are environ- Editor: mentally, economically, and socially sustainable. Deutsche Gesellschaft für Technische Zusammenarbeit (GTZ) GmbH ITDP is an international non-governmental P. O. Box 5180 organisation that particularly focuses upon the D - 65726 Eschborn, Germany http://www.gtz.de promotion of , non-motorised transport, travel demand management, and Division 44, Environment and Infrastructure improved land-use planning. ITDP works Sector Project "Transport Policy Advice" exclusively in developing countries and econo- Commissioned by mies in transition, where the consequences of Bundesministerium für wirtschaftliche Zusammenarbeit und Entwicklung (BMZ) inadequate basic mobility are the most keenly Friedrich-Ebert-Allee 40 felt, and where the adverse social and environ- D - 53113 Bonn, Germany http://www.bmz.de mental effects of rapid motorisation are causing the greatest economic and environmental prob- Manager: lems. To fulfil its mission, ITDP has three core Manfred Breithaupt activities: Editing: Manfred Breithaupt (I) Catalysing demonstration projects with progressive municipalities; Cover photo: Karl Fjellstrom (II) Communicating successful options and Bangkok’s Victory Monument, Dec. 2001 technical information; and Layout: (III) Encouraging better policy making at the Klaus Neumann, SDS, G.C. local, national, and multi-lateral levels. Eschborn 2002 (revised July 2005)

i 1. Introduction 1 4.2 Planning & construction time 20 Project development and planning 20 Construction 20 2. Mass Rapid Transit concepts 2 4.3 Passenger capacity 22

2.1 Terminology 2 4.4 Flexibility 23

2.2 Defining features of MRT 3 4.5 Speed 24 Use of space 3 4.6 Institutional capacity for Speed and passenger capacity 3 successful implementation 24 Integration 4 The scope of the challenge 24 Level of service 4 Role of the private sector 25 2.3 The strategic importance of Supportive policy setting 25 MRT systems 4 4.7 Long term influence on city development 26 3. Current applications in MRT and city form 26 developing cities 5 MRT and development 26

3.1 5 4.8 Poverty alleviation 27 Latin American experience 6 4.9 Environmental impact 27 Asian experience 8 North American experience 9 5. Conclusion 29 European experience 10 Australian and New Zealand programs 10 Resource materials 30 3.2 Transit 11 Current applications 11 LRT and lines in Shanghai 11 3.3 Metros 12 The Bangkok Skytrain (BTS) 13 3.4 15 Current applications 15

4. Comparison on key parameters 16

4.1 Cost 16 Positive experience with concessioning of commuter rail services 16 Capital costs for rail-based MRT 17 Capital costs for Bus Rapid Transit 18 Operating costs 18 Rolling stock 19 Public finances 20

ii Module 3a: Mass Transit Options

1. Introduction leading consideration is cost (including cost of construction, rolling stock, and operation); Choices on public transit options are choices others include planning & construction time, about a city’s future. Will there be congestion? flexibility in implementation, passenger capacity, Will there be high levels of air and noise pollu- speed, and institutional issues. Longer term tion? Will transport be affordable? Will services effects on poverty, city form, and the environ- be available to all? The type of public transit ment are also assessed. In terms of maintaining system will have a big impact on the answers to a transit-friendly city form and ensuring the these questions (Figure 1). urban poor have access to employment, contacts This module aims to provide policy-makers in and services, a crucial factor when comparing developing cities—and those advising them— systems is the potential for a Mass Rapid Tran- with guidance on choosing appropriate Mass sit system to secure long term advances—or at Rapid Transit (MRT) systems. The module least stabilisation—in the share of people travel- begins by briefly describing some basic concepts ling by public rather than private transport. and defining features of MRT in developing cities. Current applications of each of the main “Choices on transit options are MRT options are then described, focusing on choices about a city’s future.” applications in developing cities. Since Metros and Light Rail Transit are still relatively un- The module ends with a discussion of what the common in low income developing cities, most comparison of the different options reveals. It is of this discussion focuses on the recent develop- seen that although there is no single MRT solu- ment of Bus Rapid Transit systems throughout tion fitting all cities, for all but the major cor- the world. ridors of relatively wealthy and dense developing The main section of the module then compares cities which are planning to develop a MRT each of these MRT options in the light of key system, the best option will often be a form of parameters for developing cities. Naturally, a Bus Rapid Transit.

Fig. 1 Which future? Choices about Mass Rapid Transit concern the kind of city we want to live in. Lloyd Wright, 2002

1 Sustainable Transport: A Sourcebook for Policy-makers in Developing Cities

2. Mass Rapid Transit concepts passengers within urban areas, or between urban areas and their suburbs, but differs from 2.1 Terminology Metros and LRT in that the passenger The distinction between many MRT concepts generally are heavier, the average trip lengths is fluid, and many different approaches are are usually longer, and the operations are car- commonly used to distinguish the different ried out over tracks that are part of the railroad modes and features of various MRT systems. system in the area. Apart from basic defining features such as cost, Light Rail Transit capacity, and technology, other features used to A light rail transit (LRT) system is a metro- delineate MRT systems might include distance politan electric railway system characterised by between stops, extent of right-of-way, opera- its ability to operate single cars or short trains tional regimes, and guidance systems. For the along exclusive rights-of-way at ground level, purposes of this module we have distinguished aerial structures, in subways, or occasionally in between four general forms of Mass Rapid streets, and to board and discharge passengers Transit: Bus Rapid Transit, Metros, Commuter at track or car floor level (TCRP, 1998). LRT Rail, and Light Rail Transit. systems include tramways, though a major dif- Mass rapid transit ference is that often operate without an Mass rapid transit, also referred to as public exclusive right-of-way, in mixed traffic. transit, is a passenger transportation service, Bus Rapid Transit usually local in scope, that is available to any Many cities have developed variations on the person who pays a prescribed fare. It usually theme of better bus services and the concept re- operates on specific fixed tracks or with sepa- sides in a collection of best practices rather than rated and exclusive use of potential common a strict definition. Bus Rapid Transit is a form track, according to established schedules along of customer-oriented transit combining stations, designated routes or lines with specific stops, al- vehicles, planning, and intelligent transport though Bus Rapid Transit and trams sometimes system elements into an integrated system with operate in mixed traffic. It is designed to move a unique identity. large numbers of people at one time. Examples include heavy rail transit, light rail transit, and Bus Rapid Transit typically involves busway Bus Rapid Transit. corridors on segregated lanes—either at-grade or grade separated—and modernised bus technol- Heavy rail transit ogy. However, apart from segregated busways A heavy rail transit system is “a transit system BRT systems also commonly include: using trains of high-performance, electrically Rapid boarding and alighting; powered rail cars operating in exclusive rights- Efficient fare collection; of-way, usually without grade crossings, with Comfortable shelters and stations; high platform stations” (TCRP, 1998). Clean bus technologies; Metro Modal integration; Metro is the most common international term Sophisticated marketing identity; for subway, heavy rail transit, though it is also Excellence in customer service. commonly applied to elevated heavy rail sys- tems. In this module we use “metro” to refer to Bus Rapid Transit is more than simply opera- urban grade-separated heavy rail systems. They tion over exclusive bus lanes or busways. Ac- are the most expensive form of MRT per kilo- cording to a recent study of at-grade busways metre, but have the highest theoretical capacity. (Shen et al., 1998), only half of the cities that have busways have developed them as part of a Commuter rail systems systematic and comprehensive package of meas- Commuter rail or suburban rail is the portion ures as part of the city mass transit network that of passenger railroad operations that carries we would identify as a BRT system.

2 Module 3a: Mass Transit Options

While Bus Rapid Transit systems always include 2.2 Defining features of MRT Fig. 2 some form of exclusive right-of-way for , Use of space The amount of space the applications we consider in this module are required to transport Similar space-efficiency considerations (see mostly at-grade, street-level busways. Elevated the same number of Figure 2) apply to all the MRT modes, al- passengers: car, bicycle, busways or tunnels may be needed for travers- though in practice it arises as a policy issue and bus. ing some city centres, but in many developing Poster displayed at the City only with regard to buses and some versions of of Muenster Planning Office, cities funds will not be available for extensive LRT, since rail systems are already fully segre- August 2001 grade separation. gated from other traffic. BRT and LRT often Bus lane (or priority bus lane) involve re-allocation of existing road space in favour of more efficient modes, whereas A bus lane is a or street reserved prima- Metros are normally fully grade separated and rily for buses, either all day or during specified have no impact on road capacity, unless they periods. It may be used by other traffic under are elevated in which case there may be a small certain circumstances, such as while making a reduction in road capacity. turn, or by taxis, bicycles, or high occupancy vehicles. Speed and passenger capacity Bus lanes, widely used in Europe even in small All forms of MRT operate with relatively high cities, are increasingly applied in developing cities speed and passenger capacities, and the basic such as Bangkok, where counter-flow buses can requirement of MRT in a developing city is move rapidly through peak period congestion. that it carries large amounts of passengers, rapidly. Where Metros are applied in devel- Busway oping cities they are often by far the fastest A busway is a special roadway designed for mode of MRT, with an average speed of up exclusive use by buses. It may be constructed to 40–50 km/h, while LRT and BRT systems at, above, or below grade and may be located in typically operate at average speeds of between separate right-of-way or within highway cor- 20 and 30 km/h. ridors. Some form of busway system is a feature of many Bus Rapid Transit systems.

3 Sustainable Transport: A Sourcebook for Policy-makers in Developing Cities

Integration Passenger information; All MRT systems require interchanges with Climate control; other elements of the public transport system, Modal integration; and integration with other modes of transport Integration with major trip attractors. such as cars, walking, and cycling. Shanghai, Rail-based systems have historically performed for example, provides excellent Metro/bicycle better on ‘level of service’ indicators, although and Metro/pedestrian interchanges, and good recent Bus Rapid Transit successes are challeng- Metro/bus interchanges at some major stations. ing these traditional conceptions. City’s Metro is physically integrated with the international airport and major bus 2.3 The strategic importance stations. Curitiba’s BRT system includes excel- of MRT systems lent integration with pedestrian streets and taxi stands. Sao Paulo’s BRT integrates well Developing cities are experiencing rapidly with the Metro system. Poor integration is a worsening traffic and related environmental feature of some under-performing rail-based conditions. As a first step, political commitment MRT systems, such as in Kuala Lumpur and to give priority to efficient modes of transport Manila. (transit, walking, cycling) is needed.

Level of service Experience in developed cities shows that MRT systems tend to have little impact on MRT systems usually offer a superior level land use patterns. This leads many experts of service compared to unsegregated road- to recommend that ‘adaptive’ MRT systems based modes such as regular buses, taxis, and should be used, not to attempt to influence paratransit. land use patterns, but rather to adapt to Superior service is evident for example with: the existing land use patterns (e.g., Cervero, Terminals & interchanges; 1998). In many developing cities, however, Cleanliness; the influence of MRT on land use patterns is Sophisticated marketing image; likely to be much stronger, since such cities are often undergoing rapid spatial expansion. Current trends—e.g., geared toward gated communities and greenfield housing estates in many Southeast Asian cities—often favour car-dependent urban forms, but a quality MRT system can help counteract such trends by maintaining growth along main corridors and in city centres (Figure 3). While theoretically we are told that cities should follow a ‘balanced’ approach, using ‘complementary’ MRT systems appropriate to local circumstances, in practice—especially in developing cities—once a particular MRT sys- tem is developed, resources tend to be devoted to that system, while other transit modes are neglected. Developing cities often lack the in- stitutional capacity to simultaneously develop multiple systems. This is apparent in almost all Fig. 3 developing cities which have recently pursued rail-based systems, including for example Corridors in Bogotá where the TransMilenio system operates: Many developing cities, even though increasingly traffic-saturated, retain a Kuala Lumpur, Bangkok, Cairo, Buenos Aires corridor orientation which is conducive to Mass Rapid Transit. and Manila. In all these cities, bus transit has Enrique Penalosa, 2001 been neglected.

4 Module 3a: Mass Transit Options

3. Current applications in in Table 1 are discussed in more detail below, developing cities and in Module 3b: Bus Rapid Transit.

We now survey world-wide applications of the 3.1 Bus Rapid Transit different MRT systems, focusing on developing Various BRT systems operate in cities. An ex- cities. tensive list can be seen in Module 3b: Bus -based systems in developing country Met- Transit, but here is a list of the most significant ros carry about 11 billion journeys each year, projects. surface rail about 5 billion, and light rail about In Asia: Ankara, Istanbul, , Kunming, 2.5 billion. While the proportion of public Nagoya, Taipei. transport trips by rail exceeds 50% in Seoul and In Europe: Besançon, Bradford, Clermont- Moscow, rail systems dominate only in a very Ferrand, Dijon, Eindhoven, Essen, Grenoble, few cities (World Bank, 2001). Ipswich, Leeds, Limoges, Lyon, Montpellier, Some typical MRT systems in developing cities Nancy, Rennes, Rouen, Runcorn, Strasbourg, are outlined in Table 1. Several of the systems West Sussex.

Table 1: Performance and costs of various MRT systems World Bank, Cities on the Move, Urban Transport Strategy Review (Oct. 2001)

KUALA BOGOTÁ CARACAS BANGKOK MÉXiCO TUNIS RECIFE QUITO PORTO ALEGRE EXEMPLE LUMPUR (Tras-Milenio, (Line 4) (BTS) (Line B) (SMLT) (Linha sul) Busway Busways (Putra) Phase 1) Suburban rail Category Rail metro Rail metro Rail metro Light rail Light rail Busway Busway Busway conversion Electric, Electric, Electric, Electic, Electric, Electric, AC Electric Articulated Technology Diesel buses steel rail steel rail rubber tyre Driverless steel rail steel rail duo- diesel bus Length (km) 12.3 23.1 23.7 29 29.7 km 14.3 11.2 (+ext 5.0) 41 25 20% elevated At grade, At grade, At grade, 100% 100% 100% 95% at grade Vertical segregation 55% at grade At grade Partial signal Mainly No signal tunnel elevated elevated 5% elevated 25% tunnel priority segregated priority Stop spacing (kms) 1.5 1.0 1.1 1.3 0.9 1.2 0.4 0.7 0.4 Capital cost, ($m) of 1,110 1,700 970 1,450 435 166 110.3 213 (inf only) 25 which: Infrastructure/TA/ 833 670 560 n.a. 268 149 20.0 322 25 Equipment ($m) Not included Not included Vehicles ($m) 277 1,030 410 n.a. 167 18 80 (113 vehs.) (private (private operation) operation) Capital cost/route km ($m) 90.25 73.59 40.92 50.0 13.3 11.6 10.3 5.2 1.0 40 (convoy Initial (ultimate) vehicles or 20 (30) 20 (30) 13 (26) 30 n.a. 8 operation 160 n.a. trains / hour / direction planned) Initial maximum pass 21,600 25,000 19,500 10,000 12,000 9,600 9,000 20,000 capacitys Maximum pass. carrying 32,400 50,000 39,300 30,000 12,000 36,000 15,000 35,000 20,000 capacity 20+ (stopping) Ave operating speed (kph) 50 45 45 50 13/20 39 20 20 30+ (express) 115% Rev/operating cost ratio n.a. 100 20 >100 n.a. 100 100 100 in 1998 Public (BOT) Public Public Privat Private Ownership Public Public Public Public under infrastructure, infrastructure, (BOT) (BOT) consideration private vehicles private vehicles 1995 2000 Year completed 2004 1999 2000 1998 1998 2002 Mostly 1990s (ext 2000) (1998 prices) Source: James Urban Transport System; BB&J Consult, 2000; J. Rebelo, and G. Menckhoff.

5 Sustainable Transport: A Sourcebook for Policy-makers in Developing Cities

Bogotá’s TransMilenio: In : Belo Horizonte, Bogotá, motor vehicles. Public transport is managed by initial results Campinas, Curitiba, Goiania, Leon, Porto a public company, URBS, and is operated by 10 Results of the first few years Alegre, Port of , Quito, Recife, Sao private companies under concession contracts. of operation of TransMilenio Paulo. The public transport system runs 1,677 buses— have met the high expecta- In North America: and Contra many of which are 270-passenger bi-articulated tions of the system’s devel- buses—which carry on average 976,000 pas- opers: Country, Boston, Chicago, Honolulu, Las Vegas, Los Angeles, Miami, Orlando, Ottawa, sengers per day. The 57 km of busways along • The system is moving 900,000 passengers each Philadelphia, Pittsburgh, Seattle, Vancouver. five main routes are “fed” by 340 km of feeder day (June 2005) In Oceania: Adelaide, Brisbane, Sydney. routes that concentrate passenger demand on strategically placed interchange terminals. These • Most users of TransMilenio In Africa: Abidjan, Saint-Denis. have gained more than 300 terminals are linked in turn by 185 km of circu- BRT systems are under planning or construc- hours per year to them- lar interdistrict routes. Acting in support of this selves tion in the following cities: network are 250 km of “Speedy Bus” routes that • 11% of TransMilenio’s In Asia: Bangalore, Beijing, Chengdu, Delhi, stop only at special tube stations generally set at riders are former private Dhaka, Hangzhou, Shejiazhuang. every 3 km. For the same flat fare, the passenger car drivers In Latin America: Barranquilla, Bogotá can thus transfer from one bus to another at • Average speed is higher (expansion), Bucaramanga, Cali, Cartagena, any of the terminals, extending public transport than 25 km per hour Cuenca, Guatemala City, Guayaquil, Lima, access to 90% of the city (Meirelles, 2000). • With the 72% of the total Medellín, , City, Pereira, Curitiba has inspired improvements else- number of buses the sys- Puebla, Quito (expansion), San José, San tem moves about 60,000 where. Even Los Angeles, perhaps the most Juan, San Salvador. passengers in peak hours car-dependent city in the world, has developed In North America: Albany, Charlotte, Cleve- • Noise and air pollution have Bus Rapid Transit after a recent visit of a delega- been reduced by 30% land, Eugene, Hartford, Louisville, Mon- tion of leading city officials to Curitiba. where TransMilenio runs tomery County, Reno, Salt Lake City, San • 627 buses in operation Francisco, Toronto. Bogotá, • Ticket fare of US$0.40 In Oceania: Auckland, Perth. With over 6 million inhabitants, Bogotá has • 55 km in operation In Africa: Accra, Cape Town, Dakar, Dar-Es- proven that Bus Rapid Transit is suitable even for the largest cities. Bogotá’s new TransMilenio • 67 stations in operation Salaam. system went into operation in January 2001. • 421 km feeder routes Latin American experience The existing two lines already by December • 362 feeder buses Curitiba, 2001 served over 600,000 passenger trips • 927 million trips since the It was in Curitiba in the early 1970s that the per day, greatly exceeding initial projections beginning (as of (see margin note). When the full system is July 1, 2005) Bus Rapid Transit idea first evolved. The city has implemented many other measures such as completed in 2015, TransMilenio will serve 5 car-free zones and large green spaces to become million passengers each day with 388 km of one of the world’s urban success stories. busways. Curitiba is one of the best examples of inte- Bogotá’s TransMilenio system was briefly grated transport and urban planning. It has a described in Module 1a of this Sourcebook, and population of 1.5 million and about 655,000 is discussed in more detail in Module 3b: Bus Rapid Transit. Sao Paulo, Brazil Fig. 4 Sao Paulo operates probably the largest Bus In Curitiba, boarding Rapid Transit system in the world in terms of tubes support 5-door kilometres covered. Sao Paulo, the most impor- boardings on locally tant financial and industrial centre in Brazil, manufactured buses. has 9.9 million inhabitants and 4.8 million Doors open outwards, and ramps drop down vehicles. Bus public transport is managed by a to allow same-level public company, SPTRANS, and is operated boarding. by 53 private companies. The public transport Manfred Breithaupt, 1999 system runs 12,000 buses, which carry an

6 Module 3a: Mass Transit Options

Buses are the backbone Even where extensive rail systems have been built Even cities with several subway and surface rail lines typically serve many more passengers with bus systems than with the rail systems. Mexico City’s Metro, for example, is more than 150 km in length and has 11 lines, but serves less Fig. 6 than 15% of all motorised trips. Likewise the Buenos The on-line median busway in Quito, , Aires Metro has 5 lines but covers operating costs at a fare of only US$0.2. serves only 6% of trips in the Lloyd Wright, 2001 metropolitan area. A similar in 1999, on the heels of the emerging global situation applies in Singa- market crisis, Ecuador’s banking sector virtually pore, Sao Paulo, Bangkok and other developing cities collapsed. Two governmental administrations with high cost rail-based during the late 1990s only survived a short mass transit systems. In all time in office. However, in the midst of this these cases buses continue rather chaotic scene, Quito has developed and to serve the large majority of expanded an impressive transit system featuring public transport trips, with rail serving less than 15% 25 km of exclusive busways. The system covers of trips. all operating costs with a fare of only US$0.20. Fig. 5 In nearly all developing cities Sao Paulo has the world’s most extensive bus Quito’s existing fleet of privately run buses has the majority of public trans- lane network, with 28 km of median busways taken an environmental and health toll on the port is bus based. Excep- city. Until recently, the average bus age of the tions include the ‘motorcycle and 137 km of bus lanes. cities’ such as Ho Chi Minh US Federal Transit Administration, 2001 private sector fleet has been 17 years, with some and Denpasar, where buses average 4.8 million passengers per day. The city units as old as 35 years. The electric trolley-bus serve less than 5% of trips, has 35 bus transfer terminals, 28 km of median also delivers additional environmental gains as well as rail-dominated Moscow. busways and 137 km of bus lanes. New bus through the substitution of diesel-fuelled buses Another notable partial corridors are planned to integrate the inter-city with units powered by hydro-generated electric- ity. The overwhelming popularity of the Quito exception is Hong Kong, bus lines, suburban rail and Metro systems, and though even there buses still trolley-bus has exceeded expectations and in a the local bus routes (Meirelles 2000). serve a majority of public sense created an unexpected problem. With over transport passenger trips. The system links outlying metropolitan areas 200,000 commuters now using the system daily, Railways are forecast to to Sao Paulo’s successful underground system. its maximum capacity has been reached, and handle about 40% to 50% Thus, similar to Hong Kong and Singapore thus has prompted calls for further expansion. of the total public transport where bus services are well integrated with passenger boardings in The municipality plans to deliver an additional Hong Kong by 2016, increas- Metro systems, Sao Paulo is an example of bus 73 kilometres of busways in the near future. ing from 33% in 1997 (Env. and Metro systems being mutually beneficial. For cost reasons, Quito’s new Eco-Via line uti- Protection Dept., Govt. of Hong Kong SAR, 2002). Quito, Ecuador lises II diesel buses rather than continue with electric trolley technology. Likewise, the Shanghai, with its two new Quito’s trolley-bus system and recent Eco-Via subway lines, elevated Pearl planned expansion will be utilising clean diesel addition are dramatic examples of BRT cost-ef- LRT line and suburban rail technology for its buses. fectiveness and the applicability of BRT even line, combined with the poor and deteriorating traffic under stressed economic conditions. Ecuador Porto Alegre, Brazil conditions for buses, may be has experienced several tumultuous years of Porto Alegre, Brazil has shown that BRT can following a similar trend, at political and economic misfortune. In 1998, be delivered at a relatively low-cost. In this least in the central city area. rains from the El Niño climatic effect destroyed case, the system was reportedly built for less much of the nation’s infrastructure. Then, than US$1 million per km. The city has 17 bus

7 Sustainable Transport: A Sourcebook for Policy-makers in Developing Cities

Initial results from Taipei

Initial results from Taipei, China, have also been very positive, including: • Improved traffic orderliness • Improved operating ef- ficiency of roadways • Reduced traffic interfer- ence by bus stops

• Savings in travel times Fig. 7 • Reduced frequency and Porto Alegre, Brazil. severity of accidents Lloyd Wright, 2001 • Improved bus operation, in terms of both efficiency transfer terminals, 27 km of median busways Fig. 9 and reliability and 1 km of bus lanes, along 5 radial routes (Meirelles, 2000). Nagoya, Japan, marks the bus lanes • Increased ridership of with a coloured road surface. public transport (Jason Porto Alegre employs a unique “Convoy” tech- Courtesy of John Cracknell, TTC, and the Chang, 2002). US Transportation Research Board. nique in organising its route structure. Platoons Taipei (China), along with of buses operate on main corridors and stop routes connecting with railway stations includ- Bogotá and other leading systems, is discussed in simultaneously at station bays that provide space ing the MTR, KCR and Airport Express. more detail in the Module 3b: for three buses. At the end of the main corridors, Bus Rapid Transit. the same buses continue onto separate com- Japan munity routes. Thus, rather than switching to Japan is currently hosting a 16-city Transport feeder buses at transfer terminals, customers can Demand Management program in which eight complete their entire journey without transfers. of the cities are developing bus improvement Asian experience initiatives. Kunming, China Taipei, (China) Through a partnership with the city of Zurich, Taipei has developed a bus lane network of 57 Switzerland, Kunming has become the first city km since March 1998 (at an average cost of in China to adopt the BRT concept. Hong Kong, China The Hong Kong bus system displays many fea- tures of BRT, including bus priority measures, advanced fare collection, comprehensive cover- age, clean buses, and passenger information. The system is well integrated with Hong Kong’s Metro, with an extensive bus feeder network comprising more than 140 bus feeder shuttle

Fig. 8 Nathan Road, Hong Kong. Franchised bus operators concentrate along major traffic cor- ridors where major Fig. 10 commercial centres Taipei commuters ponder the benefits are located. of bus travel. Karl Fjellstrom, June 2001 Jason Chang, 2002

8 Module 3a: Mass Transit Options

US$500,000 per kilometre), in the context of a wider policy framework emphasising: A network of dedicated bus lanes; High quality transfer environments; Green buses; Intelligent Transport System (ITS) applica- tions, including innovative passenger infor- mation systems; Transit-oriented development. Taipei has pursued a number of innovative solu- tions to finding lane spaces for buses.

North American experience Ottawa, future transit use indicated that the building of a Fig. 11 Ottawa has one of the most successful BRT costly tunnel or any other grade-separated facil- With an initial 17 city systems in North America with 26 kilometres ity in the downtown area could be safely deferred program, Bus Rapid Transit is rapidly of exclusive busways and a total system length for 20 to 25 years (Shen et al., 1998). expanding in the of over 60 kilometres. Up to 200 articulated USA United States. buses operate on the system per hour and Courtesy of Bus Rapid Transit is a success story of technol- US Federal Transit Administration handle peak capacities of approximately 10,000 ogy transfer from the developing world to the passengers per hour per direction. The system is developed. Invented in Curitiba (Brazil) Bus currently handling 200,000 passengers each day Rapid Transit is quickly being replicated in for an annual total of over 85 million passenger North America, Europe, and Australia. In the trips. The system is well integrated with other United States, the initial 17-city program is transport infrastructure including train stations, rapidly expanding, and benefiting greatly from Park and Ride lots, and cycleways. The system a national information sharing program. also provides good examples of features such as traffic signal prioritisation and queue jumping Honolulu’s successful CityExpress system has for buses (Leech, C., personal communication, now been expanded to connect the system with OC Transpo, Ottawa, 2002). a unified intercity service called CountyExpress. Pittsburgh initiated its busway program back in Ottawa’s visionary system was developed at a 1977 and now has three lines on 26 kilometres time when many other cities were looking to of exclusive busways. much more expensive rail-based mass transit solutions, and in combination with transit- Results from the US Bus Rapid Transit program friendly land use development policies. Faced are encouraging, as Table 2 shows. In virtu- in the 1980s with anticipated increases in the ally every case, travel times have been reduced and ridership levels have seen dramatic gains, metropolitan population, employment, and though from a low base. transit ridership, the transit operating agency OC Transpo strove to increase the efficiency Table 2: Positive initial results from the and use of the existing bus system in the region. US Bus Rapid Transit program. Travel time Ridership OC Transpo considered that the region would City be best served by an “outside-in” rapid transit de- reduction increase velopment strategy. The downtown segment was Pittsburgh 50% 80–100% the most expensive to construct and was there- Los Angeles 25% 27–41% fore deferred in favour of less costly construction Miami N/A 70% in the corridor leading to the downtown. The near term benefit/cost ratios were much higher Honolulu 25–45% N/A for the relatively inexpensive outer segments Chicago 25% 70% than for the costly CBD links. Also, forecasts of Source: US Federal Transit Administration

9 Sustainable Transport: A Sourcebook for Policy-makers in Developing Cities

Fig. 12 The modern Civis The Brisbane Busway bus on a busway Impressive initial results in Rouen. Courtesy of John Marino (Irisbus) Brisbane’s Southeast Busway, which opened in and the US Transportation April 2001, led in the first 6 months of operation Research Board to an increase in ridership of 12% along the same routes, compared to the previous year. The Busway rapidly gained further popular- ity. After a year of operation, the service was recording 27,000 extra passengers per week, with patronage on core bus services up by 45%. A study in 2002 showed that property values along the Busway had risen substantially, though European experience property values have also risen elsewhere in the city over the same period. A long term solution for a rapidly France also has an ambitious Bus Rapid Transit growing metropolitan region agenda with such cities as Grenoble, Lyon, Nancy, and Clermont Ferrand in France opting The Southeast Busway, to be followed by the Inner Northern Busway (due for completion in for improved bus services. late 2003), is aiming to fulfill the long-term mo- Fig. 13 bility needs of the city. It is seen as a long-term solution for the rapidly growing metropolitan Ipswich, England. area, rather than a transitional measure toward The unpaved centre a rail-based system. strip reduces costs con- siderably, and also As in Bogotá, implementation of the BRT system is done in stages, with e.g., major extensions reduces noise. such as the Inner Northern Busway, and regular Courtesy of US Transportation Research Board ongoing improvements at particular stations, interchange facilities, etc. For more informa- tion please see http://www.transport.qld.gov. au/busways/. Great Britain Busways are becoming increasingly common in such English cities as Leeds, London, Reading, and Ispwich.

Australian and New Zealand programs Several cities in Australia and New Zealand have launched Bus Rapid Transit programs. Operating systems are in place in Adelaide and Brisbane (see margin note on the Brisbane Busway). Systems are also being planned in Perth, Sydney, and Auckland.

Fig. 14 35 The Brisbane Busway features excellent station design, 50 new natural gas “green buses”, good passenger support and information, and excel- lent modal integration and marketing. It has extensive grade-separation, elevated and underground, in the city centre area. Karl Fjellstrom, April 2001

10 Module 3a: Mass Transit Options

3.2 Light Rail Transit Light Rail Transit (LRT) systems are a relatively new and promising concept for application in certain urban locations, though more relevant to wealthy than to developing cities. Comparable to BRT systems in terms of capacity, LRT pro- duces no local emissions. As with BRT, LRT lines are usually segregated from other means of traffic by barriers or slightly elevated tracks, or by full grade separation.

Current applications LRT ranges from the conventional on-street tramways of Eastern Europe and Egypt to the elevated and segregated systems of Singapore and Kuala Lumpur. With the exception of the extensive systems of Central Eastern Eu- rope and the former Soviet Union, LRT systems exist or have been planned only in relatively station. Bicycle parking is provided near all Fig. 16 wealthy developing cities such as Hong Kong, MRT stations. The major Shanghai Stadium The MRT system in Shanghai, Tunis, and Kuala Lumpur, or for high interchange is located next to a major bus ter- Shanghai has had a income developments such as the Tren de la minal. Figure 16 (see also Figure 20) shows the positive impact on land Costa of Buenos Aires. positive influence the MRT can have on land use, with densification use in the city, with a row of high density de- occurring along Metro Recent examples of LRT systems in developing routes. cities include the elevated Putra and recently velopments focusing on the Shanghai Stadium Karl Fjellstrom, Jan. 2002 opened (July 2002) systems in Kuala area; a major transit interchange. Lumpur and Shanghai’s Pearl line. On the downside, it is doubtful that the system can be expanded at a pace to match the rapidly LRT and Metro lines in Shanghai expanding city. New developments in outer The elevated (for 80% of its length) areas combined with a frenetic road-building “Pearl” LRT line (see Figure 15) in program tend to promote car-dependency. Traf- Shanghai serves high density, high- fic conditions and speeds in the city rise apartments to the north of the centre are already poor for buses Rail system descriptions city centre. A second line is being and will worsen. and maps, world-wide built to form a rough circle with the existing LRT line. The decline of trams in For a comprehensive developing cities and reasonably up-to-date The system provides listing of current rail systems excellent examples Trams, historically a feature and projects world-wide, of well-planned modal integra- of many developing cities, including for example rail retain a role in some cities, projects and expansion plans tion. The northern point of in Bangkok, Guangzhou, the Red Metro line connects such as Hong Kong, but are in decline. In Cairo the Shanghai, Taipei, , with the long distance train Sao Paulo, Manila, Kuala percentage of all motorised Lumpur, and Hong Kong trips by tram has fallen from 15% in 1971 (several different projects) Fig. 15 to 2% in 1998 (Metge, 2000). Historically see http://www.railway- ‘Shanghai City Plan’ shows many developing cities had tram systems along technology.com/projects/in- the two Metro lines in Green major corridors, but these were dex.html. and Red, and the LRT line in dismantled to make way for Maps of rail systems purple. increasing private car traffic. world-wide are available at Shanghai Tourist Map, Tourism http://www.reed.edu/~reyn/ Administrative Commission, 2005 Tram lines, now largely paved transport.html.

11 Sustainable Transport: A Sourcebook for Policy-makers in Developing Cities

over, are still visible in streets in many develop- ing cities in Asia and Latin America. Cairo (Figure 17) is one of the few developing cities with a functioning tram system, though this has gradually dwindled to one line. Renewed interest in wealthier cities In many richer cities the trends of tram decline are reversed (see Figure 18). A European best practices report notes that the decline in tram use in Munich, for example, has been reversed and patronage has increased in the last 10 years through a program of tram priority at intersec- tions and integration with other rail services (Atkins, 2001). Many other European cities have introduced and expanded tramways, both in the inner city (e.g., Amsterdam, Vienna, Frankfurt) and serving outlying commercial and Fig. 17 leisure facilities (e.g., Oberhausen, Germany). Cairo’s dwindling, neglected tram system, though averaging only around In North America, many cities have success- 11 km/h speeds, offers a pleasant community atmosphere and a fare from fully combined public transport projects with the upmarket Heliopolis to downtown Cairo of less than US$0.07. a policy of revival of their city centre. Well-de- Karl Fjellstrom, March 2002 signed and planned LRT systems are attractive to passengers, even in car-dominated, low den- sity North American cities. In the last 20 years, 14 cities in the US and Canada have introduced LRT systems. Building ‘transit malls’ with LRT access, trees and pedestrian zones can encourage private investment in city centre office blocks, shops, and apartments.

3.3 Metros Metros in developing cities carried about 11 billion journeys in 2000, more than twice the ridership of commuter rail and more than four times the ridership of LRT systems. Both Metro and commuter rail systems require exclusive right-of-way (ROW) and safety meas- ures due to relatively high speeds. To provide exclusive ROW many heavy rail systems are Fig. 18 built underground or elevated, causing very City-centre tram LRT high costs. Metro systems may cover their lines in Sapporo, Japan operational costs in urban areas with high (top) and Frankfurt, population density, such as in Hong Kong or Germany (right). Sao Paulo, but normally they require subsidies. In both cities the trams act as feeders to A successful Metro also requires integration extensive Metro systems. with existing transport modes and policies, and Karl Fjellstrom, 2002 planned densification around Metro stations.

12 Module 3a: Mass Transit Options

Metro systems are being developed or expanded in several developing cities, such as Bangkok, Santiago de , Kuala Lumpur, Sao Paulo, Buenos Aires, Mexico City (Figure 19), Cairo (Figure 20), Manila, Shanghai, and Hong Kong (see http://www.railway-technology.com/ projects for a list). Older, generally successful systems include Mexico City, Buenos Aires, and Sao Paulo, though in all cases the Metro ridership is far below the ridership of the bus system. In this Fig. 19 module we describe the cases of Bangkok and Mexico City has an extensive Metro system Kuala Lumpur in more detail, as these cases with 11 lines. Fares are low at a flat 2 peso, though the service is often overcrowded and illustrate the strengths and weaknesses of Metro run-down. An entrance is shown here, to the applications in developing cities. While the right of a bus lane. Bangkok Skytrain system is described following, Karl Fjellstrom, Feb. 2002 the Kuala Lumpur heavy rail and LRT systems are described in next section of the module, comparing costs of the various MRT options.

The Bangkok Skytrain (BTS) Three separate mass transit schemes were initi- ated in Bangkok in the 1990s: The Bangkok Transit System (BTS or bet- ter known as the Skytrain), initiated by the Fig. 21 Bangkok Metropolitan Administration; Victory Monument, The failed Hopewell elevated rail project, ini- Bangkok. BTS trains tiated by the then Ministry of Transport and run on dual tracks, car- Communications; ried on a 9 metre wide The , initiated by the Mass Rapid viaduct, supported on single box viaduct gird- Transit Authority (a 20 km underground rail ers, each 12 metres line still in planning stage, connecting to the above the road level. suburban and BTS systems). Karl Fjellstrom, Jan. 2002

Fig. 20 Cairo’s 63 km, two-line Metro carries 700 million passengers per year. Its stations, marked by a distinctive “M”, have promoted development along its route (top) and also serve poor areas (above). Karl Fjellstrom, Feb. 2002

13 Sustainable Transport: A Sourcebook for Policy-makers in Developing Cities

Bangkok Skytrain service The Skytrain, which opened in late 1999, is an innovations elevated heavy rail system running above some Recent Skytrain innovations of Bangkok’s busiest commercial areas. It has a Fig. 22 include regular promotional peak capacity of around 45,000 passengers per Each car is air-conditioned, and the BTS offers events. All are advertised, hour per direction. Trains run on 5 to 7 minute a comfortable and fast ride through the central both in the mass media and headways from 6 am to midnight, though as at the BTS stations. city area. demand increases and for special occasions such Karl Fjellstrom, Dec. 2001 In October 2001 a free shut- as New Year’s Eve, headways can be shortened tle bus service for pass-hold- former car drivers. Interestingly, one-third of ing Skytrain passengers was to 2 minutes (Sayeg, 2001) and running time BTS trips are new trips. implemented on 5 different extended. The BTS has two lines, with a total Ridership should, however, continue to increase, routes. BTS cannot charge length of 23.1 km and 23 stations. The lines especially as densification around stations takes for these services. If they intersect at the city centre station. could, and BTS was able place (encouraged by rising land values near to determine routing, this Tender documents for a turnkey BTS system stations), road traffic to the central area becomes would put pressure on the were issued in March 1993 to five consortia. The even more difficult, integration with other BMTA to change. Hence, a agreement was later amended to cover not just modes is improved, and complementary mass multi-modal concession for the construction, but maintenance and opera- the BTS extensions (under transit systems are completed. tion of the completed network. (For further construction) may be a good Despite the initially disappointing ridership, an discussion of private sector participation in the idea. Singapore’s northeast International Finance Corporation (one of the corridor is an example of a BTS see Module 1c: Private Sector Participation system’s investors) funded study indicates that: multi-modal concession, with in urban Transport Infrastructure Provision.) SBS—a bus operator—now At present, BTS is covering operating and also running trains. Fares, ridership and operating costs maintenance costs through the fare box. ... As Fares range from 15–40 Baht, or around the marginal cost of carrying passengers on US$0.37 to $1.00. This is relatively expensive, the BTS is well below the average cost, its cost even compared to air-conditioned bus fares for recovery will increase markedly as patronage long trips, which are less than $0.50, or around grows (IFC, 2001). BTS has also developed $0.11 for shorter trips. Economy bus fares are commercial spots in the stations and gets a con- much cheaper, from around $0.05 for short trips siderable amount of revenue from this activity. up to $0.20 for long trips. Modal integration First year ridership was only one-quarter of Integration of BTS with other modes of transport forecast ridership. Though it is improving, is poor; a contributing factor to the disappoint- increasing from around 160,000 to 200,000 ing ridership. The Bangkok Mass Transit Author- trips per day in its first two years of operation ity, Bangkok’s monopoly bus services provider, (average 280,000 weekday passengers in Oct. has been slow to act. The BTS meanwhile has 2002), this is still only one-third of the fore- taken steps to provide its own feeder services (see cast. Similar disappointing ridership has been margin note), but they are severely constrained. recorded for recent urban rail systems in Kuala Some clear opportunities for modal integration Lumpur (discussed later in this module) and in were missed, with the northern line terminating Manila (Metrostar). Diversion from car drivers only around 2 km from the newly constructed to the BTS system appears to be relatively high, northern bus terminal, and no feeder service or however, with around 10% of passengers being pedestrian walkway connecting the two.

14 Module 3a: Mass Transit Options

Facilities for bicycles are either not provided, or are located in an unsupportive environment for cyclists, and are therefore unused (such as at Ekkamai station). Eight stations are directly connected to adjacent shopping complexes. Rolling stock Thirty-five three-car, 1,100 passenger capacity trains, 65.1 metres long, are currently operated. The quality, cleanliness and reliability of the sys- tem are all outstanding. The three-car trains can in future be doubled in length at peak times. Future arrangements From the start of commercial operations, all operating revenue for the following 30 years was to be handed to BTSC. However, the current situation is that the BTS has been transferred back to the BMA, although BTSC still carries out the system maintenance. Fig. 23 Commuter rail in Bombay An overloaded commuter train in Jakarta, The (inevitable) need for expansion . Commuter/suburban rail services Almost all developing cities which are consider- are in decline in many developing cities. ing MRT applications or extentions are expand- Kompas, 17–Jun–01 ing at a rapid rate. It is therefore inevitable that Metro systems, which are very expensive and therefore often limited to one or two short lines, soon come under pressure for expansion to serve new areas of the city. This has also happened in Bangkok. BTS system expansion was approved Fig. 25 in 1999, and construction has commenced but Six million passengers is proceeding slowly due to problems of cost and per day are carried complexity. The four approved extensions add by suburban rail in up to an extra 33.4 km (see further http://www. Bombay, India. bts.co.th/en/btstrain_04.asp Manfred Breithaupt, Feb. 2002, ). Churchgate Station, Bombay

3.4 Commuter rail Current applications Commuter or suburban rail services are mostly provided by general railroad companies and Fig. 24 they share track with freight and long-distance Radial commuter rail lines have influenced the transport. While in theory the capacity would urban form in Buenos Aires. Nora Turco, 2001 be limited to the number of available seats, in practice these services are often run at crush a transit-friendly city form and maintaining a passenger loads in developing cities (Figure 23). strong city centre (Figure 24). Fig. 26 Market differentiation Suburban railways in developing cities are usu- As shown in Bombay, where each day 6 million in Bombay extends to passengers are carried by suburban railways, ally radially oriented into the city centre. Al- women-only carriag- though even in relatively well-served cities like this mode may even serve as a backbone MRT es, similar to Cairo’s Bombay, Rio de Janeiro, Moscow, Buenos Aires, for a developing city. Like Metros, suburban Metro. and Johannesburg, they carry less then 10% railways need an independent institutional Manfred Breithaupt, Feb. 2002 of trips, they can be important in supporting body which allocates funds and distributes

15 Sustainable Transport: A Sourcebook for Policy-makers in Developing Cities

earnings, as well as fare and timetable intregra- 4. Comparison on key parameters tion with other transport services. Measures which can increase capacity and Though ideally cities developing a MRT system safety include the elimination of at-grade road will draw from different combinations of road crossings (or introduction of safety equip- and rail-based MRT, experience shows that ment), the purchase of double-deck-trains and most developing cities will probably focus on improvement of boarding/alighting facilities, one choice for a MRT system. Once one form though in all cases the cost implications may of MRT is implemented, it is likely that other be too large for many developing cities. As with MRT options will be neglected. It is therefore all other MRT systems, high ridership on com- important that this choice is well informed. muter lines requires feeder services (e.g., by bus) and good interchange facilities. 4.1 Cost The rehabilitation and improvement of sub- For any municipality, the infrastructure cost of urban railways show good cost-benefit-ratios a transit system is a pre-eminent decision-mak- and can contribute to poverty alleviation, as ing factor. Bus Rapid Transit is relatively eco- poorer people generally live further from the nomical to develop. Without costs of excavation city centre. and expensive rail cars, Bus Rapid Transit can The most serious obstacles to rail developments be over 100 times less expensive than a Metro are frequently institutional. When operated by system. national rail organisations, suburban railways tend to be given low priority—in particular in “New subway systems in the US comparison to the road lobby—and are poorly show that costs have been well coordinated with other urban public transport services. In many cases the weakness of publicly above, and ridership well below, owned national rail undertakings leaves their forecasts made when the projects capacity severely underdeveloped (as in Manila, were approved. This has also been Jakarta, and Surabaya). the experience of many rail transit Positive experience with concessioning of commuter rail services systems in developing countries.” In Module 1c: Private Sector Participation in Gregory Ingram, World Bank, Patterns of Metro- Urban Transport Infrastructure Provision, it was politan Development: What Have We Learned?, Urban Studies, Vol. 35, No. 7, 1998 seen that positive experience is possible where these weaknesses are addressed. A program of The cost difference extends to other infrastruc- concessioning to the private sector in Buenos ture items, such as stations. A busway station Aires revitalised the system, doubling patron- in Quito, Ecuador costs only about US$35,000 age over a five year period while at the same while a rail station in Porto Alegre that serves a time reducing the budget burden of the system similar number of persons costs US$150 M. by nearly US$1 billion per year; although the system still requires an ongoing operational subsidy and operating conditions have consider- ably worsened in 2002. In Brazil the transfer of responsibility for subur- ban railways from the highly centralised CBTU (Companhia Brasileira de Trens Urbanos) to Fig. 27 Fig. 28 local (state) control, together with a government BRT station in Quito, Rail station in Porto funded rehabilitation program, has improved Ecuador: US$35,000 Alegre: US$150 M service in most of the major cities. Assisted by a Thus, for the same amount of investment, a Bus program of concessioning, it is greatly reducing Rapid Transit system can serve as much as 100 the fiscal burden. times the area of a rail-based system. A city with

16 Module 3a: Mass Transit Options

enough funding for one kilometre of Metro Table 4: Factors influencing Metro Further information on might be able to construct 100 km of BRT. capital costs. comparisons, and transit Adapted from Allport 2000 levels of service

Capital costs for rail-based MRT Influence Factor More information on transit Capital costs usually cover planning and con- Dominant - Management/organisation quality level of service, relevant to comparisons between struction costs as well as technical equipment - New system, or progressive modes — although from and rolling stock. The capital costs of US LRT expansion of existing system a North American rather systems are on average US$21.6 million per Large - Ground conditions (underground than developing country kilometre. construction, and foundations for perspective — can be elevated viaducts) The capital costs depend on the extent of grade obtained from the Transit - Urban constraints and topography Capacity and Quality of separation and right-of-way, as well as on (utilities diversions, proximity to Service Manual (http://kit- specific geological conditions and the prices of buildings, ability to divert traffic, telson.transit.com), building materials and labour, but also extend environmental constraints, prepared for the Transit to planning procedures and institutions. All- earthquake protection) Cooperative Research port (2000) shows also that the effectiveness - Design and safety requirements Program (TCRP), 1999. of planning procedures contributes to a large - Financing costs extent to capital costs. The study found that - Depth of water table (can make cost prohibitive for underground) similar Metro systems in developing countries were much more expensive, for example, than Moderate - Land costs - Competition in the equipment a system implemented in Madrid (see Table 3). supply and construction market Table 4 provides a rough assessment of factors Small - Labour costs influencing rail-based MRT capital costs. Simi- - Taxes and duties lar factors and influences can be assumed to - System features (long trains, AC, apply to BRT systems. special access, etc.) Table 4 shows, perhaps counter-intuitively, that it is not the construction phase (with labour and equipment costs) or details in system features, Table 3: Capital costs of various rail systems. UTSR 2001; Allport 2000; GTZ 2001 Fig. 29 Fig. 30 Cost/ Railway Type km Notes BRT: US$1–10 Metros: US$55–207 (US$) million per kilometre million per kilometre

West Rail Hong Heavy 38% but rather strategic decisions on management 220m Kong Metro tunnel and organisation that have the greatest influ- Kuala Lumpur – Elevated, ence on MRT capital costs. Additionally the LRT 50m Putra driverless integration in the urban fabric and the funda- Kuala Lumpur – Heavy Largely mental decision of vertical alignment will have a 50m Star Metro elevated major bearing on capital costs. Manila– Light 50m Elevated Table 5 underlines the impacts of alignment extension Metro decisions on capital costs for rail MRT systems. Bangkok Skytrain Metro 74m Elevated Table 5: Impacts of alignment on cost: Caracas– Metro 90m rail-based MRT. Allport 2000 41m Vertical All-in cost (US$m) Ratio alignment per route km 23m Tunis LRT 13m At-grade 15–30 1 Elevated 30–75 2–2.5 Comm Recife–Brazil 12m Rail Underground 60–180 4–6

17 Sustainable Transport: A Sourcebook for Policy-makers in Developing Cities

Fig. 31 Operating costs Two systems at the same cost: (1) Rail When comparing such operating cost values between mass transit modes (e.g., BRT with rail), one must be certain that a “like for like” comparison of variables is being made. BRT sys- tems typically amortise vehicle purchase costs within the operating cost calculation, while rail systems sometimes list rolling stock as a capital cost. Further, because of rail’s high cost structure, certain maintenance and replacement part items are sometimes capitalised. To make a correct comparison, adjustments will need to be (2) Bus Rapid Transit made to ensure capital and operating costs are appropriately categorised. Rail systems do have an apparent operational cost advantage from the standpoint of labour costs, specifically with regard to the cost of a driver. Bus coaches each require a driver while several rail coaches connected together only requires a single driver. However, in developing nations, the lower wage differentials mean that this advantage is largely overwhelmed by the other components. Porto Alegre, Brazil offers a Capital costs for Bus Rapid Transit unique opportunity to compare urban rail and BRT operating costs on an even basis. The city Whereas rail-based MRTs may cost from has both types of systems operating in similar US$20–180 million per kilometre, Bus Rapid circumstances. The Trensurb rail system requires Transit systems are an order of magnitude a 69% operating subsidy for each passenger trip cheaper: US$1–10 million per kilometre. (Thomson, 2001). By contrast, the city’s BRT We can view these cost differences graphically, system has a comparable fare structure, but oper- in terms of the length of MRT system achiev- ates with no subsidies and in fact returns a profit able for roughly the same cost. to the private sector firms operating the buses. Table 6 summarises costs of Bogotá’s Trans- Profitability of bus systems in developing Milenio BRT system, discussed in more detail cities in Module 3c: Bus Rapid Transit. Public transport by bus in developing countries Table 6: Infrastructure cost components is already characterised by a high level of cost of Bogotá’s TransMilenio BRT system. recovery, and usually such services operate Lloyd Wright, 2002 at a profit. The fact that such services can be Total cost Cost per km profitable under inferior and deteriorating Component (US$ million) (US$ million) operating conditions (chiefly congestion), and Trunk lines 94.7 2.5 a poor and unsupportive regulatory and plan- ning framework, indicates that where a range Stations 29.2 0.8 of operational and regulatory improvements Terminal 14.9 0.4 encouraging competition and service innovation Pedestrian are implemented along with physical measures 16.1 0.4 overpasses such as bus priority, there is little doubt that Bus depots 15.2 0.4 BRT in developing cities will be profitable. Control centre 4.3 0.1 In addition, the form of many developing cities is Other 25.7 0.7 still suited to transit, as development is often still channelled along major arterials rather than dis- Total 198.8 5.3 persed to all areas of the city. Even car-saturated

18 Module 3a: Mass Transit Options

cities such as Bangkok can be more accurately Table 7: Fare Box Ratios, selected rail MRTs. Extra costs of new considered “car-saturated transit cities” rather TCRP 1999, Allport 2000, GTZ (edited) technologies Railway Fare Box Ratio than “car-dependent cities”. These circumstances Providing refueling infra- (unlike in car-dependent cities where activities are Regional Metro Porto Alegre 0.25 structure can also be a highly dispersed) tend to favour a high ridership. consideration. According Kuala Lumpur Putra LRT 0.50 to the International Energy Rail system operating costs Buenos Aires Metro 0.77 Agency, refueling instra- structure and other support Operating costs include salaries, fuel and main- Kuala Lumpur Star Metro 0.90 tenance of both vehicles and infrastructure. The system costs for fuel cell Sao Paulo Metro 1.06 operational costs depend partly on the amount buses cost approximately US$5 million. of cars required to provide a service. The higher Singapore Metro 1.50 1.60 A major additional cost operating speeds the lower the circulation time for new technologies such and in consequence the number of cars needed Manila Light Metro 1.80 as fuel cells, which is not for a single line. Hong Kong Metro 2.20 included in Table 8, is the cost of research and “The construction costs of Metros railway operations are able to cover operational development for the transit agency concerned. in developing countries are so high costs and to use the surplus for depreciation that they crowd out many other of infrastructure. These are exceptional: Most railway operations are subsidised by an agency investments. ... Most systems have or surpluses in other branches of the city budget. operating deficits that severely Fare Box Ratios of BRT systems constrain local budgets, as in The Fare Box Ratio of BRT systems in Porto Pusan and Mexico City.” Alegre, Curitiba, Bogotá, and Quito exceeds Gregory Ingram (op cit) one, as do most bus systems throughout the developing world. A recent US survey (GAO, 2001) confirms that operational costs of LRT systems are much Furthermore, as shown in Module 3c: Bus higher than for BRT. The report compares six Rapid Transit (see Figure 6) revenues from the US cities having both LRT and BRT systems. It TransMilenio BRT in Bogotá do not only cover refers to three categories of operating costs: operating costs for the trunk line operators, but Costs per vehicle hour; also cover a range of other costs, including the costs of the feeder services, the system planning Costs per vehicle revenue km; and regulatory body (3% of fare revenues), the Costs per passenger trip. fare collection company, the funds administra- Operating costs per vehicle hour of 5 LRT tor, and a contingency fund. systems are between 1.6 to 7.8 times higher than those of BRT systems. LRT operating Rolling stock costs per vehicle hour ranged from $89 to $434. Table 8 provides an approximation of the Similar findings were made for operating costs cost difference between buses with different per vehicle revenue kilometre. Table 8: Costs of various bus technologies, The World Bank (2001) provides some figures compared to a standard rail car. for developing countries (see also Table 1). International Energy Agency, 2002. Cost per vehicle Operating costs per passenger range from Propulsion technology US$0.61 in Hong Kong to $0.19 in Santiago, (US$) New diesel, constructed while revenues per passenger range from $0.11 30,000–75,000 in Calcutta to $0.96 in Hong Kong. in developing country New diesel (Euro II) 100,000–300,000 Fare Box Ratio CNG, LPG bus 150,000–350,000 The Fare Box Ratio gives an indication of Hybrid electric bus 200,000–400,000 economic viability of a MRT system. It de- Fuel cell bus 1.0–1.5 million scribes the ratio between fares collected and operational costs. Table 7 indicates that five Metro rail car 1.7–2.4 million

19 Sustainable Transport: A Sourcebook for Policy-makers in Developing Cities

Construction time propulsion systems, compared to a standard all in financial trouble and are nowhere achiev- advantages of bus rapid rail car. The purchase cost does not include ing profitability (see further Module 1c). Alone transit substantial and ongoing additional costs such among rail MRT systems, the Hong Kong Metro Bangkok’s Skytrain system as specialised maintenance, and research and funds all its costs (capital, asset replacement, took four-and-a-half years development needs that accompany the most and operating) from its mainly farebox revenues, to establish, from the time advanced technologies. and can be considered profitable. All other rail of signing the construction contract to first operation. MRT systems require support from the public Public finances sector; often very substantial (Allport, 2000). Bogotá’s TransMilenio BRT In terms of public sector affordability, BRT is system—with 56 stations The problems encountered by new rail MRT compared to the Skytrain’s the most favourable form of MRT system. BRT systems in developing cities are in many ways il- 25 stations and with a large systems require a relatively small initial outlay. lustrated by the experience of the Star and Putra range of associated improve- Bogotá, for example, was able to build the ments such as pedestrian rail MRT systems in Kuala Lumpur, Malaysia system's first phase of around 40 km without (see text box on the next page). and cyclist facilities, public taking out loans. parks and so on—took less than 3 years from concept Savings, meanwhile, can be used in other areas, 4.2 Planning & construction time to full implementation. The such as health and education, public space facili- actual physical construction ties, and conditions for pedestrians and cyclists. Project development and planning of the entire system, includ- The project development and planning process ing the associated public Rail systems—both LRT and Metros—require is generally quicker for BRT than for rail-based space improvements, took much greater initial outlays and ongoing MRT systems. The BRT planning process for a only around 8 months. subsidies. Though the advent of private sec- ‘world class’ BRT system, described in Module tor concessionaires was expected by many to 3c: Bus Rapid Transit, takes about one year and change this situation, the evidence is that the costs around US$400,000–US$2 million. various new Build-Operate-Transfer projects are Construction time Due to the relatively low costs, financing is also generally easier and quicker for BRT than Fig. 32 for rail-based systems. Jakarta, Indonesia, for BRT: < 18 months 4 example, decided in late 2001 to implement a Lloyd Wright, 2001 (Bogotá) BRT system, and the government was able to quickly allocate funds from the routine city development budget.

“Mayors who are elected for only

Fig. 33 three or four years can oversee a Metros: > 3 years6 BRT project from start to finish.” K. Fjellstrom, Feb-02 (Sao Paulo) Construction The simpler physical infrastructure of Bus Rapid Transit means that such systems can also be built in relatively short periods of time, often in less than 18 months. Underground and elevated rail systems can take considerably longer, often well over three years. This time difference has a political dimension. Mayors who are elected for only three or four years can oversee a BRT project from start to finish. Successfully implemented BRT systems have positively influenced the re-election and political careers of mayors in cities such as Curitiba and Bogotá.

20 Module 3a: Mass Transit Options

Gov’t completes takeover Rail-based MRT in Kuala Lumpur of two LRT operators

Malaysia has developed several new rail MRT Kuala Lumpur 1:51pm, systems, often portrayed as paragons of tech- Fri: (AFP)—The govern- nological progress and sophistication. But are ment today completed the they sustainable? The systems include STAR takeover of two debt-ridden Light Metro (operating from Dec. 1996) Putra light railway companies in LRT (from Dec. 1998), the KLIA Airport Express its largest ever restructuring (from Apr. 2002), and the Monorail LRT (from exercise, dealers said. July 2002). The various rail systems all inter- The government issued sect at the city centre, though there is no fare four tranches of bonds integration between them. totalling RM5.467 billion with maturities of five, seven, 10 and 15 years in a debt conversion scheme to settle Fig. 35 the two companies’ debts, Kuala Lumpur’s city centre monorail has expe- bond dealers said. rienced many delays in construction since 1997. The serial bonds will be Though it serves thriving commercial areas and issued to creditors of Projek interlink with the other rail systems, after the Usahasama Transit Ringan experience with STAR and Putra, the govern- Automatik (Putra) and Sistem ment must be questioning the financial viability Transit Aliran Ringan (Star) of its rail-oriented MRT strategy. in the debt replacement, they added. Fig. 34 The deal, made through Putra’s grand Dang Wangi station is often a special purpose vehicle Syarikat Prasana Negara, deserted. Pedestrian access is difficult, with would see the government no crossing provided in front of the station. acquiring 80% of the assets of both operators, In its first three years of operation Putra’s rider- the New Straits Times said. ship increased 10-fold, from 15,000 to 150,000 The railway networks are passengers per day. This increase in ridership, to be leased back to the however, was only achieved after substantial private firms to operate. fare reductions which probably had a negative Putra, which is owned by overall effect on revenue (Sayeg, 2001). Despite debt-ridden conglomerate this ridership gain, however, Putra has been a Renong, is the biggest financial failure and along with STAR the venture debtor among the two, with was nationalised in late 2001. After only 3 years total debts amounting to of operation, Putra had accummulated debts of RM 4.27 billion, the newspa- more than US$1.4 billion (see margin note).* per said. The Monorail and KLIA airport services * Note: On 1 Sept. 2002 KL’s monorail, linking the LRT lines, was due Syarikat Prasarana Negara to open in mid 2002. However, a mishap dur- Fig. 36 Berhad (SPNB), a wholly- ing a trial run in July (a wheel fell off, striking a owned subsidiary of the Min- journalist) has led to the opening being delayed This makeshift tent (above) serves as the major bus stop at Kuala Lumpur’s largest shopping ister of Finance, completed until early 2003. Major commercial areas and the sale and purchase of the trip attractors—many currently under construc- mall (top left). Buses are infrequent and over- assets and business opera- tion—line its route. loaded, and passengers are forced to scramble tions of Sistem Transit Aliran Two rail connections to the city’s International past taxis (above). The megamall is actually Ringan Sdn Bhd (STAR) and Airport, 70 km from the city centre, are also be- only around 1.5 km from an LRT station, Projek Usahasama Transit ing built. One of these, the US$260 million, 57 though no feeder bus service to the mall is pro- Automatik Sdn Bhd (PUTRA) km KLIA Airport Express line, opened in April vided, and nobody walks from the LRT station from the Renong Group. 2002 but at only 3,000 passengers per day (and to the megamall, as the walkway is pot-holed, SPNB said it will continue a hefty fare of US$10), ridership has been well very narrow, and unprotected from the sun and operating STAR and Putra. below forecasts. rain (top right). Photos Karl Fjellstrom, Dec. 2001

21 Sustainable Transport: A Sourcebook for Policy-makers in Developing Cities

Under-achieving new 4.3 Passenger capacity urban rail systems in the Asia-Pacific region Misconceptions abound about the potential of BRT, especially in dense developing cities. A Star, Putra, and KLIA Airport common misconception is that “Any city seri- Express MRTs in Kuala Lumpur, Metrostar in Manila ously wishing to move toward sustainability (17 km, Dec. 1999), the by changing the private car/public transport Sydney (10 equilibrium…must move in the direction of km, June 2000 and now electric-rail-based transit systems” (Newman in receivership), the Hong & Kenworthy 1999, p. 90). Table 9 draws from Kong Airport Express Rail (34 km, mid 1997), the Bangkok Newman & Kenworthy’s book to present—and Sky Train, and the Brisbane then counter—several typical “myths” of BRT. Airtrain airport link: all of Fig 37 Another misperception is that Bus Rapid Tran- these new MRT rail systems People walking or taking a bus to the megamall have shown disappointing sit cannot serve high passenger numbers. The ridership, generally about (see Figure 36) must cross a busy road with no results in Colombia and Brazil show that Bus one-quarter the projected help from signals or road markings. Not sur- Rapid Transit can handle passenger flows in prisingly, almost everyone gets to and from the levels. From these systems the range of 20,000 to 35,000 passengers per mega-mall by car or taxi. Long queues form all the longest in operation, hour per direction. Table 10 shows passenger Star, has stabilised at day for taxis. around 20–25% of projected Photo Karl Fjellstrom numbers actually recorded for different systems ridership. Brisbane’s Airtrain in selected cities. Some of the biggest factors opened in May 2001 and Rail at the expense of bus determining capacity is not the mode of trans- operates without government services? port but rather the techniques used for boarding subsidy. However the Airtrain and alighting. has an uncertain future, Though Kuala Lumpur has made much recent with ridership of just 6,000 progress, including many initiatives to improve Table 9: Some ‘myths’ of Bus Rapid Transit. per week compared to a conditions for pedestrians in the city centre, projected 52,000 per week. and major new rail facilities, bus services re- ‘Myth’ In fact... main unreliable, unintegrated, unprofitable, and An important factor here is May be true in some neglected (The Star, 21 Dec. 2001). the fare: the Singapore and Only rail systems cases, though a recent Hong Kong successful MRT The lack of attention to buses is reflected in the are fast enough to study (GAO, 2001) shows systems have fares com- poor conditions at Kuala Lumpur’s main bus compete with the that in 5 of 6 US cities parable to air-conditioned station. The bus station is a stark contrast to the private car (p.90) with both BRT and LRT, bus services, and, relative to shiny new expressways and rail lines of modern BRT was faster KL. Litter is scattered around and water forms income, are about one-quar- Buses are effective Success to date with ter as expensive as fares in standing pools. The litter and water, combined with the confined exhaust smoke (there are no in transit cost BRT has come from cities Bangkok, Manila, and Kuala recovery only where other than developing Lumpur (Sayeg, 2001). exhaust fans and little circulation), foul odour, slippery stairs, and poor lighting, contributes there are large Asian cities, including to a wholly unpleasant experience for passen- numbers of captive Latin America and gers. (This situation should be rectified by a users, as in newly Canada. Curitiba has the major new bus terminal under construction in developing Asian largest car-ownership in the city centre, which integrates directly with cities (p.117) Brazil, after Brasilia the Star MRT line. Further improvements were Rail systems offer a Many developing achieved with the opening in 2002 of KL Sentral “more fundamental cities have tragically Stesen, the new central rail station, which links way to recover wasted scarce the Metro and LRT systems with the commuter transit costs” (p.117) development funds on rail lines.) and are “c…heap in expensive infrastructure comparison to…any It is not just Kuala Lumpur which is preoccu- megaprojects. BRT is a highway option” pied with large-scale projects to the detriment cheaper option of bus systems and non-motorised transport. (p.155) In developing cities ranging from Jakarta to Buses cannot Passenger flows in many Buenos Aires, Bangkok to Guangzhou, Ho Chi cope with a high BRT systems regularly Minh City to Surabaya, policy-makers have passenger demand reach more than 25,000 consistently given more attention to large-scale, (p.196) pax/hr/dir expensive projects such as expressways, ring LRT is a natural BRT is implemented as roads, LRT, and Metros, rather than to lower progression ‘up’ a long term strategy in cost approaches. after BRT (p.200) many cities

22 Module 3a: Mass Transit Options

Table 10: Actual and theoretical maximum passenger numbers into a single corridor. Some- Flexibility in operation ridership, selected MRT systems. times this situation occurs due to geographical Bus-based systems’ Ridership constraints (Hong Kong), but it is often due to ability to operate both on Line Type (pass/hr/ a lack of funding for a city-wide Metro system. and off a busway or bus dir) Thus, in a sense, the high capacity figures be- lane provides Bus Rapid Transit the flexibility to Hong Kong Metro 81,000 come inevitable. However, such situations can respond to operating prob- Sao Paulo East Line Metro 60,000 be avoided by offering more distributed systems. lems. For example, buses can pass disabled vehicles, Santiago La Moneda Metro 36,000 Whether a city is utilising bus or rail transit while Light Railtrains can London Victoria Line Metro 25,000 systems, system designers may wish to keep be delayed behind a stalled Buenos Aires Line D Metro 20,000 capacity figures within manageable bounds. If train or other vehicle on the tracks. Thus, the impact of Buenos Aires Line E Metro 5,000 a system is operating at over 50,000 pphd and a technical or operational problem occurs, the a breakdown of a Bus Rapid Mexico Line B Metro 39,300 Transit vehicle is limited, entire system can become overwhelmed with Bangkok BTS Metro 50,000* while a disabled Light Rail passenger backlogs very quickly. Further, very train may disrupt portions Kuala Lumpur Putra LRT 30,000* high capacity lines can be uncomfortable and of the system (GAO, 2001). Bogotá TransMilenio BRT 33,000 unsafe for passengers if tight passenger “pack- Recife Caxanga, Brazil BRT 29,800 ing” becomes necessary. Belo Horizonte, Brazil BRT 21,100 Goiania, Brazil BRT 11,500 4.4 Flexibility Sao Paulo 9 de Julho BRT 34,911 Unlike rail-based options which are by nature Porto Alegre Farrapos BRT 25,600 more fixed, BRT allows a great deal of flexibility Porto Alegre Assis BRT 28,000 for future growth. Making new routings and Quito Trolleybus BRT 15,000 other system changes to match demographic Curitiba Eixo Sul BRT 15,100 changes or new planning decisions is fairly easily accomplished. Bogotá’s plans for a phased Ottawa Transitway BRT 10,000 (* Theoretical max., not actual ridership. Putra ridership is approx. 150,000 BRT expansion (figure following) provides a per day; BTS less than 300,000 passengers per day). good example of matching technology to the Lloyd Wright; GTZ; from various sources, 2001 dynamics of urban centres. Capacity and patronage are cardinal points Fig. 38: Growing and changing with the city when it comes to assessing the financial viability of a MRT. Capacities up to 30,000 passengers per hour per direction (pphpd) are currently handled by bus while capacities exceeding 35,000 pphpd can only be handled by Metros. The maximum recorded ridership of most LRT systems are limited to approximately 12,000 pphpd, although the Alexandria-Rami (Egypt) line serves 18,000 pphpd. TransMilenio 2001 TransMilenio 2015 The necessity for very high capacity flows in TransMilenio, SA, Bogotá, Colombia part depends upon the structuring of a system. Cities such as London and New York are fairly BRT systems provide greater flexibility than dense and enjoy high usage of their Metro LRT in implementation and operation. Im- systems. However, peak capacities are only in provements such as signal prioritisation and the area of 20,000–30,000 pphpd. This occurs, interchanges, which improve capacity and bus because these systems feature multiple lines speed, can be added incrementally. distributing passenger flows about the city. In Since buses approach and leave busways at inter- cities such as Hong Kong and Sao Paulo, the mediate points, many different routes can serve higher capacities are achieved by offering a a passenger catchment area, with fewer passen- limited number of lines and then feeding large ger transfers than would be required in a fixed

23 Sustainable Transport: A Sourcebook for Policy-makers in Developing Cities

�������������� �� ��

��

�� �� �� �� ��

�� �� �� �� �� �� �� �� �� ��

� ������ ������ ����������� ���������� ��������� ��������

���������� ����������������� Fig. 30 guided system. This is an important feature of A medium term goal Curitiba’s successful system, where express buses Fig. 31 in Bogotá is to expand combine some feeder features at the extremity In five of six cities with both BRT and LRT the TransMilenio of the route, thereby minimising transfer needs systems, BRT speeds were higher. The one ex- BRT system so that of passengers. Bus Rapid Transit can also more ception was Los Angeles, where the BRT system does not provide dedicated bus lanes. 85% of the city’s 7 closely match capacity and service quality to million inhabitants live GAO,2001 (from National Transit Database and six transit agencies) changing passenger demands and special events, within 500 meters of a and buses are more able to segregate the market, (Figure 31). Thus, low-cost bus systems can TransMilenio line. Such match the travel times of expensive rail systems. an expansion program providing a range of services (air-conditioned, would be unrealistic express, etc.). for a rail-based MRT 4.6 Institutional capacity for system. “Expanding and adjusting a rail system successful implementation Enrique Penalosa, 2001 is much more costly and complex.” Institutionally, rail-based systems are demanding: Without high standards of operations, mainte- In terms of flexibility to expand and adapt to nance and administration [Metros] will rapidly a changing city, Bus Rapid Transit offers clear deteriorate [...]. The culture, managerial standards advantages over a rail-based system (Figure 30). and attitudes often found in bus companies and Expanding and adjusting a rail system is much railway corporations of developing countries are more costly and complex. Developing cities fol- unsuitable for a Metro. Accordingly it is usually lowing rail-based MRT approaches have quickly necessary to set up a new institution with new encountered a need to expand their initial people and fresh ideas (Allport, 2000). limited systems. Bangkok is a typical example; A BRT system also poses major institutional similar situations apply in Cairo, Shanghai, challenges. The need for a ‘new institution’ cited Buenos Aires, and virtually all developing cities above probably also applies to BRT in develop- which have developed rail-based MRT systems. ing cities, as the experience of Bogotá suggests. Bogota created a new institution to plan and 4.5 Speed regulate TransMilenio. Grade separated Metros, LRTs and BRTs can The scope of the challenge operate at high speeds. Street-running LRT sys- Various basic prerequisites of successful rail- tems like Alexandria-Madina (Egypt) perform based MRT projects include (Allport, 2000): less well due to interferences from street traffic Corridors with outstanding trip volume and maintenance problems. (more than 700,000 trips per day); A recent comparative study between BRT and More than 5 million inhabitants or linear LRT systems in the same city found that bus spatial development; systems on segregated bus lanes can easily At least US$1,800 per capita annual income match in terms of velocity at the city level;

24 Module 3a: Mass Transit Options

A city management with positive experience with traffic regulation; TransMilenio & the private sector Integration of other modes/fares; TransMilenio S.A., a publicly owned company Competitive fares; that provides PLANNING, MANAGEMENT, and CONTROL. A strong institutional framework; Steady population growth combined with Infrastructure is economic prosperity; developed and paid for by the local government: City center growth. • Trunk lines Even where such circumstances exist, institu- • Stations tional capacity may be insufficient for Metro • Maintenance facilities implementation in developing cities. Even • Complementary infrastructure. where corridor size, city income, growth pros- pects, city centre growth, low cost alignment, Fare collection is fares policy, city management, and Metro managed by the private management needs are met, Allport (2000) sector: • Smart cards compared the options and concluded that: • Financial management and Metros are a different order of challenge, cost and disbursements. risk…most likely to be applicable to serve the largest corridors of the biggest and more affluent Bus operations are pro- developing cities. vided by 7 concessioned private sector bus com- Institutional challenges—and associated risks panies (plus an additional and costs—are much higher for rail-based MRT 7 companies providing compared to BRT. feeder services): • System operation Role of the private sector • Bus procurement • Employee management Private sector involvement in MRT construc- • Maintenance. tion and operation can be highly beneficial to all parties, provided the government is able to establish an appropriate regulatory setting. The implementation and excessive centralisation case of Bogotá provides an excellent illustration at the level of policy-making contributing of how to successfully draw upon the private to a lack of transparency and a poor policy sector to build and operate a BRT system (see framework for making MRT investments. text box). Buenos Aires is often cited as a suc- Bus-based systems throughout the develop- cess story of concessioning of suburban rail ing world, on the contrary, are often operated services to the private sector, although in the without subsidy by the private sector, even in case of rail-based systems the situation is more a highly unconducive policy setting and poor complicated in that the government will almost and deteriorating operating conditions. Where always be required to provide an ongoing subsidy. private sector involvement is well-regulated, a In the case of Kuala Lumpur, this ongoing quality MRT service can be provided at a rela- subsidy resulted finally in the nationalisation of tively low fare, providing profit to the private the rail MRT systems in 2001. sector operators and operating without subsidy. Reasons for the failure of the private sector involvement included: Supportive policy setting Overestimation of demand; Successful MRT projects require additional Weak sectoral policies (no private car re- measures in urban transport policy. Ideally straint; poor integration with buses; no inte- infrastructural and institutional improvement grated land use and transport policies; and a will complement one another. The high capital new tollway along a similar alignment); costs of rail based MRT—and also but to a Inadequate institutional arrangements, lesser extent BRT—will not be justified if short- with both fragmentation at the level of comings in urban and transport planning offset

25 Sustainable Transport: A Sourcebook for Policy-makers in Developing Cities

Long term benefits of the benefits and harm operating conditions. in traffic-saturated developing cities like Bang- mass rapid transit Supportive policy settings include transport de- kok. In Bangkok, 10% of all BTS passengers Perhaps the major long- mand management, suitable land use planning, were previously car drivers, although there term benefit of a mass economic instruments, modal integration with seems to be such a pent-up, suppressed demand rapid transit system, rail or non-motorised transport, public awareness and that reductions in congestion are quickly ab- bus-based, is the effect it support, viable financing, and so on (see Mod- sorbed by new trips. has in concentrating a city’s development along transit- ule 3c: Bus Rapid Transit). This integrated and The smart office buildings that line the corridors accessible lines and nodes, comprehensive approach to transport planning of Curitiba’s bus system bear witness to the and resisting urban sprawl. is evident in the successful MRT cases such as positive developmental impacts of Bus Rapid Strong public transit Bogotá, Curitiba, Singapore and Hong Kong. Transit (Figure 32). Businesses locate near bus systems and transit-oriented Experience from several developing cities shows lines and stations, because of the synergies with development are an essential ingredient in any strategy to that this supportive policy setting for MRT will customer traffic. And likewise, the development reduce the level of “auto- be easier to achieve where one institutional body helps provide a critical mass of customers to dependency” of a city. provides MRT planning and regulation. make the transit system economically viable.

Cairo’s MRT reduces 4.7 Long term influence on city MRT and development pressures for urban development Mass Rapid Transit stations help catalyse new sprawl economic and employment opportunities by MRT and city form This is evident for example in acting as nodes of development. Importantly for land use patterns and transit- Cairo, Egypt, where an This has been the experience in Bogotá, with ris- impressive 60 km heavy rail friendly development, nearly all MRT systems ing land values in the vicinity of TransMilenio metro network along major enable continuing city centre growth. A mass stations and strong demand from land-owners corridors now carries 20% transit system is an indispensible aspect of a of all motorised passenger and businesses for the construction of stations sustainable transport system for a large city, and trips in Greater Cairo. in their local areas. Bogotá implemented an in developing countries can play an important Without the metro network, innovative value capture scheme in which the north-south corridors and role in shaping future development of the city, windfall benefits to landowners in the form of the city centre would have leading to a transit-friendly city form. been overwhelmed by con- rising land values was partially diverted to help gestion, and development It may, however, be unrealistic to expect major fund the construction of the stations. would have been forced reductions in road congestion in developing Rail-based MRT systems can have similar into peripheral areas much cities. MRT infrastructure projects have only effects, though in the case of bus and rail the earlier. minor impacts on car ownership and use. Car (Metge, 2000). government plays a crucial role in promoting ownership is generally more influenced by development around stations and along routes. parking space supply and ownership costs rather than by MRT supply. This applies particularly However at the city-wide level the effects on city structure will be weaker than hoped for when unrestricted car use and weak building laws encourage urban sprawl and lower urban densities. Hong Kong’s success, for example, results both from a well-designed and highly productive MRT-system and an enforced policy of high-density residential or commercial areas around the stations. In Paris the concept of five edge cities was fostered by the implementation of a heavy rail system (RER) linking these edge cities with the centre of Paris. In the city centre Fig. 32 the RER is integrated with the underground Curitiba’s 5 BRT lines network. However even in Paris, where the city are lined with high centre is served by an excellent public transport density apartments, system, car use has been increasing and densi- offices and commercial developments. ties falling, due to the lack of a policy of strong Karl Fjellstrom, Feb. 2001 restriction of car use.

26 Module 3a: Mass Transit Options

4.8 Poverty alleviation MRTs: Poor service for the urban poor? In the World Bank Urban Transport Strategy Review, Allport (2000) points to a ‘dilemma’ in We should not assume low fares are the most important MRT policy for developing cities: factor for low income users At the centre of MRT policy for developing cities of public transport in deve- is the apparent conflict between tackling poverty loping cities. Surveys in the alleviation, for which affordable service is critical, Indonesian cities of Den- pasar and Surabaya, for and attracting car users, for whom service quality example, have revealed that is critical. factors such as reliability, Experience with BRT, and with quality bus personal safety, frequency, speed, and comfort (espe- services in general, show this may be a false cially not being cramped) are dilemma. Cases such as Curitiba, Bogotá, Sao often rated as more impor- Paulo, and Quito show that BRT systems tant than low fares. in developing cities can provide an excellent Secondly, it may be mis- service popular with high and low income users, taken to assume that a high quality MRT system would and be profitable at a low fare. In comparison, necessarily be priced beyond rail systems provide a more limited geographical Fig. 33 the reach of poor users. coverage—especially for poorer people relying A typical low income area of Cairo. Paratransit High quality BRT systems on road-based transit (see Figure 33). provides a feeder service to the Metro terminus. in developing cities can Karl Fjellstrom, Mar. 2002 operate at a low fare. One of Mass Rapid Transit can play an important role the successes of Bogotá’s in alleviating—or exacerbating—poverty. It 4.9 Environmental impact BRT is seen in its socially is the poorest people who most depend upon integrating effect, with rich Energy use by different transport modes, and poor rubbing shoulders public transit for access to jobs and services. In which is closely related to emissions, is in the bus. In many ways it is some cities the urban poor pay up to 30% of presented in Table 11. Rail is the most envi- a social experiment, not just their income on transport. The poor also typi- ronmentally friendly type of MRT in terms of a MRT system. cally live in lower rent areas on the outskirts energy use per person-kilometre, though only of the city (see Figure 34), and in some cases spend two to four hours commuting each day. Most importantly, public funds which are not poured into road-building and rail can be spent on improving health, education, public space, and quality of life of the urban poor. Concentrating on the transport modes of poor people calls for the provision of affordable forms of public transport, although public transport should not be viewed as only for the poor, as wealthy European and Asian cities show. Large cities in the developing world are centres of economic growth and magnets for poor people from the countryside, who often settle in the outskirts and along traffic arteries. They are heavily affected by noise and pollution. Improved transit possibilities will provide faster Fig. 34 access to work-places and enable more people to Miami, Buenos Aires, Paris... The rail-based MRT systems of Sao Paulo work. The MRTs in Cairo, Mexico, Bogotá and probably seem as inaccessible as the cities advertised on the billboards to the urban poor living on the outskirts of Sao Paulo. Bus Rapid Transit, with elsewhere are used extensively by poor riders its potentially greater geographical reach, offers more hope to low income who profit from quick access to the city centre communities on the outskirts of all developing cities. and hence additional employment possibilities. Karl Fjellstrom, Feb. 2002

27 Sustainable Transport: A Sourcebook for Policy-makers in Developing Cities

Table 11: Energy use per passenger Table 12: Trends in public transport use in kilometre, various modes and operating an international sample of cities, 1970 to conditions the mid 1990s Armin Wagner, 2002, from various sources Barter 1999; GTZ SUTP

Energy use per Percent of all motorised trips by public System passenger-km transport [Watt-hours] City 1970 1980 1990 ’93–’96 Bicycle (20 km/h) 22 Highly occupied Metro-systems Tokyo 65 51 48 ? 79 (Tokyo, Hong Kong) Hong Kong ? Buses (Khartoum, Sudan) 99 Seoul 81 74 63 ? Buses (Occupancy 45%) 101 Singapore 42 ? ? 51 Paratransit 184 Manila ? 70 67 70 (Mini-Bus, Khartoum) Bangkok 53 ? 39 ? Less occupied Metro systems 184–447 such as Germany Kuala Lumpur 37 33 32 24 Metro (occupancy 21%) 240 Jakarta 61 58 52 53 Paratransit (occupancy 67%/ Surabaya ? 36 35 33 317 Minibus/Aleppo (Syria)) Rail-based systems USA selection of cities. There are some exceptions in 577 (22,5 passengers per unit/USA) cities which have experienced increasing shares Buses (8,9 passengers/USA) 875 of passenger-kilometres by transit (e.g., Zürich, Vienna, Washington and New York: WBCSD, where occupancy is very high. Emissions vary 2001) and increasing transit modal shares greatly depending on the power source used to (e.g., Singapore), but in general the trend is for generate electric traction (for rail), and the bus declining transit modal shares of around 1–2% and fuel technology in a BRT system. In addi- per year in large cities. tion, not all developing nation rail systems are electrified, and thus there are sometimes local In the longer term, then, the MRT systems emission impacts. which can be expected to have the best envi- ronmental impact are those which can halt or From an environmental perspective, however, reverse the declining modal share of public the main point to note is that virtually all transport. In the case of lower income develop- MRT systems offer environmental advantages ing cities such an impact on overall modal to the extent that they replace trips by private share in the city is probably possible only with motor vehicles. Perhaps most important in the bus-based MRT, rather than rail. Due to the long term, in terms of reducing emissions, is larger cost, new rail systems can be developed in the impact of a MRT system on the modal only very limited areas of a developing city, and split, or percentage of people travelling by do not have the capacity of BRT to reach and public and private transport modes. In this cover larger areas, or the flexibility to adapt to a regard experience shows that in developing changing and expanding city. cities it is the BRT systems such as Bogotá and Curitiba that have enabled public transit In terms of air quality the crucial factor in to maintain or even increase modal share developing cities is not so much the emission compared to private transport. In other cit- performance of the different MRT modes, but ies public transit has tended to decline, with rather their potential in getting people out corresponding negative environmental impacts of cars and off motorcycles, and into transit. not just in terms of local pollutant emissions, To the extent that a BRT system can do this but also in terms of greenhouse gases, noise, better than a rail system (with much more and visual intrusion. Table 12 describes the limited coverage), BRT has a greater positive progressive decline of public transport in a environmental impact.

28 Module 3a: Mass Transit Options

5. Conclusion

After comparing MRT options, in general we can conclude that there are few reasons for developing cities to favour rail-based systems where passenger capacities would be less than 25,000 passengers per hour per direction. Unless specific circumstances apply—such as when visual image of the system is quite important and a city is sufficiently wealthy to handle the higher capital and operational costs—this kind of rail-based transit for devel- oping cities compares unfavourably with BRT systems on most terms, and especially for key parameters such as cost, flexibility, time frame, and institutional demands. There is however no single “right” transit solu- tion. The best system for a city will depend on local conditions and preferences and will involve “Think rail, use bus.” a combination of technologies. Bus Rapid Photo Karl Fjellstrom, 2002 (Curitiba, GTZ Transport Photo CD) Transit may not be the solution in every situa- tion. When passenger flows are extremely high and space for busways is limited, other options may be better, such as rail-based public transit; although we have seen that BRT can accommo- date passenger volumes to match demand even in very large cities. In reality, it is not always just a choice between bus and rail, as cities like Sao Paulo, Brazil have shown that Metro and BRT systems can work together to form an integrated transport package. It must however be recalled that city invest- ments in Mass Rapid Transit systems come at a high opportunity cost. Funds used to build and subsidise the operation of a limited Metro could be used for schools, hospitals, and parks. Bus Rapid Transit has shown that high quality public transit that meets the needs of the wider public is neither costly nor extremely difficult to achieve. Many organisations are ready to help municipalities in developing cities make efficient public transport a reality. With political leader- ship, everything is possible.

Karl Fjellstrom, Jan. 2002 (Shanghai’s Hengshan Rd. Station)

29 Sustainable Transport: A Sourcebook for Policy-makers in Developing Cities

Resource materials Transit Cooperative Research Program (TCRP), Transit Capacity and Quality of Roger Allport, Urban Mass Transit in Devel- Service Manual, Kittelson & Associates, oping Countries, Halcrow Fox, with Traffic 1999, http://www.trb.org (many excellent and Transport Consultants, 2000, http:// reports available for download) wbln0018.worldbank.org/transport/utsr.nsf Thomson, I., UN Economic Commission W.S. Atkins, Study of European Best Practice for Latin America and the Caribbean (UN- ECLAC), The Impact of Social, Economic and in the Delivery of Integrated Transport, Sum- Environmental Factors on Public Transport in mary Report, Nov. 2001, http://www.cfit.gov. Latin American Cities, International Seminar uk/research/ebp/exec/index.htm on Urban Transport, Nov. 2001, Bogotá, Jason Chang, Taipei Bus Transit System and Colombia Dedicated Bus Lane, International Workshop World Bank, Cities on the Move: An Urban on High Capacity Bus Systems, New Delhi, Transport Strategy Review, 2001, http://www. India, 20 Jan. 2002. worldbank.org/transport Robert Cervero, The Transit Metropolis: A Global Enquiry, Island Press, 1998. United States General Accounting Office (GAO), Bus Rapid Transit Shows Promise, Re- port to Congressional Requesters, Sept. 2001, http://www.altfuels.com/PDFs/GAOBRTs- tudy.pdf. Gregory Ingram, World Bank, Patterns of Metropolitan Development: What Have We Learned?, Urban Studies, Vol. 35, No. 7, 1998 International Finance Corporation (IFC), Bangkok Mass Transit (Skytrain) Externalities Study, prepared by Policy Appraisal Services et al., July 2001 (unpublished) Alexandre Meirelles, A Review of Bus Priority Systems in Brazil: from Bus Lanes to Busway Transit, Smart Urban Transport Conference, Brisbane, 17–20 Oct. 2000. Hubert Metge, The Case of Cairo, Egypt, World Bank Urban Transport Strategy Re- view, Nov. 2000, http://wbln0018.worldbank. org/transport/utsr.nsf Peter Newman & Jeff Kenworthy, Sustain- ability and Cities: Overcoming Automobile Dependence, Island Press, Washington, 1999. Philip Sayeg, Smart Urban Transport Maga- zine, 2001, http://www.smarturbantransport. com David Shen et al., At-Grade Busway Planning Guide, Florida International University, Dec. 1998, http://www.cutr.usf.edu/research/nuti/ busway/Busway.htm

30 Module 3a: Mass Transit Options

Urban Transport Strategy Review Reference on choice of MRT The World Bank’s Urban Transport Strategy Review includes a report which, like this module, offers advice on approaching Mass Rapid Transit options in developing cities. Urban Mass Transit in Developing Countries (Roger Allport, Halcrow Fox, with Traffic and Transport Consultants, 2000), includes an excellent discussion of the impacts, challenges and risks of rail-based projects, although in general it fails to draw out the experience of ‘world best’ Bus Rapid Transit applications such as Bogotá, since it was released only months after the TransMilenio system began operation. Major sections of the report include: MRT options MRT role Research results Scale of challenge Attitudes to MRT Forecasting MRT impacts Planning for tomorrow The private sector approach Affordability and the private sector Public transport integration Economic viability Poverty alleviation Land use and city structure The environment MRT planning Implementation and operations This report can be obtained downloaded free of charge at the Urban Transport Strategy Review web site, http://wbln0018.worldbank.org/trans- port/utsr.nsf Many more online resources on MRT topics can be obtained through the Univ. of Nottingham’s Sustainable Urban Travel: Comprehensive bi- bliography lists of relevant contacts, addresses, and worldwide Websites, http://www.notting- ham.ac.uk/sbe/planbiblios/bibs/sustrav/

31 Dag-Hammarskjøld-Weg 1-5 P. O. Box 5180 D - 65726 Eschborn Germany Telefon +49-6196-79-1357 Telefax +49-6196-79-7194 Internet: http://www.gtz.de