Document of I , P The World Bank FOR OFFICIAL USE ONLY Report No. 2759b-BR STAFF APPRAISAL REPORT SECOND URBAN TRANSPORT (PORTO ALEGRE BRAZIL April 17, 1980 Projects Department Latin America and the Caribbean Regional Office This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. Currency Equivalents Currency Unit Cruzeiro (Cr$) divided into 100 cents Exchange Rates: Average for Year 1975 US$1 = Cr$ 8.10 1976 US$1 = Cr$ 10.70 1977 US$1 = Cr$ 14.15 1978 US$1 =Cr$ 18.10 1979 US$1 = Cr$ 26.00 1980 US$1 = Cr$ 43.00 (January 1980) Fiscal Year Jaruary 1 to December 31 Units of Weights and Measures: Metric Metric: British/US Equivalent 1 kilometer (km) - 0.62 mile (mi) I meter (m) 2 - 3.28 feet (ft) 1 square kilometer (km ) = 0.386 square mile (sq mi) 1 hectare (ha) = 2.47 acres (ac) 1 metric ton (m ton) = 0.98 long ton (lg ton) 1 metric ton (m ton) 1.1 IJS short tons (sh ton) 1 liter 0.2642 gallons Abbreviations and Acronyms BNDE National Development Bank BNH - National Housing Bank CED Centr.tal Business District of the Municipality of Porto Alegre CEEE State Electric Cou..pany of Rio Grande do Sul CIP - Interministerial Council on Prices COHAB - State Housing Company CNDU - National Council on Urban Development CNPU - National Commission on Metropolitan Regions and UJrban Policy CTC - Centralized Traffic Control DEMHAB - Porto Alegre Housing Agency DNER - Federal Highway Department DNOS - National Department for Sanitation EBTU - Brazilian Urban Transport Corporation EMTU - Metropolitan Transport Agency FDTU - Urban Transport Development Fund FNDU - National Urban Development Fund GEIPOT - National Transport Planning Agency METROPLAN - Metropolitan Planning Agency of Porto Alegre MRPA - Metropolitan Region of Porto Alegre NEP - Project Implementing Office of TRENSURB/P.A. NMTU - Agency for the Coordination and Regulation of Transportation in MRPA NPV - Net Present Value PCU - Passenger Car Units PME - Energy Mobilization Program RFFSA - Federal Railway Company SEAP - Secretariat for Supply and Prices in the Ministry of Planning SEPLAN - Ministry of Planning TRENSURB/P.A. - Porto Alegre Metropolitan Rail Mass Transit Company BRAZ IL FOR OFMUL USE ONLY SECOND URBAN TRANSPORT (PORTO ALEGRE) PROJECT STAFF APPRAISAL REPORT TABLE OF CONTENTS Page-No. I. SECTOR BACKGROUND ... . ...... 1 A. Urban Structure and Trends .......................... 1 B. Urban Policy ........... .............................. . .......... I C. Urban Transport ....... ............................. 3 D. Bank Involvement in the Urban Sector ................ 7 II. AREA OF INFLUENCE - PORTO ALEGRE METROPOLITAN-REGION ..... 7 A. Geography .......................................... 7 B. Demographic and Economic Base .............O.......... 8 C. Distribution of Income, Housing and Public Services. 9 D. Metropolitan Planning and Coordination .............. 10 E. Prognosis for the Metropolitan Region .............. 11 F. Transport Sector ........... . 11 G. Travel Demand Forecasts ............................. 15 H. Train Ridership ................... .................. 16 III. PROJECT OBJECTIVES, DESCRIPTION AND COST ESTIMATES ....... 17 A. Project Objectives ............ . ..................... 17 B. Physical Works and Equipment...................... 18 C. Complementary Institutions and Policies ............. 21 D. Technical Assistance ............................... . 22 E. Cost Estimates ....... ........ .. ......... ......... 24 IV. PROJECT IMPLEMENTATION AND THE PROPOSED LOAN AND-BORROWER 24 A. Project-Implementing Office ........................ 24 B. Implementation Schedule . . ... .. ....... 25 C. The Proposed Loan. .... . .............. .......... 25 D. The Borrower, Onlending and Project Agreements ..... 26 E. Procurement . .......................... ........ ... 27 F. Disbursements . . . ........ . 27 G. Financing Plan . . . ........... . 27 H. Project Risk ................................ ........... 28 This report is based on the findings of an appraisal mission which visited Brazil during September-October 1979. The mission comprised Messrs. M. Staab (Economist), J. Baigorria and C. Buratti (Engineers), E. Haythorne (Lawyer), R. Mosse (Financial Analyst) and C. Turner (Consultant),. The report has been edited by Miss V. Foster. This documont has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be qisclosed without World Bank authorization. TABLE OF CONTENTS (Continued) Page No. V. THE NEW METROPOLITAN RAIL MASS TRANSIT COMPANY ........... 28 A. Ownership, Organization and Financial Policy ....... 28 B. Operations ...... ......... .. ........... 30 C. Accounts ................. O* ...0..... 0........................ 31 D. Assets and Liabilities . ............................ 32 E. Operating Costs ......................... ............ 32 F. Tariffs .......... ..... ........................... 32 G. Financial Results ....... ..... ... ......... O.. 34 VI. ECONOMIC ANALYSIS .................... .... .. ... 34 A. Alternatives .......... . 34 B. Risk and Sensitivity Analysis ..................... 39 C. Distributional Impact ........................... 39 VII. AGREEMENTS REACHED AND RECOMMENDATION .................. 40 TABLES 1.1 Urban and Regional Population Size in Brazil 43 1.2 Motorization Rates in Selected Urban Areas . .44 2.1 Population by Municipality in the MRPA. 45 2.2 The Location of Population Within the MRPA . .46 2.3 Sectoral Distribution of Employment in the MRPA 47 2.4 Spatial and Sectoral Distribution of Employment in the MRPARPA...... 48 2.5 Distribution of Household Income by Municipality in the MRPo..............A.... .. 49 2.6 Road Transport Infrastructure in the MRPA . .50 2.7 Number, Type and Fleet Size of Municipal and Intermunicipal Bus Companies in the MRPA 51 2.8 Vehicle Occupancy, Trip Distance, Speeds and Fares by Mode in the MRPA. 52 2.9 Vehicle Registration by Municipality in the MRPA 53 2.10 The Availability and Cost of Parking in the Porto Alegre CBD . .54 2.11 Private Automobile Owner/User Charges in the MRPA 55 2.12 Modal, Purpose and Time Distribution of Passenger Trips in the MRPA. 56 2.13 Municipal Shares in Trip Generation in the MRPA 57 2.14 Daily and Peak-Hour Vehicle Traffic Levels and Growth Rates on BR-116 in the MRPA 58 2.15 Station-to-Station Daily Passenger Traffic - 1985 59 2.16 Station-to-Station Daily Passenger Traffic - 1990 60 2.17 Station-to-Station Daily Passenger Traffic - 2000 61 TABLE OF CONTENTS (Continued) Page No. 3.1 Train Operations and Operating Costs (3 pages) .e 62-64 3.2 Permanent Way Maintenance Equipment .65 3.3 Depot Maintenance Equipment (2 pages) .66-67 3.4 Technical Assistance Program .68 3.5 Project Cost Estimates .69 4.1 Project Implementation Schedule .70 4.2 Items to be Financed by the Proposed Loan (2 pages) 71-72 4.3 Estimated Disbursement Schedule .73 5.1 Forecast of TRENSURB's Financial Statements .74 5.2 Financial Sensitivity Analysis .75 6.1 Economic Analysis - Alternative I .76 6.2 Economic Analysis - Alternative II .77 6.3 Economic Analysis - Alternative III .78 6.4 Economic Analysis - Alternative IV .79 6.5 Economic Analysis - Alternative V .80 6.6 Economic Analysis - Base Case .81 6.7 Economic Road Vehicle Operating Costs .82 6.8 Economic Rail Operating Costs .83 6.9 Passenger Travel Time Costs .84 ANNEXES 1. Description and Status of First Urban Transport Project .................................. 85 2. EBTU's Organization, Resources and Finances .......... 87 3. Assessment of the Development Master Plan for the MRPA .92 4. Land Use Alternatives and Travel Demand Forecasts .94 5. Selected Documents and Data Available in the Project File ...................................... 100 CHART 1. Organization of EBTU .101 MAP IBRD 14688 - Brazil - Second Urban Transport Project - Porto Alegre Metropolitan Area I. SECTOR BACKGROUND A. Urban Structure and Trends 1.01 Brazil's urban growth during the period 1950-1970 was quite rapid, with the country's urban population (as measured by an international standard definition of 20,000 or more inhabitants) increasing from 11 million in 1950 to 36 million in 1970, an increase of about 6% per annum compared with a figure of about 3% per annum for the overall rate of population growth in the country. The proportion of total population that is urbanized increased from a level of only 16% in 1940 to nearly 40% in 1970 (Table 1.1). The highest growth rates in the 1960s were recorded in medium-sized cities of between 100,000 and 200,000 inhabitants. During the 1970s, these trends are believed to have continued, but definitive data will not be available until the 1980 census is completed. 1.02 Geographically, urbanization has been unevenly distributed. Growth rates have been highest in the central western region, primarily because of the growth of Belo Horizonte and Brasilia, and lowest in the northeast region despite the rapid growth of its three metropolitan areas: Fortaleza, Recife and Salvador. The dominance of the southeast (mainly Sao Paulo and Rio de Janeiro) in the urban population has been offset somewhat by the growth of cities in the central west and in the state of Parana in the south, although not sufficiently to alter the large imbalance between the southeast and other regions in the ratio of urban to total population. 1.03 The rapid rate of population growth in the large metropolitan areas is mostly explained by migration from the interior. Large numbers of this migrant labor force, which is relatively poor and lacking in industrial skills, are forced to seek low paying jobs, sometimes in the informal sector. Municipal governments, therefore, are faced with the problem of supplying essential public services to an expanding urban population, while the tax base to support these services is not growing nearly as fast. These trends have become of increasing concern to urban planners. B. Urban Policy 1.04 Federal urban policy was first directed at the large metropolitan areas. In order to deal with the problem of coordinating urban policy in these areas, the Government, in 1973, established the administrative boundaries for eight metropolitan areas: Belem, Fortaleza, Recife, Salvador, Belo Horizonte, Sao Paulo, Curitiba and Porto Alegre. The ninth, Rio de Janeiro, was added in 1974. The states in which these metropolitan areas are located have all established metropolitan planning agencies, the responsibilities of which range between those which are purely planning agencies (e.g., Porto Alegre) and those which also fund and execute projects (e.g., Recife). However, none of these state agencies has a government status. The hierarchy of governments in Brazil goes from the municipal to the state to the federal level. There is no metropolitan government as such, which sometimes makes coordination among municipalities difficult. 1.05 A national urban policy was initiated with the Second National Development Plan (II PND), which covered the period 1975 to 1979. The urban development strategy contained within the II PND sought to: - 2 - (a) consolidate the system of metropolitan areas and introduce measures to prevent the additional concentration of demo- graphic and economic growth in the southeast region, par- ticularly in Rio de Janeiro and Sao Paulo; (b) encourage the development of medium-sized cities (between 50,000 and 500,000 inhabitants) in all regions to reinforce an existing development trend; and (c) encourage the development of settlements in the interior. Overall, federal government actions have been consistent with these objectives regarding the pattern of urbanization. Nevertheless, despite these efforts, the large metropolitan areas continue to attract many migrants - a trend which is expected to persist unabated through most of the next decade. State governments will have to increase their efforts if the national objectives are to be substantially realized. The urban strategy contained in the recently formulated Third Development Plan (1980-1985) further promotes these objectives, but, in addition, emphasizes the development of agriculture to reduce the overall rate of urbanization. 1.06 The institutional structure exists for carrying out the federal government's urban policy. In 1974, the National Commission on Metropolitan Regions and Urban Policy (CNPU) was set up as an interministerial agency to promote and coordinate national urban development policy. To finance urban development projects, the National Urban Development Fund (FNDU) was also created. A special Urban Transport Development Fund (FDTU) was established as a sub-account of FNDU to specifically finance urban transport projects. The Brazilian Urban Transport Corporation (EBTU) was later established to administer the FDTU (para 1.15). 1.07 Following the change of government in March 1979, CNPU was termi- nated and the new government created the National Council on Urban Develop- ment (CNDU). The Ministries of Planning, Finance, Industry, Commerce, Transport and the Interior are represented on the Council, together with the Secretary General of the Ministry of Planning (SEPLAN) and the presidents of the National Housing Bank (BNH) and EBTU. The Council is chaired by the Minister of the Interior. CNDU has assumed all of the responsibilities of the former CNPU and has recently defined the nation's urban policy objectives for the period 1980-1985. Consistent with the objectives of the II and III PND in seeking to redirect migration from the metropolitan areas to the secondary and smaller cities, CNDU is proposing that the nine metropolitan areas, medium- sized cities and small cities receive about 50, 40 and 10% of the available federal funds, respectively, over the period 1980-1985. This allocation would be effected and monitored by the respective executing agencies, namely EBTU, BNH and the National Department for Sanitation (DNOS). 1.08 There are, however, several reasons to be concerned about the probability of the above policy objectives being successfully implemented. First, although responsible for the overall coordination and integration of urban policy, it has yet to be seen whether CNDU will be able to exercise any influence on urban area policies other than through an advisory role. Specifically, the metropolitan area master plans are not approved by CNDU, but by the respective states. The metropolitan areas send their master plans to CNDU only after approval by the state. CNDU's role is essentially to review the various sector investment items within the master plans in order to coordinate any possible involvement of federal sector agencies. CNDU is -3- currently studying the need for new legislation that would make participation of federal sectoral agencies in a given metropolitan area conditional upon CNDU's approval of the master plan. If such legislation is adopted, CNDU could use any newly acquired influence to try to strengthen the planning agencies at the local level rather than to direct their activities from Brasilia. However, to be successful in performing even this limited additional function, CNDU, with only about 20 professional staff at present, would have to recruit more qualified personnel. C. Urban Transport (i) General 1.09 An overview of urban transport in Brazil reveals certain problems, patterns and changes which are common to many of the country's urban areas. During the last several years, most urban areas have experienced large in- creases in the demand for transport services. In the nine metropolitan regions, the number of daily passenger trips by all modes increased by 9% per annum between 1968 and 1975, almost twice the rate of growth of the urban population. As previously mentioned (para 1.03), a large portion of the popu- lation growth in these areas is explained by in-migration from the rural areas within the same state and, in some cases, from the neighboring states as well. From a transport point of view, what is important is that this migrating population, which is generally of low income, tends to locate on the periphery and non-central parts of the metropolitan region, where housing is cheaper but where transport service (including access to job market centers) as well as other public services is more limited. In contrast, transport infrastructure tends to be concentrated in those parts of the metropolitan region which are closest to the central city and which link centers of employment with the residences of higher and middle income groups whose principal mode of transport is the private automobile. It is no coincidence that the transport infrastruc- ture serving these areas tends to be primarily road based, further encouraging the use of private automobiles, normally to an extent well beyond the space lim- itations of the central area, resulting in high economic and environmental costs. 1.10 This situation is reflected in the usage and characteristics of the principal urban transport modes of bus and automobile: the users of the former being largely of low income and traveling longer commuting distances to work from areas that are not particularly well served by public transport (bus or other modes); and the users of the latter being essentially of high or middle income and traveling shorter commuting distances on roads for whose limited space buses must compete with the automobile. In the last few years, however, bus services in certain corridors in some urban areas have been given priority through the introduction of exclusive bus lanes. This approach is an important part of the First Urban Transport Project (Loan No. 1563-BR), the scope and status of which are discussed in Annex 1. For the most part, however, in Brazil, as in many other countries, this type of change is relatively new. 1.11 Urban rail transport is a significant mode of transport only in Rio de Janeiro and Sao Paulo, where each of the systems carries about 500,000 passengers,daily. Very limited suburban railway services also exist in Belo Horizonte, Salvador and Recife. The service in Porto Alegre (the site for the proposed project) is virtually non-existent with only two commuting trains daily in each direction on a deteriorated track system (para 2.20). - 4 - 1.12 In 1976, the modal distribution of total daily person trips in the nine metropolitan regions ranged as follows: for buses, from 71% in Recife and Belo Horizonte to 54% in Sao Paulo; for autos, from 35% in Sao Paulo to 22% in Rio de Janeiro; for taxis, from 7.6% in Belem to 1.6% in Porto Alegre; and for person trips by other modes, 8% in Sao Paulo to 1.0% in Curitiba. While the bulk of demand is carried by buses, their share of traffic has been falling at the expense of the car. Moreover, in terms of the number and composition of vehicles during the peak period, which determines road capacity, cars in many urban corridors represent as much as 75% of the total number of vehicles on the road despite the fact that they may account for only 30% of the total passengers carried. Since 1965, the automobile fleet has grown at an average rate of 11% per annum. In 1975, private vehicle ownership in the nine metropolitan regions reached 2.3 million vehicles, resulting in an average motorization rate of 80 automobiles per 1,000 inhabitants. A comparison of motorization rates in selected urban areas is shown in Table 1.2. Clearly, a reduction in the use of the private automobile in urban areas and a correspond- ing increase in the use and provision of public transport services should be and, in fact, is a fundamental objective of national urban transport policy. This objective is important not only because of the attendant fuel savings, which would be significant, but because, in many cases, lack of public trans- port access to job centers more than proportionately affects the urban poor. (ii) Coordination and Planning at the Local Level 1.13 The extent and quality of urban transport coordination and planning vary among the different regions but all are faced with a similar problem-- how to plan and manage a transport system in a metropolitan region consisting of many municipalities, each with its own administrative powers, when the necessary inputs for, and effects of, this transport system extend well beyond the administrative boundaries of any one municipality. A possible solution would be to create a metropolitan-wide transport coordinating/regulatory agency, which sometimes requires a state law. In some Brazilian metropolitan areas, adaptations of this type of institution already exist, and, in others, plans (or, in some cases, draft legislation) exist to create them. These organizations are sometimes referred to by the name Empresa Metropolitana de Transportes Urbanos (EMTU). In addition to regulation and coordination, some of the existing organizations are also involved in the operation of some trans- port enterprises. This practice has a high probability of creating a conflict of interest, and it is, in principle, undesirable for such organizations to become directly involved with the ownership or operation of transport enter- prises. An entity, called Nucleo Metropolitano de Transportes (NMTU), with only coordinating and regulatory responsibilities, was recently created (December, 1979) by the State Governor of Rio Grande do Sul for the Porto Alegre Metropolitan Area, the influence area for the proposed project (para 2.27 through 2.29). (iii) National Planning, Investment and Policy 1.14 Federal government involvement in urban transport is accomplished directly through an organization specifically established to serve in this capacity, the Brazilian Urban Transport Corporation (EBTU), and indirectly through the government's support to the Federal Railway (RFFSA) and the Federal Highway Department (DNER), both of which, to some extent, invest in urban - - trans,og-rt jects. However, investments of this type by- RFFSA and DNER are expected, ia th£e f£ture, to become much smaller sirice present policy, as much as possibie, is to channel all federal spending in this sector through EBTU. 1.15, EBTIJ was established by law in November 1975. It is a federal government-owned enterprise under the jurisdiction of the Ministry of Trans- port but having administrative and financial autonomv.. independent juridical staLtus and its o. assets and iiabilities. The organizational structure of EBTU is shown in Chart 1. EBTU agreed during negotiations to employ consuLtants, by December 31, 1980, to improve its internal accounting, budgeting and control procedures and to discuss with the Bank the consultants' recommendationsi The details regarding EBTU's organizational structure, financial perf?ormatnce, budget, accounting and audit systems are presented in Annex 2. The Company's basic function is to promote national urban transport policies through its support of certain projects and programs which it deems to be in this interest (EETU's relation to and support of the proposed project is presented in paragraphs 4.07 and 4.10). To finance its expenditures, EBTTJ has in the past drawn its resources almost entirely from the Urban Transport Development Fund (FDTU) (para 1.06), which is derived from vehicle registration and fuel taxes. Commencing in 1980, EBTU will be able to draw on an additional major source of funds from the Energy Mobilization Program (PME), recently created under a new law. 1/ Under this new financing scheme, beginning in 1980, EBTU will receive about 65% of the PME plus the FDTUi 2/ 1.16 The federal government has already taken some fairly progressive action with regard to the conservation of petroleum resources in the trans- port sector by promoting the alcohol fuel program (at present, automotive gasoline consists of a 20% mixture of alcohol, and all new government cars and taxis will use 100% alcohol), by eliminating the subsidy element from gasoline prices and by raising taxes on petroleum products quite substantially. EBTU, during the first three years of its existence (1976-1978), oversaw an estimated US$884.6 million (1979 prices) of expenditures on various urban transport projects. The components selected generally fit within a set of categories that broadly reflect EBTU's policy goals of promoting public transport and the efficient management of the urban transport system. However, EBTU has not had sufficient expertise to convert general national guide'lines into an integrated, high priority investment program. For this type of work, there are few experienced professional economists or policy- oriented urban transport planners on EBTU's current staff. The staff, just over 300, consists mostly of engineers and architects, the effect of which, to date, has been to impart an overly physical orientation to the planning function. While the quality of economic wqork is beginning to improve, further improvement is necessary. Under the First Urban Transport Proiec; . i t was agreed that EBTU would engage consultants to, among other things, strengthen this area of its operation. Some consultants have '1 LTaw 1691 of August 2, 1979. The PME is derived from a new surtax of 12.5%7 on the CIF value of imported crude oil plus 26% of the revenue from the existing vehicle registration tax. 2/ T'he FDTU, beginning in 1980, will be composed of 12% of the existing vehicie registration tax and 45% of the existing surtax (12%) on lubri- cauts and liquid and gaseous fuels. -6- been working with the tecLrincal directoraLe, arnd nor , io engaged shortly. This process is having a positive impact. However, addi- tional experienced, professionally trained economists and planners are needed to occupy staff9 as well as managerial, positions w-hin -he organi: In order for the technical directorate to begin to play a more criticai .-c-e in EBTU's investment decisions, improvements in the quality and quantity of tech- nical staff are essentiai. Having EBTU rely more heavily upon its technical directorate is an orientation that the Bank has continued to promote. (iv) Future Investment Plans 1.17 EBTU's investment plan for the next three years (1980-1982), which includes the proposed project, was recently formulated by the Ministry of Transport in order to make use of the resources which are expected to become available from the PME (para 1.15). The investment plan includes items which EBTU had already identified and was in the process of preparing plus some new items introduced by the Ministry of Transport, primarily in the urban rail subsector, which are expected to save fuel. The average annual expenditure of this plan (US$883 million) is about three times the amount spent per annum during the period 1976-1978. The plan's overall objective is the conservation of energy through heavier reliance on public transport. This is the general rationale; however, benefit-cost analyses, for the most part, have not yet been carried out except for a few projects, one of which is the proposed project for Porto Alegre. Given the large expenditures associated with the plan, which, if realized, would involve significant changes in intersectoral priorities that have not yet been promoted by the Government, the plan is still likely to undergo considerable change in the near future, especially after the results from further study become available. 1.18 The modal distribution of expenditures under the plan is as follows: suburban railways in seven metropolitan areas, 59%; urban roads including buses, 25%; trolley-buses where these already exist, 11%; and marine passenger ferries, 5%. This modal distribution is significantly different from EBTU's earlier plan, which emphasized road-related public transport investments. The change is largely based on the Government's desire to save fuel and, through the reduction of petroleum imports, also to mitigate its serious balance-of- payments problems. As desirable as these objectives are, they represent only a partial view without complete technical and economic studies. Furthermore, for the rail projects identified, it has not yet been demonstrated how these particular objectives are served. The suburban railway investment subprogram needs further careful scrutiny also because of the general policy of the Ministry of Transport to convert all diesel train services to electric traction, to substitute broad for all meter gauge track and to build freight bypasses, when necessary, to separate freight completely from suburban passenger trains. While, under some circumstances, such actions may be justified, as has been demonstrated in the case of the proposed project in Porto Alegre (Chapter VI), it would not be economically sound to regard this approach as having universal applicability. 1.19 In view of the foregoing, it would be desirable for the Bank to participate in discussions with the Government as the national urban transport investment plan begins to be firmed up and more concrete decisions are taken. -7- EBTU has already agreed, under the First Urban Transport Project, to adopt, by June 1980, economic and technical criteria for selecting projects. While it already appears that there may be some delays in meeting this target, and, in any case, some further work will need to be done on the criteria for selec- ting urban rail projects, work in this area is proceeding. What remains to be arranged is the review process by which the Bank can have some positive impact on the formulation of the urban transport investment plan. During negotiations, agreement was reached that, at least once per year, the Government, through EBTU, would inform the Bank of its National Urban Transport Investment Plan. These reviews would provide the Bank an opportunity to exchange views with the Government on the Plan. A draft investment policy statement by EBTU was also discussed during negotiations. This statement is a quite useful begin- ning and provides a sound general framework within which to develop more specific guidelines. The investment policy statement was formally approved by EBTU's Administrative Council. It was agreed during negotiations that, before making any substantial change in this statement, EBTU would review and discuss such changes with the Bank. D. Bank Involvement in the Urban Sector 1.20 The Bank's involvement in the urban sector focuses on two main objectives: the provision of needed services and employment opportunities for the urban poor, and the strengthening of the planning and implementation capacity of sectoral institutions at the federal and local levels. Beginning in 1970, the Bank participated in the financing of a series of water supply and sewerage projects through BNH in the states of Minas Gerais, Pernambuco, Bahia and Ceara and in Greater Sao Paulo. The first Urban Transport Project in five metropolitan areas was approved in June 1978 and is under execution (Annex 1). A Sites and Services and Low-Cost Housing Project in Greater Sao Paulo, Bahia and Pernambuco was approved in January 1979. The Medium-Sized Cities Project, approved in May 1979, contains subprojects in eight cities. A water supply and sewerage project for the southern states was just approved in March 1980. Urban projects, now under preparation, include a third water supply and sewerage project for the state of Minas Gerais, an integrated urban development project and a third urban transport project. II. AREA OF INFLUENCE - PORTO ALEGRE METROPOLITAN REGION A. Geography 2.01 The suburban railway system to be constructed under the proposed project would traverse the Metropolitan Region of Porto Alegre (MRPA) from north to south (Map IBRD 14688). MRPA covers a total area of 5,718 km2 and is located at the head of an inland waterway which runs parallel to the coast in the central part of the state of Rio Grande do Sul, the southernmost state in the country. The principal area of influence for the proposed project is MRPA's north-south corridor, which comprises ten municipalities including Porto Alegre (where the line originates) and has a population of -8- 1.9 million at present. 1/ The line would also serve, although to a more limited extent, through car and feeder bus connections, the Region's other four municipalities. 2/ The total population of MRPA is 2.38 million. B. Demographic and Economic Base 2.02 MRPA represents over 25% of the State's total population; in 1940, its share was only 13%. This quite large relative growth of the Region in the State is essentially explained by the very high rate of migration into MRPA, predominantly (about 90%) from the secondary centers and small towns in the interior of the State and, to a lesser extent, from the State of Santa Catarina to the north. Thus, while annual average rates of population growth in MRPA have been on the order of 4% since 1960 and closer to 5% since the mid-1970s, migration has accounted for about 82% of this increase, while natural population growth has accounted for only about 18%. 2.03 The period since 1950 has also experienced some changes in the dis- tribution of population growth within MRPA. In the 1950s, the municipality of Porto Alegre grew at an annual rate of 5% and, by 1960, accounted for some 62% of MRPA's total population. However, by the early 1970s, Porto Alegre's growth rate had declined to 3.3%, and, by 1974, its proportion of the total population had fallen to 54%, contrasting with the experience of the outlying municipalities, whose average rate of growth has been over 5% between 1960 and 1970, and even higher in the 1970s (Tables 2.1 and 2.2). Significantly, al- though the population of Porto Alegre itself declined in relative terms, it increased in absolute terms by some 130,000 inhabitants between 1970 and 1974. The region's north-south corridor grew by over 100,000 in the same period, which represents a greater absolute, but smaller relative, increase than that of the east-west corridor. Survey work carried out since 1974 confirms a continuation of these trends. 2.04 The dominance of MRPA within the State is also reflected in its relative share of economic activity: in 1970, MRPA accounted for almost 50% of all industrial production and 40% of total output in the State. Recent evidence indicates that the rate of economic growth within the Region is continuing at a high level. 2.05 Total employment in MRPA continues to grow rapidly, increasing from 439,000 in 1971 to an estimated 609,000 by 1976, an annual increase of 5.6%. The composition of this employment is given in Table 2.3, which shows the relative decline of the primary and secondary (industrial) sectors and the striking relative shift of employment to the tertiary or service sector, which, in absolute terms, increased from 236,000 in 1971 to 377,000 in 1976. This reflects the rapid increase in the migrant labor force, which is rela- tively unskilled. 1/ The ten municipalities in the north-south corridor include: Cachoeirinha, Compo Bom, Canoas, Estancia Velha, Esteio, Novo Hamburgo, Porto Alegre, Sao Leopoldo, Sapiranga, and Sapucaia do Sul. The line would pass directly throgh Porto Alegre, Canoas, Esteio and Sapucaia do Sul. 2/ The other four municipalities, which comprise the so-called east-west corridor, include: Alvorada, Gravatai, Guaiba and Viamao. 2.06 Despite the relative decline of the secondary sector, there has been some growth within the sector in the dynamic/higher wage paying industries such as the electronics and electrical industries. The traditional manufactur- ing industries in the region include metallurgy, machinery, chemicals and, particularly in Novo Hamburgo, leather and footwear. 2.07 The spatial distribution of total employment by sector is shown in Table 2.4. While it can be seen that Porto Alegre dominates MRPA's economic structure, its dominance is greatest in the sector experiencing the most rapid growth, namely, services. Also, employment outside Porto Alegre, particularly in the secondary sector, is concentrated in the north-south corridor. C. Distribution of Income, Housing and Public Services 2.08 With respect to the distribution of income in MRPA, interpretation of trends is difficult both because of the broad census income classes adopted and also because of the absence of comprehensive income data since 1970. How- ever, some comparison of the size of the "low income" population between 1970 and 1974 is possible on the basis of census figures for 1970 and survey figures for 1974. The income level that is commonly accepted in Brazil as the cutoff point for families in poverty is three times the average minimum wage, which at present is roughly equivalent to the Bank's relative poverty definition of one-third of national per capita disposable family income. The data reveal that, in 1970, 32.9% of all metropolitan households had earnings below two minimum wages. Although the survey data do not permit an exact comparison, they indicate that, in 1974, some 42% of all metropolitan area households had earnings of less than 2.5 minimum wages (Table 2.5). Thus, despite the Region's relatively high average income per capita in national terms, the inequality of the income distribution in MRPA is such that between one-third and one-half of urban households can be defined as the urban poor. 2.09 Reinforcing this conclusion is the available information on MRPA's housing stock and supply of essential public services. In 1970, roughly 8.3% of the Region's population was housed in what, by local standards, would be considered improvised, substandard dwellings, and this figure does not take into account access to basic infrastructure services. The highest proportions of such housing were to be found in the municipalities of Sapucaia do Sul, Sao Leopoldo, Alvorada, Novo Hamburgo and Porto Alegre. Studies by DEMHAB (the municipal housing department of Porto Alegre) indicate that, in 1972, in Porto Alegre alone, roughly 124,000 individuals, or about 12.5% of the municipal population, were living in impoverished, substandard dwellings. By 1977, it is estimated that the number of persons living in this type of housing had increased to about 150,000 in the central city alone. Recent information is not available on the housing conditions of the low income population living in the peripheral municipalities. However, given the rapid increase in the migrant population in such locations and the socio-economic characteristics of this population, it is probable that the housing conditions of the major migrant residential areas are at least as unsatisfactory as those in Porto Alegre. In this respect, it is significant to note that, outside Porto Alegre, both Alvorada and Cachoeirinha (major recipients of migration), were among the municipalities with the highest population densities in the 1970s. - 10 - 2.10 The availability of water and sewerage infrastructure within MRPA is also quite limited. In 1970, the two municipalities having the greatest proportion of improvised dwellings, Alvorada and Sapucaia do Sul, had the lowest proportion of their population served by water and sewerage systems; at that date, less than 20% of the MRPA population was served by a main sewer- age system. There appears to have been little investment in either water or sewerage since that time despite the rapid growth of population. Under the recently approved Southern States Water Supply and Sewerage Project, some of the municipalities in MRPA, including Campo Bom, Viamao and Alvorada, are expected to receive assistance. D. Metropolitan Planning and Coordination 2.11 The Metropolitan Region's development master plan was published in 1973. With a horizon year of 1992, the master plan visualized a metropolitan population by that date of 3.2 million. These projections assumed a more rapid rate of growth for the metropolitan area's economic base (measured in terms of job opportunities) than for population; total jobs per person were expected to increase from 25% in 1970 to 35% in 1992. In addition, the adopted land use strategy involved the consolidation of growth in the north- south axis between Porto Alegre and Novo Hamburgo by establishing a few large centers of development in the corridor. It was assumed that this growth would be drawn largely from the central area of Porto Alegre. The master plan also had as its objective the redirection of some population and employment growth along the east-west axis between Guaiba and Gravatai. 2.12 In August 1975, the Metropolitan Planning Agency (METROPLAN) was created for the purpose of coordinating land use, social and economic activi- ties, and public services in MRPA in order to achieve the plan's objectives. METROPLAN is funded by the State of Rio Grande do Sul, although it is not an implementing agency. Its role is, rather, one of seeking the required funding for projects which it identifies and which are then carried out by the execut- ing agencies. In the area of housing, for example, METROPLAN identifies the appropriate locations for the construction of low income housing, with the construction being carried out by the State Housing Company (COHAB), and, in the case of the municipality of Porto Alegre, its own housing agency (DEMHAB). To carry out its functions, METROPLAN employs about 55 professional staff. 2.13 The basic problems confronting METROPLAN essentially relate, in the first place, to the fact that it does not have the legal authority to require the municipalities to follow the metropolitan master plan; further, it needs additional technical expertise to assess and plan for the type of basic structural economic and demographic changes that the metropolitan area is presently experiencing (paras 2.02-2.07) and that were not anticipated by the master plan. The master plan did not forecast the large increase in the rela- tively unskilled, low income migrant labor force or the concentration of informal job opportunities in the north-south corridor and the attraction of this axis for the new migrants. A more complete assessment of the master plan is given in Annex 3. Technical assistance is included in the proposed project to help METROPLAN revise its present master plan in order to address these unforeseen large demographic and economic changes (para 3.27). - 11 - E. Prognosis for the Metropolitan Region 2.14 Having assessed the master plan in the light of the most recent evidence on the growth, structure and spatial distribution of popula_on al^e employment and found its projections to be at variance with the evidence, it is necessary to put forward a new prognosis concerning the metropolitan area's development. It is also necessary to assess the role of the suburban railway in the light of this prognosis. 2.15 Considering first the question of migration, it is likely that migration would continue to exceed that assumed in the plan, although this high rate could be reduced by the implementation of an effective state invest- ment program for the interior. At present, however, there is no evidence to suggest that such a policy would be forthcoming in the short term, and, further- more, it is far from evident whether such a policy to try and reverse present migration trends would be desirable on economic grounds. Given this situation, the continuation of present migration rates through the 1980s would not seem unreasonable. 2.16 The realization of this migration rate would produce a metropolitan area population of over 3.2 million by 1985 (the master plan had forecast that this level would be reached in 1992), 4 million by 1990, and, depending on the rate at which migration subsides in the future, from 4.5 to 5 million by 2000. This population is also likely to be of a different socio-economic composition (poorer and less skilled) and to have different locational charac- teristics than those assumed in the plan. 2.17 It is also probable that the activity rate of registered employment to population would not increase to the extent assumed by the master plan, if, indeed, at all. What is likely to increase is informal sector employment, particularly in tertiary sector activities. Thus, even allowing for the fact that the adoption of a constant activity rate represents a conservative em- ployment growth assumption, the clear indication is that population growth is likely to be higher and registered employment growth lower and of different sectoral structure than previously assumed. METROPLAN and local staff of the National Transport Planning Agency (GEIPOT) are generally in agreement with this scenario. 2.18 The suburban railway project should be considered in the light of these more probable trends for the future. The most important point is that recent events have combined to produce a greater need and stronger justifica- tion for the suburban railway project than had followed from the objectives of the master plan, essentially because the north-south corridor is developing much more rapidly than was anticipated. Moreover, the creation of new activity centers along the line, if introduced under the right conditions (para 3.27), is likely to facilitate the master plan's objective of dispersing tertiary sector employment from the Porto Alegre Central Business District (CBD). F. Transport Sector (i) Basic Network 2.19 MRPA's transport system is shown on Map IBRD 14688; it consists essentially of roads. A main north-south trunk road, BR-116, centrally located - 12 - in the corridor, connects the southern- and northernmost municipalities of the Region. This is a four-lane divided highway with limited access, and, through- out most of its 42 km except near the city of Porto Alegre, it has two service lanes on each side to serve the local traffic. Several smaller two-lane east- west roads feed into BR-116 from both sides. In addition, an east-west trunk road, BR-290, with four lanes, crosses BR-116 within the municipality of Porto Alegre at the southern end of the Region. The total route length of the pri- mary road system within MRPA is about 400 km. The route lengths and number of lanes for the main roads are shown in Table 2.6. 2.20 An old meter gauge railway line parallels BR-116 between Porto Alegre and Rio dos Sinos (39 km); it is double track from Porto Alegre to Km 12 and single track from Km 12 to Rio dos Sinos. There are also the remnants of a roadbed, where a railway line used to exist, between Rio dos Sinos and Novo Hamburgo (6 km). The existing track is over 100 years old and is beyond economical repair. The rolling stock consists of some 15 dilapidated wooden cars over 50 years old and two diesel locomotives. The federal railway (RFFSA), which owns these assets, runs two commuter trains daily in each direction and carries about 700 passengers daily. A suburban train service, therefore, is virtually non-existent. Freight trains destined for or leaving the port (three trains daily on average) use this line for some 10 km. The port itself would be considered medium-size in Brazil. It handled about 5 million tons in 1977, of which grains represented about 80%. The main port for the state of Rio Grande do Sul is Rio Grande, which is about 300 km to the south of Porto Alegre. 2.21 So far as planned transport projects in MRPA are concerned, the local DNER in Porto Alegre has a project proposal (presently shelved) to build a new highway in the north-south corridor between Porto Alegre and Scharlau (24 km). The design for this project is completed, but the project has not been approved, nor is there any indication that it is likely to be approved in the foreseeable future. Nevertheless, in view of the considerable adverse effect that such a project would have on the proposed suburban railway project and the fact that the economic analysis (Chapter VI) indicates that the proposed project is preferable to this road project if considered an alter- native, it was agreed with the Federal Government during negotiations that the Porto Alegre-Scharlau road project would not be implemented until a proper economic feasibility study, satisfactory to the Bank, is carried out and shows that this project is economically justified. For other projects, not yet known, a more general agreement was reached; EBTU, through 1988, agreed to inform the Bank of any proposed passenger transport related investments exceeding US$10 million in the influence area of the Bank project, provide the Bank with information on the expected impact of the proposed investment on the Bank project, and consult with the Bank before a commitment is made. (ii) Operations and Traffic 2.22 The dominant mode of travel within MRPA is bus. Of a total of 1.95 million daily person trips by all modes in 1979, buses account for 1.3 million, or 67%. Cars and taxis account for 27.5 and 1.6% of total person trips, respectively. All of the bus companies operating in MRPA are privately owned. Ten companies are licensed by the State of Rio Grande do - 13 - Sul to provide intermunicipal services and 32 companies (22 in Porto Alegre alone) are licensed by the municipalities to provide a municipal service (Table 2.7). The average fleet size per company providing intermunicipal services is fairly large (117 buses). The fleet sizes for the companies providing municipal services vary widely, with those in Porto Alegre aver- aging about 68 buses and those in the other cities about 22 buses (Table 2.7). Tariffs for the intermunicipal and municipal services are regulated by the State and the municipalities, respectively, both of whose decisions are reviewed by the National Interministerial Council on Prices (CIP), which was recently incorporated into the newly created Secretariat for Supply and Prices (SEAP) within the Ministry of Planning. Mid-1979 tariffs for municipal and intermunicipal bus services average around Cr$ 0.20 and Cr$ 0.30 per pass-km, respectively. With these tariffs and present costs, all of the bus companies in MRPA are operating at a profit, although often at the expense of comfort to the passenger; during the peak hours on many routes, occupancy rates are double the seating capacity (Table 2.8). 2.23 Despite the dominance of buses as a mode of transport in MRPA, cars represent, on average, about 50% of the total vehicles on the road system; on some main roads, such as BR-116, the figure is as high as 74%. During the last 10 years, car ownership in MRPA has increased from 103,000 to 251,000 vehicles, a compound annual growth rate of about 12%, almost three times the rate of growth of the population. The bus fleet, on the other hand, increased by only about 7% per annum during this period (Table 2.9). To date, little has been done to discourage the use of the private automobile. The creation of exclusive pedestrian walkways in the Porto Alegre CBD (about 2,520 meters of street or about 8% of the total street length of the CBD) has helped to a degree, but a substantial amount of day-term car parking is available in the central area at relatively low cost (Table 2.10). Private automobile sales, property and user taxes are shown in Table 2.11. 2.24 Of the total daily person trips by all modes in MRPA, roughly 65% are between work and home, 25% are study-related and 10% are made for other purposes. About 72% of all trips are made during the three main peak periods comprising six hours: 6-8 in the morning, 12-2 in the afternoon and 5-7 in the evening (Table 2.12). Efforts by the municipality of Porto Alegre to convince private businesses to stagger working hours and thus to spread this peak have met with limited success. 2.25 Because of the spatial distribution of population and jobs within MRPA (paras 2.02-2.07), travel demand is high. With a limited residential population of only 63,000 or 2.6% of the total Region's population, the CBD, because of its attraction as a regional job center for the residential popu- lations living in the suburbs, generates about 266,000 passenger trips daily, corresponding to a trip generation rate of 4.25 trips per inhabitant per day. The rest of the municipality of Porto Alegre, which has a fairly large resi- dential population of more than one million inhabitants, generates another 300,000 trips involving other municipalities. The average trip generation rate for MRPA is 0.82 (Table 2.13). Average journey distances range from 9 km for automobile passengers to 21 km for-intermunicipal bus passengers (Table 2.8). About 55% of all trips at present are intramunicipal (Table 2.13). - 14 - 2.26 Traffic on many sections of BR-116, the main trunk road in the Region serving the north-south corridor, has already reached saturation levels. Daily and peak hour vehicle traffic and growth rates on BR-116 are shown in Table 2.14. During the peak period, traffic on BR-116 ranges from some 845 vehicles (1,020 passenger car units, PCUs) per hour in one direction in the Esteio-Sao Leopoldo section to some 1,371 vehicles (1,771 PCUs) and 2,294 vehicles (2,814 PCUs) in the Esteio-Canoas and Canoas-Porto Alegre sections, respectively. These congestion levels of traffic between Porto Alegre and Esteio give rise to very high operating costs and long travel times which are reflected in average bus speeds of 15-20 km per hour during the peak period. Much of this congestion is due to the large number of automobiles on the road. As mentioned (para 2.23), car traffic, which has been growing at about 4.5% per annum on this road, now accounts for about 74X of the total vehicles on BR-116. An exclusive bus express lane will soon be introduced on a section of road within the municipality of Porto Alegre after some road resurfacing is completed under the First Urban Transport Project (Annex 1). This action should help to mitigate the problem somewhat, but, on this section, and certainly on the rest of the road which serves the metropolitan area, high levels of congestion will remain and become much worse unless measures are taken to make the available corridor capacity more effi- cient by reducing car usage and increasing public transport. To accomplish this goal, in addition to the type of operational improvement provided under the First Urban Transport Project, a major expansion of exclusively public transport infrastructure will be required. It is in this context that the suburban rail solution, designed to attract car and bus users, has been considered and several mutually exclusive alternatives have been analyzed (Economic Analysis, Chapter VI). (iii) Policy, Planning and Coordination 2.27 A transport master plan for MRPA, which, among other things, recom- mended a suburban train service for the north-south corridor, was completed by GEIPOT in 1976. 1/ Until recently, a specialized metropolitan-wide trans- port agency to promote the plan's implementation did not exist. Transport planning in MRPA was essentially the responsibility of the individual muni- cipalities, and all licensing, franchises and fares were regulated either by the municipalities or the State of Rio Grande do Sul, depending on whether the service, facility, etc. was within or outside the jurisdiction of a single municipality. During several project preparation missions, the Bank discussed this situation with local and state government officials who were acutely aware of the need for coordination at the metropolitan level. In response to this problem, in December 1979, the State Governor of Rio Grande do Sul created the Nucleo Metropolitano de Transportes Urbanos (NMTU) to coordinate and regulate the activities of all transport-related activities within MRPA. 2.28 NMTU replaces the several state agencies previously involved in the coordination and regulation of transport in MRPA. In addition, under an existing federal law, the newly created NMTU, in the interest of the Metro- politan Area and in order to improve its transport system, has the authority 1/ Plano Diretor de Transportes Urbanos da Regiao Metropolitana de Porto Alegre, 1976. - 15 - to also perform many of the regulatory functions which were previously the exclusive responsibility of the municipalities. During negotiations, the State of Rio Grande do Sul agreed that NMTU would per.form the following functions: (a) grant and review franchises for intra- and intermunicipal bus companies and taxi operators; (b) specify bus routes; (c) coordinate levels of service at intermodal connections; (d) specify, regulate and review safety standards for all modes; (e) provide transport enterprises with guidelines for the setting of tariffs, monitor such tariffs and, if necessary, require such enterprises to propose tariff changes. (While CIP's approval of tariffs is still required (para 2.22), it is hoped that, at some time in the not too distant future, this regulatory control would be relinquished and only the monitoring activities of NMTU would be retained.) (f) regulate the use of the private car through the coordination of municipal parking policies, user charges and taxes and the increased availability of public transport; (g) review and advise on all proposed transport infrastructure investment decisions in the metropolitan area; and (h) prepare and annually update an indicative rolling five-year transport plan to serve as a guide to public and private investment decisions. 2.29 So far, it is too early to judge the performance of this newly created agency. Many of NMTU's authorized powers, particularly with regard to regulation, have not yet, in practice, been transferred from the municipalities. This aspect should be carefully monitored during project supervision to ensure that NMTU achieves its intended objective. Technical assistance to help NMTU perform its designated functions is included in the proposed project (paras 3.23-3.26). G. Travel Demand Forecasts 2.30 For the purpose of forecasting traffic in MRPA, the latter was divided into a total of 46 zones. With the assistance of METROPLAN, the employment and population characteristics of each zone were forecast under four alternative land use scenarios. After analyzing these four alternatives in the light of recent trends in the growth and spatial distribution of population and employment (paras 2.02-2.07) and the revised prognosis for the development of MRPA (paras 2.14-2.18), it was concluded that two alternatives were most representative of the future evolution of the metropolitan area. Both involved the same overall rate of growth of employment and population in the Region, but with somewhat different spatial distributions. Specifically, - 16 - the growth in tertiary employment that was forecast for the CBD of Porto Alegre under one alternative was redistributed to a few commercial centers along the railway right-of-way under the second land use alternative. The alternative with the growth in tertiary employment concentrated in the CBD of Porto Alegre was adopted as the population and employment base for the travel demand forecasts to be applied to all the project alternatives examined in the economic analysis (Chapter VI) with the exception of the two alternatives involving rapid rail mass transit, namely, the proposed project and a light rail alternative. In both of these cases, the second land use alternative, which takes into account the development of commercial centers along the line, was adopted. Then, for each of these two land use alternatives and for each of the forecast years (1985, 1990 and 2000), the number of trips generated by each zone, the distribution of trips between zones and, finally, the modal distribution of trips between zones were estimated. A description of the four land use alternatives and how the above steps in the transport model were accomplished is presented in Annex 4. 2.31 The forecast of total daily person trips in the region by mode, under the assumption that the proposed suburban railway project is imple- mented, is summarized below: Forecast Total Daily Person Trips by Mode in MRPA Year/Mode Car Bus Train Total 1985 693,371 1,361,255 260,779 2,315,405 1990 746,191 1,533,560 386,180 2,665,931 2000 924,805 1,882,874 586,069 3,393,748 2.32 Under the second land use alternative, in which the rate of economic growth is the same but its spatial distribution is different, the total number of trips in the Region is forecast to be the same as that above, but the distribution of trips, trip lengths and the modal composition are different, reflecting the fact that the growth in employment expected around certain stations would, instead, without the rapid rail mass transit system, be concentrated in the CBD of Porto Alegre. The assumption that the rate of economic growth in the Region would be the same with or without the rapid rail mass transit system is generally consistent with the evidence accumulated from elsewhere, which shows that the introduction of such a system is of itself unlikely to bring about an increase in a metropolitan area's rate of economic growth. It does appear, on the evidence available, however, that a rapid rail mass transit system can influence the distribution of economic growth and job creation when the growth potential already exists, as appears to be the case in the north-south corridor of MRPA. H. Train Ridership 2.33 After obtaining the forecast of suburban train passengers as described above, the 46 x 46 matrix of train passenger flows was collapsed into a 14 x 14 station-to-station matrix. The station-to-station, origin and destination flows of train passengers in 1985, 1990 and 2000 are presented in Tables 2.15, 2.16 and 2.17. The most important stations in the system, - 17 - in terms of passengers, are: Mercado Central, Sapucaia, Farrapos, Niteroi and Canoas. Of the 261,000 train passengers forecast for 1985, about 21,400 passengers are expected to be diverted from cars and the rest from buses. The car diversion figure is expected to reach 72,500 passengers in 1990. 2.34 In 1985, the maximum peak hour flow is forecast at about 39,000 passengers or about 15% of total daily traffic, of which about 80% or 31,000 passengers are forecast to move in the principal direction. This figure of peak hour traffic in the principal direction is forecast to reach 46,000 passengers in 1990 and 70,000 passengers in 2000. The practical design capacity of the system under the project is estimated at 48,000 passengers per hour in one direction, stretchable to 72,000 passengers with some relatively small investments (para 3.06). Thus, sometime in the early 1990s, some small changes in the system may be required, but otherwise, the proposed project should be able to serve the demand quite efficiently through at least the remainder of this century. III. PROJECT OBJECTIVES, DESCRIPTION AND COST ESTIMATES A. Project Objectives 3.01 The principal objectives of the proposed project to provide a suburban rail mass transit system in the metropolitan region of Porto Alegre may be grouped into three main categories: (i) institutional, (ii) economic efficiency and (iii) urban development. (i) Institutional 3.02 As part of the preparation of the proposed project, a new company (the Porto Alegre Metropolitan Rail Mass Transit Company, TRENSURB/P.A.) was created in mid-April 1980. This company, when fully established, would be the first of its kind in Brazil, independent of the federal railway. TRENSURB/P.A. would own and operate the suburban train facilities and equip- ment provided under the project. This new institution would also break with past tradition insofar as it is proposed to be run on a commercial basis, whereby tariffs would be set at levels sufficient to cover at least all operating costs. (ii) Economic Efficiency 3.03 The proposed project is expected to achieve a significantly more efficient modal distribution of traffic and use of all transport facilities in the metropolitan region's north-south corridor by diverting as many car passengers as possible and a significant proportion of the intermunicipal bus passengers to the suburban train service while retaining buses on feeder lines and on those longer routes which are most economically served by bus. The reduction of trips made in automobiles, accompanied by the substitution of public for private modes of transport, is an important element of the project design. - 18 - (iii) Urban Development 3.04 The suburban railway system (with stations approximately every 2 km) is expected to contribute to the attainment of the metropolitan region's land use strategy of developing selected employment growth poles in the north-south corridor by providing the growth stimulus for the development of centers around certain railway stations along the line. On the basis of all support- ing evidence, it is unlikely that this type of spatial development could be achieved by providing additional road capacity which would tend to support a more linear expansion of the city of Porto Alegre. B. Physical Works and Equipment 3.05 The proposed project entails the construction of a surface suburban rail mass transit system with electric unit trains between the central market in downtown Porto Alegre and the municipality of Sapucaia, 26.7 km to the north (Map IBRD 14688). The line basically follows the same right-of-way as RFFSA's existing obsolete railway line (para 2.20), except for a few deviations to improve the new line's geometry and also to provide a 2.3 km access to the central market in Porto Alegre where no railway right-of-way exists at present. RFFSA has already transferred the right-of-way to TRENSURB/P.A. in exchange for capital shares (para 5.03). This project constitutes Phase I of a larger scheme which includes, in Phase II, the stretch between Sapucaia and Novo Hamburgo (an additional 16 km). Detailed engineering for the entire 42.7 km has been completed by GEIPOT. An economic and financial feasibility study of Phase II is included in the proposed project (para 3.22). The major items of the proposed project (Phase I) are: tracks, pedestrian and road vehicle under/over passes, stations, a train maintenance depot with related yard and equipment, substations and catenary for the provision of electric power, electric unit trains, automatic signaling, telecommunications, satellite automobile parks near some stations to attract car users, related civil works and technical assistance. 3.06 The automatic signaling system is designed for a minimum headway of three minutes between trains, which would permit a maximum of 20 trains per hour in each direction. The station platform lengths, which can be increased in the future as the situation warrants, are 176 m and can accom- modate trains with compositions of up to eight cars. Thus, with a maximum occupancy of 300 passengers per car (para 5.07), eight cars per train and 20 trains per hour, the system is estimated to have a practical design capac- ity of 48,000 passengers per hour in one direction. By lengthening the platforms to accommodate trains with 12 cars and by making a few other rela- tively small changes in the passenger areas of the stations, the catenary and substations, and by lengthening some of the sidings, the system's capacity could be increased to 72,000 passengers per hour in each direction. A more detailed description of the project is presented below. (i) Track, Bridges and Related Works 3.07 The new line will consist of two broad gauge (1.6 m) tracks. It will begin in the CBD of Porto Alegre near an old, and architecturally significant, city landmark--the central market building. From Mercado Central station, which is the most central point in the CBD, the line will proceed north following Avenida Maua (presently five lanes with traffic in one direction - 19 - which will become four lanes with the railway line) until it reaches the second station at Km 0.9 (Rodoviaria). There the line will join BR-116, which at this point is called Avenida Castelo Branco. At this juncture, a road bridge will be built over the railway line, which will continue on its northerly course on the western side of BR-116 in the existing railway right-of-way until Km 19.5. At this point, another road bridge on BR-116 will be built over the railway line, which will then continue on the eastern side of BR-116 until it reaches the terminal station of Sapucaia (Km 26.7). The line will be protected along its entire length by a concrete post fence on the urban access to the CBD and by a precast concrete and wire fence for the remainder. There are no level crossings; along the route there are 16 pedestrian over/under passes, ten road vehicle bridges and six railway bridges. 3.08 The old track will be removed and the new track put in place. The new track will have welded rails. The line will be relatively straight (minimum radius in the curves of 200 m, which occurs in only one place) and level (maximum §radient of 2%, which occurs in a few places on the line). About 172,000 m of ballast, 93,000 concrete sleepers, 65,000 wooden sleepers and 9,800 tons of rail will be used. Rail fittings, switches and crossings, and rail welding materials are also included. 3.09 In order not to cut off the access of freight trains to the port, which presently use the existing line (para 2.20), a 10-km meter gauge track, parallel to the two suburban tracks, will be built between Canoas, where RFFSA's freight line enters the Region, and Km 3.8, where the freight line deviates to follow its own separate alignment to the port. To the extent possible, salvageable materials from the old meter gauge tracks will be used on the parallel freight line. Also, RFFSA's long-distance passenger trains will originate/terminate in Diretor Pestana station (near the airport), which is in close walking distance to the suburban line. 1/ Thus, only the suburban trains will be run on the suburban system. (ii) Stations 3.10 A total of 14 stations, spaced approximately every 2 km, will be built along the line (Map IBRD 14688). Mercado Central station (Km 0) with convenient access to bus transportation and the rest of the Porto Alegre CBD, will be partly underground and connected to the railway platforms and to the street above by underground tunnels. Rodoviaria station (Km 0.9), which is also partly underground, will be conveniently located next to the city's main bus terminal. Aeroporto station (Km 6.1) will have a 330 m underground tunnel connecting it with the terminal at the airport. 2/ Satellite automo- bile parks adjacent to several stations, having a combined capacity of 13,500 automobile spaces, are also included in the project. Some of the most important stations, from the point of view of demand for satellite automobile 1/ RFFSA's long-distance passenger service involving Porto Alegre is insignificant, with, on average, about one train per day. This situation is not expected to change in the foreseeable future. 2/ About one million passengers a year arrive at and leave from Porto Alegre Airport. The visitor-to-passenger ratio is about three to one on average. At present, more than 90% of all passengers and visitors arrive at or leave from the airport in private automobiles or taxis, which clearly represents a potential market for the rail mass transit system. - 20 - parking facilities, include: Farrapos (Km 4.8), Niteroi (Km 10.5), Canoas (Km 14.8), Luiz Pasteur (Km 24.7) and Sapucaia (Km 26.7). Consultants, engaged under the technical assistance program (para 3.23), would identify the precise locations and sizes of the satellite automobile parking lots. 3.11 Four stations, including the two terminals--Mercado (Km 0) and Sapucaia (Km 26.7)--and two intermediate stations--Ceasa (Km 8) and Matias Velho (Km 16.2)--will be provided with railway switches and sidings to allow for the formation and return of trains. Passengers will enter the platforms through automatic turnstiles after inserting their tickets. These tickets will be purchased from ticket sellers in booths in the stations. (iii) Substations and Catenary 3.12 The line between Mercado Central and Sapucaia will be electrified with 3 kV (kilo volt) direct current fed into the overhead catenary by two 6 MVA (Mega Volt Ampere) substations located at Km 4.5 and Km 17.9. Electric energy (hydrogeneration) will be supplied by the State Electric Company (CEEE). The proposed project can be accommodated within CEEE's present expansion plans. Under a Bank loan approved in March 1980, CEEE will expand its sub- transmission and distribution system in the State. (iv) Electric Unit Trains and Spare Parts 3.13 The project includes 25 electric unit trains (and related spare parts) of the same type now being used in Rio de Janeiro. This number should be sufficient for the first few years of operation. The requirement is worked out in Table 3.1. Each unit train consists of four cars (two motor coaches and two trailers). The unit trains can be coupled together, the limit essen- tially determined by the lengths of sidings and platforms. Initially, two unit trains will be coupled together during the peak periods (para 5.06). The cars will be made of stainless steel and be 22 m long, 2.94 m wide and 3.91 m high. Each car will haye 108 seats. The trains will have a maximum acceleration rate of 0.8 m/sec and a maximum speed of 90 km/hr. (v) Train Depot and Administration Building 3.14 A train depot for the normal maintenance of trains will be built at Km 7.3 in an area now owned by RFFSA (Diretor Pestana Yard). The depot will have six main tracks with a capacity to provide normal periodic main- tenance for nine unit trains. The average unit train, which is expected to run about 95,000 km a year, should visit the maintenance depot approximately once a week for 15 man-hours of work and every two months, four months, and 12 months for 70, 100 and 150 man-hours of work, respectively. General overhauls should not be required for at least 0.5 million km, or during the first five years of operation. Thus, a workshop for general overhaul is in- cluded only in Phase II. A small administration building will be constructed next to the train depot, which will house the train operations center and all the administrative offices. (vi) Track Maintenance and Maintenance Depot Equipment 3.15 The proposed project also includes equipment for track maintenance and for the train maintenance depot (Tables 3.2 and 3.3). - 21 - (vii) Signaling 3.16 The signaling system will be bidirectional centralized traffic control (CTC) which, if necessary, will permit trains on the same track to run safely in opposite directions between stations. The CTC control room will be located in the operations center of the administration building. The operations center, from information recorded on its display panel, will be able to follow all line operations, train locations and switch and signal positions. (viii) Telecommunications 3.17 The telecommunications system will provide for telephone units every 500 m along the line and, through an underground cable, the capability for communication between the control room and every station, between stations, and between the management and stations/depot/operations center and points outside the system. There will also be telephone communication between the trains and the stations and operations center. (ix) Land 3.18 The project cost estimates (para 3.28) include the expropriation of some pr vate land which is required for the satellite automobile parks (297,000 m ) and for the construction of some under/over passes and substations (42,000 m ) which are partly outside RFFSA's existing right-of-way. No parti- cular difficulties are expected in acquiring this land for the project. In addition, the project cost estimates include provision for the expropriation of some private land to be reserved for the construction of additional satellite car parks in the future (for about 9,000 automobile parking spaces) and for the continuation of the line to dovo Hamburgo under Phase II when it becomes economically justified (63,000 m ). The public lands, which are included in the project (but not in the cost estimates) and which have been contributed to the project in exchange for capital shares in TRENSURB/P.A. (para 5.03), include: RFFSA's right-of-way and the area for the train depot and administra- tion building (500,000 m ) and State land near Aeroporto Station (12,000 m Porto Alegre municipal land for Mercado Station and the urban access along Avenida Maua (about 24,000 m ) has been designated for this purpose and is expected to be transferred to TRENSURB/P.A. in the next few months in exchange for equity shares in the company (para. 5.03). C. Complementary Institutions and Policies 3.19 To complement the physical work, certain institutional/organizational changes are also included in the project: (a) the development of the newly established NMTU so that it may become an effective transport coordinating and regulatory agency at the metropolitan level (paras 3.23-3.26); and (b) the setting up of TRENSURB/P.A. to own and operate the suburban rail service (Chapter V). Also, in addition to the incentive policy of providing satellite automobile parks in some of the outer stations to attract automobile passengers to the suburban train system (para 3.10), experience elsewhere has shown that it would be necessary to have a disincentive policy as well, to discourage commuters from using their cars. The automobile parking restrictions program recently adopted by the municipality of Porto Alegre provides a useful frame- work within which to work. A program, including the necessary enforcement, to reduce substantially the availability of day-term automobile parking spaces in the CBD of Porto Alegre, in conjunction with the introduction of the suburban - 22 - train service, is included in the proposed project. During negotiations, the Federal Government, through an existing agreement with the Municipality of Porto Alegre, agreed to cause the latter to implement this program which consists of (i) the elimination of all commuter-related day-term street and open-lot automobile parking in the CBD of Porto Alegre, and within the first ring road that surrounds it, within three months of the commencement of TRENSURB's operations, and (ii) the continuation of the Municipality's prohibi- tion on the construction of new automobile parking garages in this area. The State, which employs the police force in MRPA, agreed, during negotiations, to enforce the above program. Further measures to control the use of private automobiles, such as area license and other pricing schemes, taxes, restricted zones and car pooling would be examined by consultants during project implemen- tation (para 3.24). D. Technical Assistance 3.20 The scope, timing and cost of the technical assistance provided under the proposed project are shown in Table 3.4. There are basically three parts to this technical assistance program which relate to (i) the setting up and operation of TRENSURB/P.A.; (ii) other complementary transport policies and programs which would be carried out by NMTU; and (iii) the urban impact of the proposed project. The consultants, to provide the technical assistance, would be retained by TRENSURB/P.A., NMTU and METROPLAN as described below (paras 3.21-3.27). Consultants for the supervision of the physical works and procurement of equipment are discussed in paragraph 4.03. (i) TRENSURB/P.A. (a) Setting Up TRENSURB/P.A. 3.21 Consultants are needed to assist with the setting up of TRENSURB/P.A. All that exists at present is a legal structure for a company. The consultants would be retained by TRENSURB/P.A. They-would advise on staffing, the training of personnel before and after train service begins, organizational structure, management, accounting, finance and operations. In addition, the consultants would assist with the implementation of some of their recommendations during the first year of operation. (b) Phase II Economic and Financial Feasibility Study 3.22 Consultants are needed to carry out an economic and financial feasi- bility study for Phase II, which involves extension of the system from Sapucaia to Novo Hamburgo (16 km). Only if the extension of the system in the next five years is economically justified would the consultants work out the details regarding the required changes in operations, staffing, management, accounts, tariffs, etc., the changes needed in bus routes and the financing plan for this new capital investment. The consultants would be retained by TRENSURB/P.A. (ii) Other Complementary Policies and Programs (a) Satellite Automobile Parks, Auto Restraint Policy and Monitoring 3.23 While some preliminary analysis has been carried out with respect to the demand, and available space, for satellite automobile parks, and provision for a certain global capacity of 13,500 automobile spaces has been included in the project, more details are required before a definitive construction plan can be executed. Consultants would be engaged to identify the precise location and sizes of the satellite automobile parking lots, to design such facilities and to identify the land which should be expropriated now for the future provision of an additional 9,000 automobile parking spaces at satellite automobile parks. - 23 - 3.24 The consultants would also study the various options, in addiLion to parking controls (para 3.19), to restrain auto use in MRPA in general and in the north-south corridor in particular, and to recommend a complete auto restraint program which the Bank and the State Government would then discuss in order that the State, through NMTU, might take appropriate action. In addition, the consultants would, during the first two years that the railway is operating, monitor the impact of the auto restraint program on the entire Region, as well as the extent of car passenger diversion to the rail mass transit system. Should certain problems arise, the consultants would propose solutions which the Bank and the State Government through NMTU would discuss prior to a course of action being adopted. The arrangements for these Bank- Government discussions at the end of the study period and during the monitoring phase, and the State Government's commitment to take appropriate action, through NMTU, were discussed and agreed during negotiations. The consultants to carry out all of the above work would be retained by NMTU. (b) Bus Routes 3.25 With the train service, some existing bus routes become uneconom- ical; others become more important as feeder links to the railway line, and others should simply continue because they serve a population which is more economically served by bus. In the modal split and economic analyses, these factors have been taken into account in a general way, but it was not possi- ble to identify the specific bus routes in each case. Before the railway service is operational, however, this detailed information would be needed by NMTU, which would be responsible for implementing the new bus program. Therefore, consultants would be engaged to identify those specific bus routes which should be discontinued and those which should serve as feeder links to the railway line. The consultants would also make recommendations on actions regarding those bus companies which would be adversely affected by these changes. The consultants would be retained by NMTU. (c) Tariff Policy 3.26 Consultants are needed to study the levels and structure of all transport tariffs in the Region, including the charges for satellite and other automobile parking, and to make specific recommendations on how they should be set after considering all economic efficiency, income distribu- tional and fiscal factors. The consultants would be retained by NMTU. (iii) Urban Impact (a) Commercial Center Development and Master Plan Assistance 3.27 With the introduction of the suburban railway service, the areas around some stations are expected to become more attractive for certain kinds of commercial development. This type of development has been assumed to occur under the land use alternative adopted for the analysis of the pro- posed project. However, this analysis was undertaken with the limited objec- tive of deriving the distribution of employment and population by zone in order to generate person trips in the transport model. Specific questions concerning the potential for different types of businesses and the conditions necessary for its realization and implementation were not addressed. There- fore, consultants are needed to assist METROPLAN, which has agreed to participate in such an exercise, in identifying the potential for specific types of commercial development at certain locations along the line where such development would be beneficial from a social as well as private point - 24 - oL view, and to recommend how best to create the planning and organizational conditions necessary for the achievement of such development. This would include consideration of the issue of whether or not a public real estate company or some other institution would be needed to manage the development. The real estate within the stations would be managed by the new railway com- pany. The consultants would be retained by METROPLAN. The same consultants would also assist METROPLAN with the updating and revision of its land use master plan (para 2.13). E. Cost Estimates 3.28 The cost estimates for the proposed project are based on the GEIPOT engineering design study of 1976, adjusted for changes suggested in discussions with the Bank. Base period unit prices are estimated as of January 1980. The cost of the technical assistance has been estimated on the basis of some 351 man-months of consulting services at an average cost per man-month of US$10,000 and US$6,000 for foreign and local consultants, respectively (Table 3.4). The total cost estimates for the entire project, broken down by foreign and local costs for each year, are shown in Table 3.5 and summarized in paragraph 4.06. The total project cost is estimated at Cr$ 10.0 billion in January 1980 prices (US$232.8 million equivalent) with a foreign exchange component of US$117.2 million or 50%. To these base period estimates, price and physical contingencies have been added. Physical contingencies include 5% for cement, 3% for steel, 5% for cables, 5% for substation spare parts and 10% for fuel. Price contingencies were based on the following cost increases for both civil works and equipment: 10.5% in 1980, 9% in 1981, 8% in 1982, and 7% in both 1983 and 1984. No differentiation was made between local and foreign price contingencies on the assumption, which is considered reasonable, that the cruzeiro would be devalued frequently in order to maintain a relatively constant relationship between local and foreign prices. The physical and price contingencies taken together represent about a 34% increase over the base cost estimates. IV. PROJECT IMPLEMENTATION AND THE PROPOSED LOAN AND BORROWER A. Project-Implementing Office 4.01 To implement the proposed project, which is estimated to take three- and-a-half years (para 4.04), a project-implementing office or Nucleo Executivo de Projecto (NEP) would be established within TRENSURB/P.A. The structure and key staffing of NEP were discussed during the preparation of the project. The establishment of NEP within TRENSURB, with staff satisfactory to the Bank, would be a condition for loan effectiveness. Since the first order of priority would be the commencement of the physical works, initially TRENSURB/P.A. would consist almost entirely of NEP with a few other personnel to handle the administrative responsibilities of the new company until its full structure and staffing (to be recommended by consultants, para 3.21) are in place. All aspects of TRENSURB/P.A., in addition to its relation to NEP,. are discussed in Chapter V. 4.02 Some steps have already been taken toward the formation of NEP. A highly qualified GEIPOT engineer, who has been supervising the project design since 1976, has been chosen as the future head of NEP. In addition, some other engineers from GEIPOT, RFFSA and EBTU are already working with him. - 25 - Also RFFSA, EBTU, GEIPOT and METROPLAN have been designated by the Minister of Transport to provide NEP with the necessary technical services. It is envisaged that RFFSA would play a key role in carrying out railway physical works. During negotiations, agreement was obtained from RFFSA, EBTU, GEIPOT and METROPLAN that they would provide the necessary technical services to NEP. 4.03 During project implementation, the principal functions of NEP within TRENSURB/P.A. would be (a) to arrange for the expropriation of all land under the project, including the land for the satellite automobile parks; (b) to prepare contracts and to call for and evaluate bids; and (c) to award contracts to carry out the physical work as well as the technical assistance according to the agreed schedule. Supervision of construction and equipment procurement would be provided by consultants. Provision for the cost of operating NEP and the consultants is included in the project costs (Table 3.5). At the end of project implementation, NEP within TRENSURB/P.A. would be disbanded, but it is possible that NEP and much of its staff could be incorporated into the technical department of TRENSURB/P.A. These matters would be studied by consultants under the technical assistance program (para 3.21). B. Implementation Schedule 4.04 The project implementation schedule is shown in Table 4.1. The preparation and award of contracts are expected to be accomplished during the latter part of 1980 and the early part of 1981. Actual physical work is expected to begin on a large scale by mid-1981 and to be completed by the early part of 1984. Trains are expected to be running by mid-1984. The first full year of operations would be 1985. The work on various items would begin at essentially the same time at several different points along the line. Thus, work on track infrastructure, stations, bridges and buildings is expected to begin by mid-1981, followed by the construction of the track. This work is expected to be divided into many different contracts. In order not to inter- rupt freight traffic for the port during construction, one of the existing tracks would continue in use until the new freight line is in place on the side of the suburban line. The supply of equipment and material for electri- fication and signaling would begin in early 1982 and the corresponding physical work in mid-1982. The supply of equipment and material for the telecommunica- tions system and the first deliveries of unit trains would begin in mid-1982. During the construction of the elevation of Avenida Castelo Branco and the urban access to Mercado Station on Avenida Maua, certain temporary traffic arrangements would need to be made, for which the assistance of the Muni- cipality of Porto Alegre would be required. C. The Proposed Loan 4.05 A loan of US$159 million is proposed, which would cover the full foreign exchange costs (including contingencies) of the project. The equip- ment and services, which would be financed by the proposed loan, are shown in Table 4.2 and include unit trains and spare parts, track materials, signaling equipment, catenary and substation equipment, maintenance depot equipment, track maintenance equipment and technical assistance. - 26 - 4.06 A comparison of the total project costs and the proposed Bank loan are shown below: US$ thousand equivalent (including all contingencies) Project Costs Proposed Bank Local Foreign Total -Loan-- Track infrastructure 13,159 1,841 15,000 - Track superstructure 20,416 14,003 34,419 12,615 Pedestrian and road vehicle under/over passes 30,968 4,617 35,585 - Stations 15,589 7,448 23,037 Administration and maintenance depot buildings 9,936 795 10,731 - Maintenance depot equipment 208 2,542 2,750 2,542 Permanent way maintenance equipment - 474 474 474 Telecommunications 2,094 3,341 5,435 - Signaling 3,383 9,228 12,611 8,410 Electrification 4,630 12,480 17,110 10,683 Satellite car parking 3,560 846 4,406 - Expropriations 23,266 - 23,266 - Unit trains 22,891 91,560 114,451 114,451 Unit train spare parts 1,915 7,665 9,580 7,665 Technical assistance 1,752 2,160 3,912 2,160 Total 153,768 159,000 312,768 159,000 Note: The division between local and foreign costs is based on an analysis of whether Brazil is a net importer of a particular item and on the likely source of supply. Specifically, in the case of unit trains, it was assumed that local bidders would be successful. D. The Borrower, Onlending and Project Agreements 4.07 The Borrower of the proposed loan would be the Federative Republic of Brazil. EBTU, in its capacity as the financial agent for the Government, would then onlend to TRENSURB/P.A. the total equivalent of the Bank loan in cruzeiros. TRENSURB/P.A. would repay EBTU such principal amount over a period of 12 years, with payments beginning after the third full year of train operations. The value of the cruzeiros borrowed would be subject to a monetary correction factor in accordance with changes in the national treasury bond index. The interest rate would be 5% per annum, which is the prevailing rate of interest charged by the National Development Bank (BNDE) for similar projects in Brazil. This onlending arrangement is considered practicable in view of TRENSURB's prospective financial capability (para 5.15). These matters were agreed during negotiations. The signing of a Subsidiary Loan Agreement, which includes the above conditions, between EBTU on behalf of the borrower and TRENSURB/P.A., would be a condition for effectiveness of the proposed Loan Agreement. The Bank would also enter into a joint Project Agreement covering all aspects of the project with EBTU, GEIPOT, RFFSA, the State of Rio Grande do Sul and METROPLAN. 27 - E. Procurement 4.08 All items to be procured under the proposed loan would be subject to international competitive bidding (ICB) in accordance with Bank guidelines. In bid evaluation, Brazilian manufacturers of equipment would be allowed a preferential margin of 15% of the CIF cost of competing imports, or the relevant prevailing level of customs duties, whichever is lower. F. Disbursements 4.09 Disbursements would be made as follows: (a) 100% of the foreign expenditures for imported equipment, materials and unit trains; and/or (b) 100% of the ex-factory cost of locally manufactured equipment, materials and unit trains; and (c) 60% of the cost of the technical assistance. An estimated schedule of disbursements is given in Table 4.3. This schedule is based on the assumption that the proposed loan would become effective by November 1, 1980. G. Financing Plan 4.10 The sources of finance for the capital cost of the proposed project include: (i) the Bank loan of US$159 million which would be onlent to TRENSURB/P.A. (para 4.07); (ii) the cash contributions of shareholders of US$153.8 million in exchange for equity shares in TRENSURB/P.A. (the share- holding arrangements are discussed in paras 5.01 through 5.04); and (iii) the land contributions of the shareholders (estimated at US$27.8 million), also in exchange for equity shares in TRENSURB/P.A. The financing plan is as follows: Financing Plan (US$ million) FY 1980 1981 1982 1983 1984 1985 Total Loan - 0.10 26.60 63.40 68.50 0.40 159.00 Equity: EBTU 1.44 4.32 24.60 25.30 17.90 0.10 73.66 RFFSA 23.00 a/ - 19.30 7.60 17.90 0.10 67.90 State of R.G. do S. 0.50 a/ 0.62 4.30 6.60 5.40 0.10 17.52 Municipality of P.A. 4.30 a! - - - - - 4.30 Other(s) b/ - 1.42 4.80 7.30 4.60 0.10 48.22 Total 29.24 6.46 79.60 110.20 114.30 0.80 340.60 a/ Land b/ See para 5.04 for an explanation of the term "other(s)". - 28 - H. Project Risk 4.11 With regard to the execution of the physical works included in the proposed project, there are no unusual risks. NEP within TRENSURB/P.A., with the technical support of RFFSA, GEIPOT, EBTU and METROPLAN, should be able to complete the project according to the agreed implementation schedule, which is considered realistic. Significant delays in this area, therefore, are not expected. With respect to the institutional and policy aspects of the project, which refer essentially to the operation and financial performance of TRENSURB/P.A. in the context of the urban structure and transport system of MRPA, more caution is warranted. (These matters are discussed in more detail in the next chapter, which deals entirely with TRENSURB/P.A.) Suffice it to say that the proposed project involves several innovations in institutions and policies in Brazil (e.g., the first metropolitan rapid rail mass transit company, the adoption of a pricing principle that allows tariffs to cover at least all operating expenses, a significant program of auto restraint measures and the diversion of a large number of car users to the railway), all of which will have to be monitored very closely. Therefore, there is a risk inherent in new undertakings of this kind. However, the Government shares the Bank's desire to have this project succeed in the institutional and policy areas and thereby serve as a model for the rest of the country. Sufficient agreements with all the agencies involved, with clearly defined commitments, have been reached during negotiations. These commitments, coupled with the technical assistance program, have reduced the risks to an acceptable level to justify the Bank's participation. During project supervision, particular attention should be given to monitoring the development of the NMTU (para 2.29) and to ensuring that the shareholding structure of TRENSURB/P.A. changes in accordance with the agreed schedule (para 5.03). V. THE NEW METROPOLITAN RAIL MASS TRANSIT COMPANY A. Ownership, Organization and Financial Policy 5.01 During the preparation of the proposed project, several alternative approaches to establishing the company to own and operate the suburban railway service were examined. After considering the various issues, it was agreed that the most practicable approach, at the present time, would be to initially establish the new company as a subsidiary of RFFSA (a step which does not require federal legislation but only the approval of RFFSA's Administrative Council), and then to increase, as quickly as possible, the shareholdings of some other shareholders such as EBTU, the State of Rio Grande do Sul, the municipalities of MRPA and possibly some others (e.g. BNDE), until RFFSA's shares are equaled or exceeded by those of at least one other shareholder. As long as no federal agency other than RFFSA would become the largest single shareholder, federal legislation would not be required. Thus, perhaps more gradually but with less political uncertainty, the desired result could be achieved. The intention would be to fully accomplish the shift from the status of a subsidiary of RFFSA to an independent company with a more diver- sified share structure well before train operations begin. - 29 - 5.02 Because all the other approaches involved federal legislative action which could be lengthy with quite uncertain outcomes and because RFFSA would play the most important role during the first few years of physical implementation in any case, the subsidiary approach, as a temporary arrangement, was considered acceptable provided agreement could be reached on the time frame for achieving the desired result of an independent metro- politan rail mass transit company. This time frame was agreed. 5.03 Thus, in mid-April 1980, RFFSA created the subsidiary company of TRENSURB/P.A. RFFSA's initial participation is based on its contribution of land. Other shareholders at this time include EBTU, through its contribution of cash and the State of Rio Grande do Sul through its contributions of land. The municipality of Porto Alegre is expected to join in the next few months in exchange for its contribution of land. To confirm their commitment to estab- lishing TRENSURB as a company, independent of RFFSA, the shareholders and the Borrower agreed during negotiations to a schedule indicating the transition of TRENSURB/P.A. from a subsidiary at the outset in which RFFSA holds 78% of the shares to a non-subsidiary legal status in which, by June 30, 1983, RFFSA would hold the same proportion of the total shares as EBTU, namely, 39%. The agreed schedule is as follows: Cumulative Percentage Shareholding in TRENSURB/P.A. % by % by % by % by 6/30/80 6/30/81 6/30/82 6/30/83 and thereafter RFFSA 78 65 51 39 EBTU 5 16 31 39 The State 2 3 6 9 Municipality of PA 15 12 5 3 Other(s) - 4 7 10 5.04 In the above schedule, the shareholder category "others" could include some other municipalities besides Porto Alegre (although their present difficult financial position makes it impossible for them to commit at this time) as well as BNDE. The prospect for BNDE's participation looks promising, but, again, no commitment can be made at this time. In the event that the balance of the total shares is not taken up, EBTU and RFFSA would take up the balance in equal proportion. Hopefully, this contingency plan would not need to be implemented, and both the other municipalities and BNDE would be able to participate in the near future. This aspect, as well as the whole matter of transition from a subsidiary status to corporate independence, would need to be carefully monitored during project supervision. 5.05 It is expected that TRENSURB/P.A. would be fully operational by the time the physical works are completed in mid-1984. Most of the details re- garding TRENSURB's organizational structure, staffing, operation, etc, would be - 30 - worked out under technical assistance provided in the project (para 3.21). It is expected that because the company would be inexperienced at the outset, although considerable training during project implementation is provided in the technical assistance program, TRENSURB/P.A. would have to rely to some extent on RFFSA or on some pool of railway expertise for the provision of certain technical services. However, TRENSURB/P.A. like any other commercial enterprise, would obtain these services by entering into normal business contracts. TRENSURB/P.A. would be run as a commercial enterprise and be governed by normal financial practices. The financial analysis (para 5.16) indicates that this is a realistic principle provided that the recommended operating plan and tariff policy (paras 5.08 and 5.15) are adopted, both of which are considered reasonable, equitable and economically efficient. B. Operations 5.06 During the initial years of operation, TRENSURB/P.A. is expected to run 11 trains per hour during the peak hours (6-8 in the morning, 12-2 in the afternocn and 5-7 in the evening) and four trains per hour during the off-peak period. Thus, the headway between trains would be about 5-1/2 minutes during the peak hours and 15 minutes during the off-peak period. Train formations are planned to consist of eight cars or two unit trains coupled together during the peak hours and four cars or one unit train during the off-peak period. The fleet requirement has been derived on the basis that, at any given time, 15% of the fleet would be in the maintenance depot and therefore unavailable. The average commercial speed of a train, which includes station stops, is expected to be about 42 km/hr. Thus, with the new train service, it should take about 38 minutes to go from one end of the line (Central Market, Porto Alegre) to the other (Sapucaia, Km 26.7). 5.07 The fleet requirement has been derived on the assumption of a maxi- mum car capaciti of 300 passengers. This implies 192 passengers standing in an area of 32 m or five to six passengers/m . This parameter was a matter of considerable interest since comfort will be an important determinant of the extent to which the train can attract car passengers. It was generally agreed that 300 passengers per car represents a fairly good, but not excessive, level of service during the peak hours and that occupancy rates above this level should be avoided by increasing the size or frequency of trains. 5.08 To run the entire operation during the first few years of train serv- ice, it is estimated that almost 650 employees would be required (Table 3.1). These are indicative estimates which, in addition to the detailed structure and total operation of the new company, need to be studied by consultants during project implementation (para 3.21). However, because these personnel estimates, along with the projections of train operations (para 5.06), form the basis for the costing analysis (para 5.11), on which the tariff policy (para 5.15) is based, agreement was reached during negotiations on the staff- ing and operating plans for TRENSURB/P.A., subject to the findings and recom- mendations of the consultants. The agreed plans are presented below: - 31 - Staffing Plan Levels of Daily Ceiling Passenger Traffic No. of Staff 100,000 400 150,000 500 200,000 550 250,000 650 300,000 700 350,000 800 400,000 850 450,000 950 500,000 1,000 Initial Operating Plan a/ Operating Day 18 hours Peak : 6 hours Off-Peak : 12 hours Headway between Trains Peak Hours : 5.5 minutes Off-Peak Hours 15 minutes Train Composition Peak Hours : 8 cars (2 unit trains) Off-Peak Hours : 4 cars (1 unit train) Average Commercial Train Speed 42 km/hr Average Fleet Availability 85% Maximum Staff 650 employees a/ Considered applicable for the first few years of operation. C. Accounts 5.09 TRENSURB/P.A. would maintain an accounting system based on sound commercial practices. The accounts would be designed by consultants under the technical assistance program (para 3.21). During negotiations, TRENSURB/P.A. agreed that its financial statements, in addition to the project accounts, would be audited annually by independent external auditors satisfactory to the Bank and that these statements would be produced annually and furnished to the Bank within four months of the end of TRENSURB's fiscal year. - 32 - D. Assets and Liabilities 5.10 The forecast of TRENSURB's assets and liabilities in constant mid-1979 prices for the period 1985-1990 is shown in Table 5.1. TRENSURB's assets during this period are forecast to consist exclusively of the items in the proposed project, including the right-of-way and some other small parcels of land (para 3.18). In addition, it is estimated that TRENSURB would accumu- late reserves, built up from its forecast net surplus, which, it is assumed, TRENSURB would hold in the form of interest-bearing bank deposits. TRENSURB's liabilities during this period are forecast to consist entirely of one long- term debt based on the loan from EBTU (para 4.07). TRENSURB's equity would consist of (a) share capital, which represents the equity contributions from TRENSURB's shareholders (para 5.03); and (b) reserves, which represent the accumulation of TRENSURB's yearly net surplus. It would not be financially prudent for the new company to enter into additional new borrowings during the firs- three years of operation. During negotiations, TRENSURB/P.A. agreed that that it would not enter into borrowings, other than the one mentioned above, before the fourth year of full operations, and then only if such borrowings would not result in a total annual debt service, including the loan from EBTU, which is greater than the average annual net operating income plus depreciation for either the immediately preceding two years of operation or for a later 24-month period ended prior to the incurrence of such debt service, whichever is greater. E Operating Costs ,.1i A forecast of TRENSURB's operating costs, in constant mid-1979 prices, at different levels of traffic, is shown in Table 3.1. During the initial period of operation when train ridership is expected to reach about 260,000 passengers daily, total working costs are estimated to be about Cr$ 2.66 per passenger. After allowing for depreciation on all fixed assets, total operating costs are estimated at Cr$ 4.82 per passenger. Depreciation, personnel, material and energy costs represent about 45%, 38%, 9% and 8% of the total operating costs, respectively. At the level of traffic forecast for 1990 (around 386,000 passengers daily), unit operating costs fall to about Cr$ 4.06 per passenger, indicating some modest economies of scale. 5.12 These estimates are based on various operating assumptions which were analyzed in considerable detail in joint discussions between the Bank and the technical experts from RFFSA and GEIPOT. In particular, staffing, which represents about 38% of the total operating costs, is a critical factor. After examining each category of work, it was mutually agreed that some 634 employees constitute a reasonably efficient but practical number of staff -or the initial operation. However, if this number were to increase by 30, 60 and 100%, the cost per passenger would increase from Cr$ 4.82 to Cr$ 5.50, Cr$ 6.17 and Cr$ 7.07, respectively. A cost per passenger of Cr$ 7.07 could jeopardize the financial feasibility of the new enterprise (para 5.14). For this reason, it is imperative that the operating plan for TRENSURB (para 5.08) be carried out generally in accordance with what has been assumed. F. Tariffs 5.13 The question arises of whether TRENSURB would be able to charge a competi-tive tariff and still cover its operating costs. Determining a compe- titive tariff for the train service is somewhat problematic since TRENSURB, - 33 - with its higher speed and greater comfort during the peak periods, would be offering a better quality of service than the bus companies and, therefore, would be competing on more than price grounds alone. Furthermore, the train would be competing with the private automobile as well, and, for these pas- sengers, the demand for the train is fairly price inelastic; convenience in terms of station connections (e.g., satellite station parking and feeder buses), waiting time, service frequency and comfort tend to be more important. The proposed project is designed to provide present automobile users with an acceptable substitute. Nevertheless, if the train is to replace buses on many routes on which the former provide a more economical solution, the train tariff should be somewhere in line with the bus tariff even if the train provides a higher quality of service, in order that the rail tariff does not impose an additional financial burden on those who would otherwise be bus users (i.e., the lower income part of the population), which would be poli- tically unacceptable. The task of identifying and implementing a new system of bus routes would be carried out by NMTU with technical assistance included in the project (para 3.25). 5.14 An analysis of bus tariffs in the influence area of the train indi- cates that, indeed, TRENSURB ought to be able to charge a tariff that covers its total operating costs (estimated at Cr$ 4.82 per passenger in constant mid-1979 prices, para 5.11) and yet is below the average bus tariff, which, in mid-1979, in the area to be served by the train, was Cr$ 7.0 per passenger. For the longest route (Porto Alegre-Esteio-Sapucaia), the fare was about Cr$ 9.0 per passenger. The minimum fare for some of the shorter routes was Cr$ 5.0 per passenger. Thus, not only could TRENSURB cover its operating costs, but it seems likely that it could also earn a return on invested capital as well and still not exceed the average bus tariff. This result can be expected to be reinforced in the future as petroleum prices rise and, other things being equal, bus costs increase more rapidly than rail costs, given the simple fact that the electric (hydrogeneration) rail service does not use this input. The question of the levels and structure of tariffs for all modes of transport in MRPA would be studied by consultants under the technical assistance program (para 3.26). Among other things, these consultants would examine the practicability of introducing distance-related and peak/off-peak rail pricing. 5.15 The administration of tariff policy should be the responsibility of TRENSURB's management. During negotiations, the Bank obtained agreement from TRENSURB that it would promptly, as necessary, seek the authorization of CIP to charge tariffs which shall provide it with revenues sufficient to cover at least all operating expenses including overhead costs, adequate maintenance expenses, taxes, depreciation and interest. During negotiations, agreement was also reached that, on the basis of a review to be undertaken not less than once every six months or whenever tariffs of passenger services competitive with those of TRENSURB are adjusted, if adjusted more frequently, the Borrower would authorize such adjustments in tariffs to permit TRENSURB to cover at least all of the above operating expenses. If economic circumstances, in the judgment of the Government, do not permit TRENSURB to charge tariffs which would cover operating expenses, the Government and TRENSURB have agreed to promptly inform the Bank of such circumstances and the tariff levels proposed. 34 - G. iinancial Results 5.16 In the indicative analysis that has been carried out, the rail tariff has been assumed to be Cr$ 6.0 per passenger, a level consistent with TRENSURB's agreed tariff policy (para 5.15), and generally constrained by average bus tariffs in the influence area. The forecasts of TRENSURB's income accounts and balance sheets in constant mid-1979 prices for the period 1985-1990 are shown in Table 5.1. TRENSURB's net surplus in 1985 expressed in mid-1979 prices is estimated at Cr$ 99.0 million or US$3.8 million. This figure is expected to increase gradually, reaching US$7.2 million in 1987. In 1988, TRENSURB's net surplus is estimated to fall to US$0.3 million because of debt service payments on the loan from EBTU (para '4.07). Thereafter, the net surplus is forecast to increase in the succeeding years. Thus, based on a tariff that is approximately equal to the average bus tariff, TRENSURB is expected to have sufficient resources to cover its operating costs, repay its debt and still have a small surplus during the period analyzed. The working and operating ratios in 1985 are estimated to be 44 and 80 respectively, improving further to 38 and 68 by 1990. TRENSURB's estimated operating profit per passenger in mid-1979 prices is forecast to increase from Cr$ 1.18 in 1985 to Cr$ 1.94 in 1990. This profit corresponds to a return on revalued net fixed assets of 1.5% and 4.3% in 1985 and 1990 respectively. By 1995, this return is forecast to reach 8%. 5.17 Of course, if the costs of operation are higher, as indicated in the sensitivity tests (para 5.11), the financial results deteriorate significantly. With a cost per passenger of Cr$ 7.07 corresponding to a 100% increase in staff, TRENSURB's working and operating ratios in 1985 increase from 44 and 80, respec- tively, to 75 and 111. If, in addition, the tariff is set at Cr$ 3 instead of Cr$ 6, the working and operating ratios increase further to 149 and 221. (More details of this sensitivity analysis are presented in Table 5.2.) There is no doubt that the attainment of the goals set forth in the operating plan (para 5.08) and the adoption of an economic tariff policy are essential to the financial success of the new company. 5.18 In addition, TRENSURB's estimated operating revenues refer only to the proceeds of passenger fares. TRENSURB may also generate additional revenues from other sources such as satellite car parking fees and from the leasing of commercial space inside the stations. Projections of these revenues have not been made. The matter of parking fees would be part of the overall tariff study included in the technical assistance (para 3.26). VI. ECONOMIC ANALYSIS A. Alternatives 6.01 For the economic analysis, five mutually exclusive project alterna- tives were analyzed: six, if one considers the base case, i.e., without project situation. The five project alternatives include: (i) the removal of the existing tracks and the construction of an express busway in the present railway right-of-way; (ii) the improvement of the existing antiquated diesel railway service; (iii) a rail mass transit system (i.e., the proposed project); (iv) a light rail system; and (v) a new road with a bus expressway. - 35 - 6.02 For each of the five project alternatives, a net present value (NPV) was calculated. 1/ The alternative with the highest NPV represents the preferred solution. The rate of return for this solution was then calculated. This analysis established the superiority of the proposed project or alter- native (iii). A summary of the analysis follows. 6.03 All prices used in the analysis, both on the benefit and cost sides, exclude taxes and are held constant at mid-1979 levels. The economic costs for each proposed project alternative consist of the capital investment costs plus the incremental road and/or rail maintenance costs. Physical contingen- cies are included in the capital cost estimates. Summaries of the capital investment costs distributed over the implementation period of each project alternative and the right-of-way maintenance costs are presented in Tables 6.1 through 6.5. 6.04 The benefits associated with each project alternative were calcula- ted as follows: first, each project alternative implies a different transport system, the characteristics of which were specified (for all practical purposes, the transport system in each case is limited to the north-south corridor, which fairly well defines the total area of influence for all the alternatives); second, the travel demand in the corridor associated with each project alter- native was estimated (as explained in paragraph 2.30, one land use scenario was adopted as the base to forecast trips for the without project situation and project alternatives (i), (ii), and (v), and another land use scenario was adopted as the base to forecast trips for alternatives (iii) and (iv)) 2/; third, all the transport costs associated with meeting the forecast travel demand in the corridor under each alternative transport system (i.e., project alternative) were calculated; fourth, the transport costs associated with meeting the fore- cast travel demand in the corridor without any of the project alternatives (i.e., the base case transport system) were calculated; and last, for each project alternative, the costs under step three were subtracted from the costs under step four to obtain the cost savings or benefits associated with each project alternative. The types of transport costs which were compared for the benefit calculation include vehicle operating and passenger travel time (including transfer or walk/wait time) costs. Summaries of these costs over the life of each project alternative and in the base case are given in Tables 6.2 through 6.6. The unit cost basis for the road and rail operating cost calcula- tions are derived and presented in Tables 6.7 and 6.8. The derivation of the value of passenger travel time is given in Table 6.9. The highlights of the base case, project alternatives, and results are presented below. 1/ When comparing mutually exclusive alternatives, the rate of return criterion may give ambiguous results. Thus, net present value was used. 2/ Since the two alternative land use scenarios do not involve different overall rates of economic growth for the region, and, thus, the total number of trips is the same for both alternatives, it was not necessary to deal with the question of generated traffic. The differences in the distribution of trips, trip lengths and modal composition arising from the different spatial distributions of growth under each land use alter- native are accounted for in the calculation of transport costs. - 36 - 6.05 The base case or without project situation, which is compared with each alternative, essentially refers to the existing transport system with committed operational and physical improvements. The most important committed improvements for the north-south corridor include pavement repair, the instal- lation of lights and signals and the erection of barriers for the creation of an express bus lane on Avenida Farropas which connects BR-116 with downtown Porto Alegre. These improvements are being made under the Bank's First Urban Transport Project (Annex 1). (i) Bus Expressway in the Railway Right-of-Way 6.06 This alternative involves the removal of the existing tracks and the construction of an exclusive bus expressway (25 km) in the existing railway right-of-way plus the relocation of the existing freight line to the side of the new expressway. An elevated road access to the CBD of Porto Alegre is also included. In considering the present solution, two alternative designs were examined: three and four lanes. Since some sections of the rail- way right-of-way are quite narrow, some expropriation of land and demolition of buildings would be required for both designs but more for the four-lane solution. These costs have been taken into account. The three-lane alter- native was found to have a higher NPV than the four-lane alternative. Thus, the three-lane design was adopted for comparison with the other alternatives. For this alternative, one can begin to see an appreciable reduction in the level of operating and passenger travel time costs when compared with the base case. However, because of the large number of buses circulating in the densely developed CBD of Porto Alegre during the peak periods, congestion costs, as reflected in vehicle operation (as well as environmental costs, such as air and noise pollution, which have not been quantified), in this central area begin to rise quite steeply. (ii) An Improved Diesel Train Service 6.07 Under this alternative, the existing track system, which is over 100 years old, would have to be totally replaced and all new installations provided. Because of the existing state of the railway, this alternative is almost equivalent to starting from nothing. The present alternative also includes an urban access line to Mercado Station on Avenida Maua as in alternative (iii) following. The use of diesel locomotives in a densely developed urban area would involve some additional environmental costs in terms of noise pollution that should be kept in mind but which have not been formally taken into account in this analysis. From the outset, under this alternative, the diesel rail system would have to operate at full capacity during the peak hours (estimated at 12 trains per hour in one direction). Despite this consideration, the road system in the corridor (basically BR-116) would still have to absorb the major share of passenger trips because of the large total demand. Furthermore, the diesel system is not expected to be competitive with the private automobile. Its ridership would be drawn essen- tially from bus users. Under these conditions, road transport costs in the corridor are expected to increase quite rapidly, especially after 1990. (iii) Rail Mass Transit System (Proposed Project) 6.08 The proposed project is described in detail in Chapter IV. The economic capital investment costs are shown in Table 6.3. Rail passenger- km costs under the present alternative are estimated to be about 25% higher - 37 - than under the diesel alternative. However, road transport costs in the corridor under this alternative are substantially lower than under the diesel rail alternative because of the rail mass transit system's higher capacity and ability to divert car passengers from the road system. This alternative also has considerably less congestion cost in the CBD than any of the other alternatives, essentially because of the high capacity, non-stop, and fixed right-of-way characteristics of the urban rail connection. (iv) Light Rail 6.09 Light rail systems are sometimes considered prior to introducing a full rapid rail mass transit system. They sometimes involve a fixed right- of-way, but not always. Monorails, most often seen in airports, are illustra- tive of the former type, whereas the trolley-bus controlled by overhead cables is an example of the latter. In the Porto Alegre Metropolitan Region, it is quite clear that the long distances involved and the large population to be served make the monorail solution, with its very limited capacity, impractical. Also, the population density in a fairly narrow band along the existing railway right-of-way is quite high, suggesting that there is probably no advantage to deviating from this route. Thus, the system chosen to be tested has essentially the same overall design as alternative (iii) but a lower capacity (approximately one-third, or 17,000 passengers per hour in one direction), resulting mostly from the use of smaller and lighter unit trains and from the substitution of automatic level crossing guards for many of the vehicle under/over passes. The type of unit train considered is the one now in use on a branch line in the Rio de Janeiro suburbs. Under the present alternative, car passengers are expected to divert to the train in approxi- mately the same proportion to the total train passengers as in alternative (iii). Rail passenger-km costs are only slightly higher than for the full rail mass transit system, alternative (iii). However, given the limited capacity of the light rail system, road vehicle operating costs are forecast to rise quite substantially. (v) lNew Road with Exclusive Bus Expressway 6.10 Because of soil, flood and present land use conditions, it is impossible to construct a new road close to the centers which would be served by all the other project alternatives without considering a totally elevated highway, which would be prohibitively expensive, esthetically unattractive and generally unacceptable to local authorities. The alignment chosen, therefore, was the one once proposed for a new road in this corridor, i.e., the Porto Alegre-Scharlau highway (para 2.21). This highway would run from Porto Alegre (starting on Avenida Assis Brasil at the intersection with BR-290) to Scharlau following a route, which, on average, is about 6 km to the east of BR-116 and which crosses BR-116 just north of Sapacaia. To compensate for the lack of access to the same urban centers served by the other alternatives, four lateral road connections with lengths of 5, 4, 3 and 2 km each were included in this alternative. The road would have six lanes and be 24 km long. Under this alternative, road transport costs in the north-south corridor (excluding intramunicipal trips) are forecast to rise from Cr$ 2,410 million in 1985 to Cr$ 5,645 million in 2000. These relatively large numbers reflect the large growth in private automobile traffic on the road system and in the central area of Porto Alegre that is expected to occur with the provision of more road capacity, a phenomenon that is experienced worldwide. - 38 - (vi) Results 6.11 The NPV associated with each of the five project alternatives is presented below: Project Alternative (i) (ii) (iii) (iv) (v) NPV 5,269 4,409 7,070 2,362 3,388 (Cr$ million) at 11% It is interesting to note that all of the mutually exclusive project alter- natives have a positive NPV. This result is' not surprising in view of the fact that, without any of the project alternatives (para 6.05), the Region's only north-south trunk road, BR-116, an already congested arterial (para 2.26), would become completely saturated, and, as speeds diminished further, road transport costs would rise very steeply. A major increase in transport capacity in the corridor is clearly justified. The analysis shows that, among the five alternatives, the full rail mass transit system, alter- native (iii), is the preferred solution with an NPV of Cr$ 7,070 million, which is about 34% higher than the second best solution, alternative (i), the construction of a new bus expressway in the present railway right-of-way. 6.12 The rate of return for alternative (iii) is 24%, and the first year rate of return is 11%, indicating that the proposed project is well justified and that the starting date is optimal. The considerable spread between the first year rate of return and the rate of return over the project's economic life is basically explained by the fact that the benefits are quite substantial in the later years as road transport costs in the base case relative to the project situation build up quite rapidly. 6.13 Petroleum savings, which are accounted for in the benefits as part of the road vehicle operating cost savings, are worth mentioning separately. In 1985 (the first full year of train operations), the proposed project, when compared with the base case, would result in petroleum (diesel fuel and gasoline) savings of some 20.7 million liters, increasing to about 47.8 million liters in 1990 and to about 63.4 million liters by the year 2000. In terms of constant mid-1979 prices, these savings amount to US$5.5 million for 1985, US$13.8 million in 1990 and US$20.5 million in 2000. Such savings arise from the substitution of passenger trips made in electric powered (hydrogeneration) unit trains for trips which would otherwise be made in cars and buses and from less congested (more fuel-efficient) traffic conditions for the vehicle fleet on the existing road system. The above savings in 1985, 1990 and 2000 represent about 2.5, 6.0 and 7.7%, respectively, of the total annual gasoline and diesel oil consumption in MRPA at'the present time. These percen- tages would be even smaller if the savings were compared with estimates of total consumption in MRPA in the same future years. Thus, while these savings are substantial and are a part of the justification for the proposed project, when compared with total gasoline and diesel oil consumption in MRPA, let alone in the State and entire country, the impact is relatively small. The facts simply do not permit one to argue that the proposed project makes a significant contribution to reducing the country's dependence on imported oil--a claim often made to justify projects like the one proposed. The proposed project should be, and has been, justified on the basis of a total benefit-cost analysis. - 39 - B. Risk and Sensitivity Analysis 6.14 The following sensitivity tests were performed: (a) a 15% increase in the capital investment costs of each project alternative; (b) the elimination of all passenger travel time benefits from all the alternatives; (c) a shortfall in the forecast of car passenger diversion to the suburban railway in alternatives (iii) and (iv); specifically, in alternative (iii), 10% of total intermunicipal corridor car passengers are assumed to divert to the railway over the entire life of the project instead of, as originally forecast, 10% in 1985, increasing to 25% in 1990 and remaining at that rate until 2000 (the same relative decrease for alternative (iv)); and (d) a 15% decrease in the total benefits associated with each project alternative. 6.15 These changes tend to cover the most likely types and magnitudes of risk associated with the various project alternatives as well as the concern sometimes expressed that transport projects in developing countries should be justified on resource grounds without consideration of time savings (sensitivity test (b)). The results are summarized below: Net Present Value (Cr$ million, mid-1979) Project Alternative (i) (ii) (iii) (iv) (v) Sensitivity Test (a) 5,000 4,135 6,641 1,960 2,975 (b) 855 778 2,681 -649 -419 (c) 5,269 4,409 5,903 2,175 3,388 (d) 4,214 3,458 5,564 1,597 2,469 Note: at 11% rate of discount 6.16 Under all the sensitivity tests, the ranking of project alternatives remains the same as in the base case with the proposed project, alternative (iii), clearly emerging as the most economical solution. It is interesting to note that, under sensitivity test (b) (without passenger travel time savings), project alternatives (iv) and (v) (the light rail and new road solutions) become uneconomical. Also, with a shortfall in the diversion of automobile users to the suburban railway (sensitivity test (c)), the rail mass transit alter- native maintains its superiority but is only marginally better (12% higher NPV) than alternative (i). Clearly, automobile diversion is an essential component of the proposed project. The rates of return for alternative (iii) under sensitivity tests (a),(b),(c) and (d).are 22%, 17%, 19% and 22%, respectively. C. Distributional Impact 6.17 In order to assess the equity or distributional impact of the project, it is necessary to identify how the benefits are distributed among the dif- ferent income groups that are affected by the project. First, it should be - 40 - noted that passenger travel time savings for all modes and automobile vehicle operating cost savings accrue directly to the users, whereas public vehicle operating cost savings accrue directly to the bus companies. However, to the extent that there is effective competition among the bus companies in MRPA (and evidence suggests that there is), one can reasonably assume that bus tariffs would be lower under a regime of lower operating costs than what they would otherwise be without the project. Further, fare regulation relates tariff increases to operating costs. Thus, users of buses also benefit from reductions in bus vehicle operating costs, although somewhat indirectly. 6.18 While it was not possible to obtain a detailed income profile of the entire beneficiary population, average income by general user group has been estimated and is considered indicative. It is estimated that the average bus passenger has an average monthly income which is approximately two times the minimum wage, or Cr$ 4,200 and that the average car user has a monthly income which is about seven times the minimum wage, or Cr$ 14,700. Train passengers consist of persons who would otherwise travel by car or bus. (Over the life of the project, potential car and bus users are, on average, expected to comprise 19% and 81% of the total train ridership.) Therefore, the benefits that have been measured (passenger travel time savings, car operating cost savings and bus operating cost savings), can be related to two beneficiary groups with incomes corresponding to the average automobile and bus user. The distribution of benefits among these two income groups is as follows: % of Total Project Benefits Average Time Car Operating Bus Operating Income Savings Cost Savings Cost Savings Total Cr$ 4,200 34 0 39.5 73.5 14,700 8 18.5 0 26.5 42 18.5 39.5 100.0 6.19 As can be seen from this analysis, almost three-fourths of the total project benefits are expected to accrue directly or indirectly to the lowest income group. At least 34% is expected to accrue directly to this group in the form of less travel time. Thus, the project may be said to have quite desirable distributional effects. VII. AGREEMENTS REACHED AND RECOMMENDATION 7.01 During negotiations, agreement was reached with the concerned parties on the following: I. With the Federative Republic of Brazil (Loan Agreement) (a) to, not less than once per year, through EBTU, inform the Bank of its National Urban Transport Investment Plan (para 1.19); - 41 - (b) to not implement the Porto Alegre-Scharlau Highway Project until a proper feasibility study, satisfactory to the Bank, is completed and shows that this project is economically justified (para 2.21); (c) to, on the basis of an existing agreement with the Municipality of Porto Alegre, cause the latter to implement the agreed park- ing restrictions program for the CBD of Porto Alegre (para 3.19); (d) to onlend the Bank loan to TRENSURB/P.A. on the agreed terms (para 7.02(b) following) (para 4.07); (e) to, on the basis of a review to be undertaken not less than once every six months or whenever tariffs of passenger services compet- itive with those of TRENSURB are adjusted, if adjusted more frequently, authorize such adjustments in tariffs that would permit TRENSURB to cover at least all of its operating expenses as set out in the agreement under V(e) following (para 5.15); and (f) to take all action as necessary to assist in ensuring that all other agreements under V following are fully complied with. II. With EBTU (Joint Project Agreement) (a) to employ consultants, by December 31, 1980, to improve its internal accounting, budgeting and control procedures and to discuss with the Bank the consultants' recommendations (para 1.15); (b) to review and discuss with the Bank any proposals to substantially change its investment policy statement which was discussed during negotiations and recently approved by EBTU's Administrative Council (para 1.19); and (c) to inform the Bank through 1988 of any proposed passenger trans- port-related investments exceeding US$10 million to be undertaken in the influence area of the Bank project, provide the Bank with information regarding the expected impact on the Bank project and consult with the Bank prior to making a commitment (para 2.21). III. With the State of Rio Grande do Sul (Joint Project Agreement) (a) to cause the newly created NMTU to carry out its agreed functions (para 2.28); (b) to provide the necessary personnel to enforce the parking restrictions program for the CBD of Port Alegre, set out in the agreement under I(c) above (para 3.19); and (c) to discuss with the Bank the recommendations of the consultants on auto restraint measures for MRPA when they become available and, through NMTU, proceed to implement an appropriate plan of action as a result (para 3.24). - 42 - IV. With EBTU, RFFSA, GEIPOT and METROPLAN (Joint Project Agreement) (a) to provide the necessary technical services to NEP within TRENSURB/P.A. (para. 4.02). V. With TRENSURB/P.A. (Joint Project Agreement) (a) to accomplish the shift in TRENSURB's status from a subsidiary of RFFSA to a more diversified share structure according to the agreed schedule (para 5.03); (b) to carry out the agreed operating and staffing plans (para 5.08); (c) to have its accounts, including those with respect to the proposed project, audited annually by independent external auditors satisfactory to the Bank, and to have these accounts produced annually, and furnished to the Bank within four months of the end of TRENSURB's fiscal year (para 5.09); (d) to not enter into borrowings, other than the loan with EBTU, before the fourth year of full operations, and then only if such borrowings would not result in a debt service which is greater than the average annual net operating income plus depreciation for either the immediately preceding two years of operation or for a later 24-month period ended prior to the incurrence of such debt service, whichever is greater (para 5.10); and (e) to promptly, as necessary, seek the authorization of CIP to charge tariffs which shall provide it with revenues sufficient to cover at least all operating expenses, including overhead costs, adequate maintenance expenses, taxes, depreciation and interest (para 5.15). 7.02 As conditions for loan effectiveness of the proposed loan agreement, the following should be accomplished: (a) the establishment of NEP within TRENSURB with staff satisfactory to the Bank (para 4.01); and (b) the signing of the Subsidiary Loan Agreement, on the agreed terms, between EBTU and TRENSURB/P.A. (para 4.07). 7.03 With the agreements and understandings obtained, the proposed project would be suitable for a Bank loan of US$159.0 million. The loan would be amortized in 15 years, including a three-year grace period. April 17, 1980 BRAZIL SECOND URBAN TRANSPORT PROJECT Urban and Regional Population Size :in Brazil (1872-1970) Urban Population (Percent of Total) Municipa- Total lities of Regional Population Regional Urban Population b/ Population Official a/ 20,000 or (Percent of Total) (Percent of Region) Year (000) Definition more Northeast Southeast Frontier Northeast Southeast Frontier 1872 9,931 N.A.- 7.9 46.7 47.7 5.6 N.A. N.A. N.A. 1900 17,434 N.A. 10.0 38.7 55.2 6.1 N.A. N.A. N.A. .1940 41,236 31.2 16.0 35.0 58.4 6.6 N.A. N.A. N.A. 1950 51,944 36.2 21.1 34.6 58.5 6.9 26.4 42.9 28.0 1960 70,119 45.6 28.1 31.6 60.5 7.9 34.0 52.9 37.1 1970 93,140 55.9 38.8 30.2 60.5 9.3 41.8 64.4 46.8 a/ Officially, urban places are defined according to political administrative criteria. This administrative definition leads to a larger urban share than would be the case with a specific size limit, as the data in the third column, based on 20,000 or more inhabitants, indicates. b/ Official definition. Source: Bank Report, Brazil Division, Human Resources Special Report January 1980 - 44 - TABLE 1.2 BRAZIL SECOND URBAN TRANSPORT PROJECT Motorization Rates in Selected Urban Areas (Automobiles/i,OOO inhabitants) Urban Areas 1970 1975 1977 Metropolitan Areas of Rio de Janeiro and Sao Paulo 61.4 103.1 111.1 Metropolitan Areas of Recife, Porto Alegre, Belo Horizonte and Salvador 43.9 68.5 75.8 Metropolitan Areas of Curitiba, Fortaleza and Belem 39.5 69.4 N.A. Capital Cities outside the Metropolitan Areas 33.6 74.1 85.5 Medium Sized Cities 45.3 90.1 102.0 Brazil 25.0 52.1 59.9 Source: Ministerio Dos Transportes, GEIPOT, Plano Nacional de Tranportes, Diagnostico, VIII. Januarv 1980 - 45 _ TABLE 2.1 BRAZIL SECOND URBAN TRANSPORT PROJECT Population by Municipality in the MRPA (1970, 1974, 1979) 1970 1974 1979 Municipalities Number % Number % Number Alvorada 40,376 2.64 66,340 3.56 101,000 4.24 Cachoeirinha 31,020 2.03 46,370 2.49 78,000 3.28 Campo Bom 16,621 1.09 23,320 1.25 34,ooo 1.43 Canoas 153,823 10.04 188,o80 10.09 241,ooo 10.13 Estancia Velha 8,896 0.58 11,890 o.64 15,000 0.63 Esteio 34,587 2.26 40,390 2.17 53,000 2.23 Gravatai 52,454 3.42 71,570 3.84 118,000 4.96 Guaiba 33,692 2.20 43,740 2.35 69,ooo 2.90 Novo Hamburgo 85,353 5.57 104,960 5.63 139,000 5.84 Porto Alegre 885,944 57.84 1,015,710 54.51 1,217,000 51.12 Sao Leopoldo 64,309 4.20 72,910 3.91 94,0oo 3.95 Sapiranga 16,403 1.07 23,380 1.25 30,000 1.26 Sapucaia do Sul 41,750 2.73 58,690 3.15 89,ooo 3.74 Viamao 66,375 4.33 96,150 5.16 102.ooo 4.29 MRPA TOTAL 1,531,603 100.00 . 1,863,-500- 100.00 2,380,000 100.00 Source, METROPLAN 1979 Janrary 1980 - 46 - TABLE 2.2 BRAZIL SECOND URBAN TRANSPORT PROJECT The Location of Population Within the MRPA (1970, 1974, 1979) 1970 1974 1979 Location Number % Number _ _% Number % North-South Corridor 421,742 27.54 523,620 28.10 695,000 29.2 East-West Corridor 157,542 10.29 228,020 12.24 366,ooo 15.4 Porto Alegre 885,944 57.84 1,015,710 54.51 1,217,000 51.1 Viamao 66,375 4.33 96,150 5.16 102,000 4.3 MRPA 1,531,603 100 1,863,500 100 2,380,000 100 Source: METROPLAN, 1979 November 1979 - 47 - TABLE 2.3 BRAZIL SECOND URBAN TRANSPORT PROJECT Sectoral Distribution of Employment in the MRPA (1971-1976) 1971 1973 1976 - Sector % of total % of total I of total Primary 7 5 4.5 Secondary 39.3 34.5 33.5 1Tertiary 53.7 60.5 62 | Total ~439,262 503,851 608,764 Source: METROPLAN November 1979 - 48 - TABLE 2.4 BRAZIL SECOND URBAN TRANSPORT PROJECT Spatial and Sectoral Distribution of Employment in the MRPA (1973) Secondary Sector Tertiary Sector Total Location Number Number % Number % North-South Corridor 73,113 43.17 35,591 11.68 108,704 22.93 East-West Corridor 10,266 6.o6 8,180 2.69 18,446 3.89 Porto Alegre 85,813 50.66 258,193 84.75 344,oo6 72.57 Viamao 185 0.11 2,687 0.88 2,872 o.61 Total MRPA 169,377 100.00 304,651 100.00 474,028 100.00 Source: METROPLAN, 1979 November 1979 - 49 - TABLE 2.5 BRAZIL SECOND URBAN TRANSPORT PROJECT Distribution of Household Income By Municipality in the MRPA (In Percent) Income Classes .- Municipalities Cr$ 0-1;810 Cr$ 1.811-5,410 lCr$ 5,411-21,050 Above Cr$ 21,050 Alvorda 8 63 28 1 Cachoeirinha 3 50 43 4 Campo Bom 3 47 43 7 Canoas 4 46 1 42 8 Estancia Velha 6 50 37 7 Esteio 6 43 43 8 Gravatai 7 55 34 4 Guaiba 7 53 33 7 Novo Hamburgo 4 41 43 12 Porto Alegre 3 30 43 24 Sao Leopoldo 5 42 38 15 Sapiranga 3 38 51 8 Sapucaia do Sul 8 57 32 3 Viamao 10 60 28 2 Total MRPA 4 38 41 17 Note: 1974 distribution in terms of 1979 prices Source: METROPLAfl, 1979 November 1979 - 50 - BRAZIL SZCOND L'BAN T?P.AISFORT PROJECT Road Transnort Infrastructure in the MRRA BR-290 Km Number of L&res S.Antonio da Patrulka/Gravatai -Intersecao RS-118(Gravatai) 32 4 Intersegao 118-Porta Alegre 21 4 Co=omr section with BR-116 '16 2 BR-116 - Guaiba/Arroio dos Ratos 36 2 BR-116 Intersegao BR-290 - Guaiba/Bar- re. do Ribeiro 22 2 Porto Alegre - Canoas 14 4 + 2 service lanes Canoas - Esteic 7 4 + 2 service lanes Esteio - Sao Leopoldo 9 4 + 2 service lanes Sao Leopoldo - Acesso Novo Ham- burgo 8 4 Novo Famburgo - Estancia Velha/ Ivoti 4 2 BR-386 Canoas/Montenegro - BR-116(Canoas) 19 2 BR-122 Intersegao BR-li6(Scharlau)-Sao Leo- poldo/Portao 6 2 RS-239 BR-116(Novo Hamburgo) - Sapi- ranga/Taquara 31 2 RS-llc^ BR-116(Esteio/Sarucaia do Sul) -BR-290(Gravatai) 22 2 BR-290(Gravataf) - ES-040 17 2 !w,-/'rn _ '**o) _-_.,.<_ RS-020 RS-030(Cachoeirinha) - Gravatal'! /Taquara ?. PS-C Av. Antonio de Carvalho(P.A.) -Rs-118 (viamao, 17f RS-l18(Vianao) - Viarao/Osorio 2 Source: DNER, Porto Alegre January 1icU BRAZIL SECOND URBAN TRANSPORT PROJECT Number, Type and Fleet Size of Municipal and Intermunicipal Bus Companies in the MRPA (1979) ___. Type of Company = _ _ ______ LOCAL jtmber of Private Public/ I Fleet Average _________ __ ___ "on,nies _Ccrporation Private Corp. Partnership Individual Size Fleet/Company Inter-municipal - total 10 3 7 _ 1,168 117 Municipal Portc Alegrc 22 _ 1 21 _ 1,500 68 Alvorada 1 1 _ 11 11 Campo Bomn 1 1 _ 7 7 Canoas 1 _ 1 - 68 68 Gravatai 1 - 1 _ 8 8 Novo llamburgo 2 - 2 _ 85 42 Sao Leopoldo _4 __ _ 4 37 9 Total I 3 1 38 . 2,90o4 69 Source: Cadastro das Empresas de Transporte Coletivo. January 1980 BRAZIL SECOND URBAN TRANSPORT PROJECT Vehicle Occupancy, Trip Distance, Speeds and Fares by Mode in the MRPA (1979) Passenger Vehicle Occupancy (pass/veh) Average Pass Principal Other Daily Trip Distance Average Speed Fares VEHIlCLE Direction Direction Average (km) (km/hr) (Cr$/pass-km) Automobiles 1.33 9 Peak 1.66 1.27 30 Off-peak 1.27 1.27 45 Urban bus 45 13 0.2048 I'eak 85 30 15 Off'-peak 40 40 20 Interurbarn bus 42 21 0.3058 Peak 81 15 20 Off-peak 40 I40 50 Taxi Cr$ 10 + Cr$ 4.0/km Peak 2.32 (includes driver) 9 30 Off-peak 50 Train 178/train 12 29 o.1666 (at present) Soltrep! (WTPOT, Pnrto Al.egre January 1.980 45*1 *lS'%# 411M( S*i *5445 s. 5 -- m u S IN | |*a L in' "Is' gm O' K W K '1 I"W ,T 1 U T 37 1 1 s wr t |- SW a5 54 9*s nU c i s s "I OS £, U n | se S' I *5 - g s s 4 4 me. .me We W,¢eisss 400 uslt n v*e.8 ssit * s s &.I $2 oil 1" s W _,_ 'mS 'S et SAts t Iand Noh Departments ] Peronne t |AdhLinistration ||innc || ibreet |oWrkiL nstuo [ndPr o:amL I i | t t and Butget 1 0 | f P RelJLtions | | ~~~~~~~Evalualtionl Yi Sourcee EBTU November 1979 fh,o map nas lenPrep,ar b, tntT oPea,axa oS, - SANTA the nce ,eflC fthe-red-a f Th, d__f,t,e,n odadte . ,t the-er sd the AV OUTH heodoreoSOec c ths cap '~ '"NOVO HAMBADRGO-km42 A7 Ware -ekad .fcafntSaNOVO .F ~ -~AMERICA accefat,'fecfnfff hondesas /FENAC - km 40.7*-. B Ai 7o55.13c, Ca! :' ~~~~~~~~~~RIO GRANDE ' - Brcsilid INDUSTRIAL -km 39.7 DO SUL LIBERDADE -km~ '~ 38 .4oD. 5- N / ..~~~ (1K'-a-t.'-~~~~~ ~~ URU~G . N 6000 1255 ~~0 ff0 200 ~~~~~~~~~KILOMETERS Rl1,00 SINOS - km 35.8 ~S40LEOPOLDO SECOND URBAN TRANSPORT PROJECT kml 315 s1A_O LEOPOLDO ~~~~Porto Alegre Metropolitan Area t~UNISINOS - km 31.6 -~--- Passenger Railway Line - Phase One -~. Passenger Railway Line -Phase Two '------Freight Railway Line - . - - . -. . - . -+--- -'i- ~Freight Railway Line -Estimated Completion June 1980 -" ~ ~ ~ ~ ~ ~ ~~~ ~~ Railwaiy Station - Phase One Railway Station - Phase Two I, I / ~~~~~~~~~~~~~~~~~~Train Depot and Admrinistration Building SAPUCAIA - km 26.7 Port / ~~~~~~~ADOSLRivers APUCAIA DO ~~~~~~LJL - ~~National Roadls Built-up Areas / ZPASTEA)t-ksm24.7 State Roads -' Municipal Boundaries tIi MERCADOkmO~~~~~~~~~~~~~. . ____ ___ ___ ___ ____ ___ ___ ___ ____ ___ ___ ___ ___ ____ ___ ~ [0 T~~~ -l .~ kmOM6EF1