Jose Carlos Melo 34442 The Experience of Condominial Water and Sewerage Systems in Brazil: Case Studies from Brasilia, Salvador and Parauapebas The World Bank Jose Carlos Melo The Experience of Condominial Water and Sewerage Systems in Brazil: Case Studies from Brasilia, Salvador and Parauapebas August 2005 Acknowledgements Report written by: Jose Carlos Melo English Translation: John Penney Reviewed and edited by: Vivien Foster and Miguel Vargas-Ramirez ­The World BanK Production Coordination: Oscar Castillo and Beatriz Schippner ­Water and Sanitation Program, Latin America (WSP-LAC) Photography: Jose Carlos Melo, Jaime Salcedo, and WSP-LAC Design and layout: Ana Maria Origone Printed in Lima, Peru, by LEDEL S.A.C. The findings, interpretations, and conclusions in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. The World Bank does not guarantee the accuracy of the data included in this document and accepts no responsibility for any consequences of their use. Index I. Introduction...............................................................................................................................6 II. Brasilia: Full Institutionalization of Condominial Sewerage ..........................................................9 II.1 City Profile..........................................................................................................................9 II.2 Adoption of the Condominial Model....................................................................................9 II.3 Engineering Aspects.........................................................................................................11 II.4 Financial Aspects .............................................................................................................16 II.5 Social Aspects .................................................................................................................17 II.6 Operational Aspects.........................................................................................................18 II.7 Institutional Aspects .........................................................................................................20 II.8 Summary .........................................................................................................................21 III. Salvador de Bahia: Large-scale Experimentation with Condominial Sewerage ........................23 III.1 City Profile........................................................................................................................23 III.2 Evolution of the Condominial Model..................................................................................23 III.3 Engineering Aspects.........................................................................................................25 III.4 Financial Aspects .............................................................................................................33 III.5 Social Aspects .................................................................................................................34 III.6 Operational Aspects.........................................................................................................36 III.7 Institutional Aspects .........................................................................................................37 III.8 Summary .........................................................................................................................37 IV. Parauapebas: Community Mobilization for Condominial Water.................................................39 IV.1 City Profile........................................................................................................................39 IV.2 Adoption of the Condominial Model..................................................................................39 IV.3 Engineering Aspects.........................................................................................................45 IV.4 Financial Aspects .............................................................................................................47 IV.5 Social Aspects .................................................................................................................48 IV.6 Operational Aspects.........................................................................................................49 IV.7 Institutional Aspects .........................................................................................................52 IV.8 Summary .........................................................................................................................53 V. Conclusions.............................................................................................................................55 3 Shallower excavation works, Parauapebas Foreword This study presents the experience of applying the condominial approach to expand water and sewerage networks in three contrasting Brazilian cities. It is based on the findings of a Study Tour to Brazil organized in December 2003 for utility managers and public officials from the Peruvian water sector, who were in the process of applying large-scale condominial systems in Lima for the first time. The goal of the visit was to permit the Peruvian delegation to see mature and functioning condominial systems on the ground, and to interact with local policymakers, utility managers, residents and specialists, to obtain better idea of the challenges and potential advantages and disadvantages of implementing this system. The three cities visited ­Brasilia, Parauapebas and Salvador­ were chosen for their contrasting urban settings and specific experiences with the condominial model. The Study Tour was funded by the Bank Netherlands Water Partnership Program.The current study is offered as a vehicle for sharing the results of the visits with a wider audience of water-sector professionals. 5 I. Introduction The so-called condominial approach to the construction of the block (under sidewalks, front yards or backyards), while in water and sewerage networks was developed in Brazil during the case of water they are generally located under sidewalks to the 1980s as a response to the challenges posed by expanding allow for individual metering. This design permits the adaptation services into peri-urban neighborhoods. While the condominial of the network to local topographic conditions and different model has proved capable of meeting the considerable social and urbanization patterns. engineering challenges posed by these areas, it is also a generic alternative to the design of water and sewerage systems. Indeed, In addition, an integral condominial design contemplates the the Brazilian experience illustrates how the model has been decentralization of the drinking water supply or sewerage successfully applied to urban neighborhoods as diverse as the treatment facilities to avoid the costs associated with Rocinha slum in Rio de Janeiro and the affluent Lago Sul and transportation of fluids over long distances. This is in contrast to Lago Norte districts of Brasilia. the conventional approach, which emphasizes the concentration of fluids at a single geographical point. In the case of sewerage, The term condominial is sometimes misunderstood, and it decentralization takes place through the use of drainage basins, is therefore important to begin with a brief description of the taking into account factors such as land availability and local approach. The condominial water and sewerage system is based environmental conditions. In the case of water, the exogenous on two key concepts that differentiate it from the conventional nature of water resources may limit the decentralization of model. treatment processes. Nevertheless, there may be opportunities for decentralizing storage reservoirs with a view to achieving greater The first concept effectively redefines the unit to which service uniformity of pressure in the entire service area. is provided. Whereas conventional systems essentially provide services to each housing unit, condominial systems deliver The second distinctive concept behind the condominial approach services to each housing block or any group of dwellings that is the development of a much closer relationship between service could be termed a neighborhood unit or "condominium." This is similar to the concept of providing a single connection to an apartment building, except that in this case the condominium is Maintenance Equipment for Condominial Branches, Brasilia physically horizontal and institutionally informal. As a result of this novel concept, the public network no longer needs to run through every plot of land or to be present in every street, but merely to provide a single connection point to each city block. Therefore, the required length of the network is considerably shorter than that of a conventional system. It needs about half the length for sewerage and about a quarter of the length for water service. The household connections characteristic of conventional systems running perpendicular to the network are replaced by condominial branches running parallel to the blocks. Condominial branches for sewerage can be located in the most convenient part of 6 providers and users, encouraging the two parties to come to an Sao Caetano, Salvador de Bahia agreement to facilitate service expansion and adaptation to local needs and constraints. Thus, the condominium becomes not only a physical unit of service provision, but a social unit for facilitating collective decisions and organizing communal actions. Members of the condominium must select the appropriate design of the condominial service and commit themselves to complementary actions ranging from sanitary education to direct participation in the construction and/or maintenance process. The purpose of this paper is to illustrate the diversity of experience within Brazil in the application of the condominial model and to document the results to date. This is done by examining three case studies chosen because of their contrasting urban contexts and the substantial differences in the way the condominial model was applied. The three cases presented are Brasilia, one million people. In contrast to Brasilia, the condominial system Salvador and Parauapebas. in Salvador was adopted in a gradual and experimental manner, motivated by the extremely dense and chaotic urbanization In Brasilia, the nation's capital, the condominial approach was used patterns as well as the challenging topography of the city. to expand sewerage service to 500,000 people in some two dozen urban areas. The two most prominent features of this experience Parauapebas, a small but fast-growing mining town in northern were the achievement of universal access at very low financial Brazil, presents a rare example of applying the condominial cost to the utility company as well as the conscious, consistent model to the water sector and illustrates the potential operational way in which the technology was adopted by the utility company, advantages that condominial designs bring to water systems. providing service access to a wide socioeconomic spectrum. Moreover, the city was able to mobilize large-scale community participation in the construction of the network. The result was the In Salvador, the state capital of Bahia, the condominial model was very rapid expansion of water coverage at a fraction of the cost of also applied to sewerage on an unprecedented scale, serving over a conventional system. 7 Excavation works in the sidewalks, Brasilia II. Brasilia: Full Institutionalization of Condominial Sewerage II.1 City Profile that the residential plots would be sufficiently large to absorb effluents. However, this did not ultimately prove to be the case, Brazil's federal district is located on the central plateau at the leading to the discharge of untreated sewage into Lake Paranoá, geographical center of the country. The city is a planned capital which began to show signs of serious contamination, including that has experienced rapid population growth, from 140,000 eutrophication. inhabitants in 1960 to its present level of 2.1 million. From an urban development perspective, the city can be divided into three II.2 Adoption of the Condominial Model broad areas with contrasting characteristics. First, there is the City of Brasilia (the nation's capital proper) with only 19% of the By 1993, the lack of sanitation in the peri-urban areas, combined population of the district, and which consists of a monumental with growing environmental awareness of contamination in modern urban complex that has been named a World Heritage Lake Paranoá, prompted the city's public water utility CAESB site. Next come the 16 "satellite cities," which either pre-dated (Brasilia Water and Sewerage Company) to take serious action. the capital or were planned and established simultaneously. Solving these problems would entail connecting 600,000 people to In addition, there are several dozen small and medium-sized the sewerage network and building sewage treatment capacity for peri-urban neighborhoods (or bedroom communities), with 1.7 million people in the form of two tertiary treatment plants populations ranging from 1,000 to 10,000. on the lakeshores. The last group, with a total of 600,000 inhabitants, is of fairly A major concern was the huge financial cost of these solutions. recent origin, growing rapidly beginning in the early 1990s and In search of lower-cost alternatives, CAESB staff made field visits following an unusually ordered pattern of urbanization. In response to cities in the states of Rio Grande do Norte and Pernambuco to intense social pressure for housing, the local government of (Petrolina), which already had extensive experience with the use the time set up large-scale housing program offering "urbanized of the condominial model. This resulted in experimentation with plots" with a ready-made package of infrastructure services to pilot projects that were used to adapt the condominial approach to families that were willing to build their homes by a given deadline local conditions. Thereafter, the condominial model was applied on and to comply with predetermined building specifications. These a massive scale in Brasilia, both in the peri-urban neighborhoods plots were organized in a number of planned peri-urban areas or and in the more affluent areas of the capital. settlements and were generally provided with roads and sidewalks, and often with electric and water services. These communities Funding for the project was provided by the Caixa Econômica were gradually transformed into real cities, some of which rapidly Federal (Federal Development Bank) and the Inter-American became as large as some of the earlier satellite cities mentioned Development Bank, but also included contributions from both above. However, the urbanized plots were generally deficient with the capital and federal district governments. However, it was respect to the provision of sewerage and drainage infrastructure, not possible to determine the total amount invested or how it which over time led to deteriorating sanitary conditions. was distributed by system and source. From 1993 to 2001, an estimated 188,000 condominial sewerage connections were Furthermore, sewerage networks were inadequate even in made in the federal district, benefiting approximately 680,000 the capital itself. This was because the original city design was people (Table II.1). The condominial model in Brasilia has based on the adoption of individual septic tanks in the belief been applied to neighborhoods spanning the socioeconomic 9 spectrum, ranging from the modest homes of the peri-urban financial, social, and operational aspects of the experience are areas to the large luxury villas of the capital. The engineering, described in detail below. Table II.1: Overview of adoption of condominial approach in Brasilia 1993 to 2001 Population Estimated No. of No. of participants Condominial Public Administrative Region served connections condominiums at meetings branches (kms) Network (kms) Brasília 8,015 1,308 94 785 15.2 9.7 Brazlândia 8,410 1,529 69 897 16.5 7.0 Candangolândia 4,075 791 46 488 8.4 5.1 Ceilândia 23,591 4,711 177 2,345 51.2 13.9 Cruzeiro Velho 2,690 490 49 305 3.0 1.9 Gama 6,445 798 41 525 8.5 5.3 Guará II 8,742 1,644 86 911 16.6 11.9 Lago Norte 11,785 3,660 185 1,709 60.7 26.4 Lago Sul 12,133 2,092 103 752 46.0 28.4 Paranoá 38,143 11,475 274 3,744 69.0 28.4 Planaltina 51,026 9,992 377 4,742 89.7 26.2 Recanto das Emas 112,339 32,680 880 8,601 216.5 101.9 Riacho Fundo 58,720 18,831 401 4,759 103.7 63.4 Samambaia 146,663 45,455 862 10,516 282.5 199.8 Santa Maria 93,000 31,265 645 7,257 192.2 91.4 São Sebastião 55,000 13,085 408 4,733 109.3 36.2 Sobradinho II 23,338 4,070 137 2,040 33.6 17.8 Taguatinga 15,303 3,815 143 1,476 33.7 20.4 Total 679,418 187,692 4,977 56,585 1,356.3 695.0(*) (*) Note: This total contains some networks built to conventional specifications at the outset of the process. It is estimated that approximately 625,000 kilometers were built following condominial specifications. Source: CAESB (2001) Sinopse do Sistema de Esgotamento Sanitário do Distrito Federal 10 Although the issue of sewage treatment will not be discussed in large size (100,000 inhabitants), plus the fact that it aptly typifies detail in this paper, it is important to note that with the expansion the group as a whole. Furthermore, the Santa Maria case has been of the sewerage network, substantial progress was also made carefully documented. with respect to effluent treatment. As a result, the federal district is expected to achieve universal coverage of sewage treatment (a) Public network during the course of 2004. The processes used for effluent treatment combine anaerobic reactors and waste stabilization With regard to Figure II.2 shows the detailed design for one ponds, while the largest of the units (planned to serve over one of the sewerage collection micro-systems in Santa Maria. million people) relies on activated sludge technology. Extensive A micro-system in this context refers to the network that drains into a single collection point for treatment, elevation or experimentation aimed at adapting sewage treatment technology interconnection with the trunk network. The figure clearly illustrates to local conditions resulted in significant cost savings, for example, the standard features of the condominial model with public by substituting reinforced concrete structures with artificial ground network sewers bordering each city block in order to provide a strengthening techniques or by employing lower-cost materials. suitable collection point. The system uses a very low network A further advantage was the free provision of land for sewage density, occupying less than half of the city streets. As indicated treatment facilities given that most of federal district the land in Table II.1 above, this results in an average public network length belongs to the government. of approximately 2.8 meters per connection, considerably lower than the equivalent parameter for conventional systems, which II.3 Engineering Aspects is approximately 5.6 meters per connection. Adapting engineering aspects of the condominial approach to Brasilia proved to be relatively straightforward thanks to a Excavation works, Brasilia number of favorable characteristics of the local setting. First, the local topography is relatively uniform, with slightly sloping gradients well-suited to the drainage requirements of sewerage networks. Second, the unusually orderly pattern of urbanization resulting from the local government's housing program facilitated the introduction of sewerage networks, whereas construction work was made easier by the absence of roads and sidewalks in much of the area. Third, the condominial approach was introduced as a result of a conscious central decision, resulting in a coherent, uniform application of the model in the federal district. In order to illustrate the engineering issues that arose in the design of the condominial systems, this paper will focus on the Santa Maria system (Figure II.1). This system was chosen due to its 11 Figure II.1: Overall design concept for condominial sewerage system in Santa Maria Source: Mello, P.L. de (1998), Sistemas Condominiais de Esgotos e sua Aplicação na Capital do Brasil 12 Figure II.2: Illustration of typical micro-system in Santa Maria Source: Mello, P.L. de (1998), Sistemas Condominiais de Esgotos e sua Aplicação na Capital do Brasil 13 Wherever hydraulics and technical norms on drainage permitted, the network made use of the minimum pipe diameter of 100 Figure II.3: Alternative options for routing condominial branches millimeters, applied to 51.2 kilometers (equivalent to 56% of the total length of the system). An additional 26.5 kilometers of the network (representing an additional 29% of its total length) were rendered in pipes 150 millimeters in diameter, while a mere 2.7 kilometers of the network utilized pipes 250 millimeters in diameter (in other words, no more than 3% of the total). Conventional network design usually calls for a minimum pipe diameter of 150 millimeters. Furthermore, pipes were buried at a depth of up to 1.20 meters, but ensuring a minimum depth of 0.50 meters and allowing the substitution of the traditional high-cost manholes with simple inspection chambers. These accounted for 84% of Source: Mello, P.L. de (1996). Sistemas Condominiais de Esgotos e sua Aplicação na Capital do Brasil the total of 2,132 inspection devices that needed to be installed, providing one inspection point for every 24 meters of the network, which is more than adequate for operational purposes. Minimum depths in conventional networks are the main network through inspection boxes or chambers; and usually 1 meter to 1.30 meters. The cost of inspection chambers minimum depths of 0.60 meters (outside the plot) and 0.30 is only one tenth of that of traditional manholes, thanks to meters (inside the plot). Exceptions to the above were made only savings in excavation, compacting and back-filling as well as in response to highly localized circumstances that demanded in chamber materials. adjustments to accommodate specific features of the physical environment. (b) Condominial branches As for the location of the condominial branches, the requirements The condominial branches adopted not only in Santa Maria established by CAESB and the local government permitted a but in all the condominial systems throughout the metropolis great deal of flexibility. Three location alternatives were offered were required to conform strictly to standards of pipe location, to the population: routing through the backyard, the front yard hydraulic capacity, material specifications and building regulations. or the sidewalk (Figure II.3). The backyard option, and to a Few changes were made to the usual norms on hydraulic capacity lesser extent the front yard option, had the advantage of being employed by CAESB and other Brazilian utility companies. significantly cheaper to build due to shorter lengths and shallower To summarize: minimum diameter pipes of 100 millimeters depths of excavation. However, the potential disadvantage (virtually throughout); minimum slope of 0.005 meters per meter; is that they are inaccessible to utility company staff, placing use of PVC pipes for sewerage systems; interconnections with the responsibility for maintenance on the household. These 14 advantages and disadvantages have led to considerable technical It is interesting to compare the preferences of the peri-urban debate about the relative virtues of the different alternatives. settlements with those of the affluent neighborhoods in the capital (Lago Norte and Lago Sul), which participated in the project In the case of Brasilia, the final choice on the routing of the beginning in 1997. The predominant choice in those areas was branches was taken by majority vote at the level of each for internal installations, particularly for branches routed through condominium. Table II.4, for example, shows the choices finally backyards. Pragmatism evidently played a role in this decision made by the "condominiums" that were served by the condominial given the existence of very spacious backyards, reasonably system at that time (1996). These consisted of a universe of favorable topography and especially residents' desire to avoid the 114,000 inhabitants in the peri-urban settlements (excluding the costs involved in digging up and replacing expensive paved areas more affluent neighborhoods of the capital). The population clearly along the front of their plots. was divided over whether to place the branches at the front, the back or under the sidewalk. Overall, 51% opted for routing In their various forms, the condominial branches of Santa Maria under the sidewalk, while 43% chose the front yard. The covered a total length of 192.2 kilometers, which is to say backyard was the least popular option, accounting for only double the equivalent length of the public network. The network 6% of connections. While this decision was usually based design incorporated 23,350 inspection boxes, equivalent to one on weighing economic savings against the inconvenience of inspection point for every eight meters of piping, which in practice assuming responsibility for maintenance, in some cases choices amounted to one box for each connection on the network. were dictated by local topographic conditions or urban density This provided plenty of access to the system for inspection considerations. In some quarters, sidewalk branches were viewed during the construction process, as well as for cleaning during as conferring a higher social status. subsequent maintenance. Table II.4: Community choices regarding type of condominial branch Condominial Branches (units/type) Localities Population Backyard Front yard Sidewalk TOTAL M-Norte, Veredas, Areal e Outras 61,992 0 3,709 8,066 11,775 Paranoá 38,143 511 5,106 1,678 7,295 Vila Planalto e Guará 9,752 787 484 484 1,755 Candangolândia 4,427 0 96 1,099 1,195 Total 114,314 1,298 9,395 11,327 22,020 Percentage 6 % 43 % 51 % 100 % Source: CAESB (1996) Esgoto Condominial no Distrito Federal 15 II.4 Financial Aspects local circumstances such as suitable topography and the type of prevailing urbanization, both of which allowed the excavation The total cost of the public network amounted to US$1.7 million, depths to be reduced. whereas the cost of the condominial branches totaled US$2.8 million. The condominial branches accounted for 60% of the total CAESB took great care to guarantee that the fee structure for cost of the expansion. Given that these costs were covered by the connection to and use of the condominial systems clearly reflected beneficiary communities, this arrangement represents a significant the relative costs of the different options and that consumers reduction in the financial burden assumed by the utility company. benefited fully from the savings associated with the lower-cost At the same time, it is estimated that consumers paid no more for choices (Table II.3). The connection fee was equivalent to the the condominial branches than they would have had to pay for actual average costs of constructing the infrastructure for each conventional household connections. type of condominial service. With regard to the public network, the cost of one meter of The connection fee for the three different types of condominial network built equals US$19. This can be broken down into its service ranged from US$47 (backyard) to US$84 (sidewalk) in the constituent parts: 13% spent on inspections (inspection chambers peri-urban settlements, and from US$123 (backyard) to US$256 and boxes), 19% on materials and 68% on laying the network. (sidewalk) in the capital. This geographical differentiation reflects These low costs reflect the advantages of the condominial model the much larger plot sizes of the houses served in the capital, (in terms of reducing network size, excavation work and keeping necessitating a greater length of network per connection and disruption to a minimum), as well as the fortuitous presence of hence a higher cost. Households had the option of either having Micro-meter and Condominial Box, Brasilia Table II.3: Connection fee for alternative types of condominial connections (US$) Cost of implementation of condominial branches Backyard Front yard Sidewalk Peri-urban settlements 47 59 85 Capital neighborhoods 89 to 123 236 256 Sewerage charge as percentage of water bill Conventional 100 Condominial 60 100 Source: CAESB (2001) Sinopse do Sistema de Esgotamento Sanitário do Distrito Federal 16 the utility execute this work directly for a fee or of doing the work Beneficiaries of the Condominial Model, Brasilia themselves under supervision by the utility company. In the latter case, the beneficiaries did not have to pay the connection charge but were expected to cover the cost of materials. About 1.5% of the connections made under the program were constructed using this approach. Similarly, system use fees were differentiated to provide a discount of 40% on the standard sewerage charge for households opting for backyard or front yard condominial branches, and therefore assuming the associated maintenance responsibility. II.5 Social Aspects The success of the condominial approach in Brasilia was largely due to the coherent discourse adopted by CAESB and the The process of social intermediation began with high-level meetings local government authorities. From the beginning of the project, in each of the localities, with the participation of community leaders authorities carefully responded to the inevitable question: "Why and senior representatives of the utility company. substitute the existing system model for a new one in the federal district?" Local residents were largely unfamiliar with the The purpose of the meetings was to explain the condominial system new system, whose lower cost seemed to denote lower quality. and to present the three basic options available for its design. Authorities responded by providing a clear technical justification, In addition, meetings were used to elect a representative for each underscoring the political support for the system and promoting a condominium, who was a key figure in terms of encouraging consistent policy to apply the new model in the company's entire sustained social participation, facilitating agreements among service area. This last point was particularly important for avoiding neighbors and inspecting the works. During the meetings, the draft any potential perceptions of inferiority. The following excerpt from the terms of agreement was circulated. Each condominium had to project literature illustrates how effectively this issue was addressed. complete this document to confirm its commitment to participate as well as to indicate the specific details of its chosen system. "The new model was studied by CAESB technical staff Great care was taken to transmit the concepts to the population responsible for the sewerage division of the company, and visits in simple language, as well as to promote the exchange of ideas were made to localities where the model had functioned for a among participants and to allot time for answering questions. number of years (in particular, the city of Petrolina in the State of Pernambuco). Large-scale experiments were conducted in a number of different areas of the federal district before it was Some 5,000 meetings were held with a total of 57,000 promoted as a universally accessible system. It was approved participants. The meetings always took place in the evenings at the technical level and accepted by all relevant authorities in a location near the condominium (generally a school or one beginning with the governor. It was subsequently adopted as the of the condominium houses). At least two individuals nominated single model to be implemented throughout Brasilia." by CAESB were present: the coordinator plus an assistant. The 17 Peri-urban households, Brasilia former directed the meeting whereas the latter was responsible within the participating neighborhoods. Social mobilization was for administrative aspects (attendance records, recording the more limited, however, in the affluent neighborhoods of the federal names and addresses of elected representatives, etc). Members district, where people faced greater time constraints, as is typical of the CAESB supervisory team attended some of the meetings in larger cities. when necessary. In those communities that opted to build their own condominial In spite of the inevitable disagreements that arise between networks rather than have them constructed by the utility neighbors, not one of the condominiums failed to reach company, social inter mediation was more intense, often requiring consensus over the appropriate course of action and to sign additional meetings. the corresponding terms of agreement. Nor did the consensus- building process lead to any delays in the execution of the works. II.6 Operational Aspects Moreover, in general, the degree of mobilization tended to increase with the number of meetings held. As the numbers It is interesting to compare the different maintenance challenges of attending meetings grew, time spent on clarifying doubts and the condominial and conventional systems. Unfortunately, CAESB seeking adherence to the program tended to diminish. In many makes no distinction between the two types of sewerage systems cases, people had had limited contact with their neighbors before in its routine maintenance records, or at least not in a way that the projects began, but this changed as the program progressed. would permit rigorous performance comparisons. Nevertheless, In effect, the project was instrumental in building social capital Tables II.5 and II.6 below represent a specific effort by the 18 company to provide information on obstructions in the sewerage length. These differences suggest that there are fewer maintenance network. The first table covers a one-year period beginning in incidents per customer for the condominial system, even while the June 1996 for the western sector of the federal district, where number of incidents per kilometer is quite similar. approximately 60% of the condominial systems were operating at the time. The second table covers the entire Brasilia network A number of conclusions may be drawn from the information for the period May 1997 to December 1998. provided in Tables II.5 and II.6. A general finding that emerges from both sets of data is that, in condominial systems, obstructions When making comparisons between the conventional and are much more likely to occur in the public network than in the condominial systems, two considerations should be taken into condominial branches. Thus, the data show between 3.2 (Table account. First, in order for such comparisons to be valid, it is II.6) and 3.9 (Table II.5) times as many obstructions per kilometer necessary to have maintenance data for incidents occurring in of public network than per kilometer of condominial branches. the household connections of the conventional system since these There are two potential explanations for this: either the condominial play a similar role to the condominial branches in the condominial branches are less prone to obstructions, or the users are better system. This information is only available in the case of Table placed to resolve simple obstructions on their own initiative. II.6. Second, maintenance incidents are reported per kilometer Comparing conventional and condominial systems, the data show of network. However, the condominial system serves the same that the overall incidence of obstructions per kilometer is relatively population with a shorter overall network length, therefore, simple similar for the two systems, with the ratio for conventional versus comparisons of incidents per kilometer of network will tend to condominial ranging from 0.90 to 1.20. However, the data on understate the difference between the two systems. Indeed, for conventional branches, which are usually laid in the middle of water service, the length of the condominial network is at most 25% streets, show 1.7 times as many obstructions per kilometer as of the conventional network length, whereas in the case of sewerage condominial branches that perform the same function. Finally, the the percentage is approximately 45% of the conventional network time taken to perform repairs is roughly similar for the two systems. Table II.5: Occurrence of sewerage maintenance incidents in Brasilia (June 1996 to June 1997) Length Percentage Monthly Percentage Requests per (kms) of total service kilometer requests of total per month Condominial (A) Condominial branches 713 22% 782 9% 1.10 (B) Public network 398 12% 1,710 20% 4.30 Total (A+B) 1,111 34% 2,492 29% 2.24 Conventional (C) Public network 2,153 66% 5,970 71% 2.77 Overall total (A+B+C) 3,264 100% 8,462 100% 2.59 Source: Data provided by CAESB 19 Table II.6: Occurrence of sewerage maintenance incidents (May 1997 to December 1998) Length Total Monthly Requests per Average (kms) service service kilometer duration of requests requests per month repair (mins) Condominial (A) Condominial branches 1,080 18,666 933 0.86 141 (B) Public network 550 30,652 1,533 2.79 148 Total (A+B) 1,630 49,318 2,466 1.51 149 Conventional (C) Connecting conventional pipes 1,045 30,775 1,539 1.47 144 (D) Public network 2,172 57,282 2,864 1.32 178 Total (C+D) 3,217 88,057 4,403 1.37 161 Overall total (A+B+C) 4,847 137,375 6,869 1.42 158 Source: Data provided by CAESB II.7 Institutional Aspects Nevertheless, internal training activities served to raise awareness As indicated above, the condominial model in Brasilia was and understanding of the condominial approach. Furthermore, implemented in a deliberate, cohesive manner and received the management division specifically responsible for implementing support from the highest levels of the utility company and local the condominial process within the company survived these authorities. internal changes. This division oversees all activities related to the development of the model, planning its expansion and At the outset of the expansion process, CAESB had recently improvement and undertaking mobilization initiatives such as restructured into two specialized divisions (water and sewerage). delivering information and sanitation education to users. Since the condominial innovation was concentrated in the sewerage service, this structure facilitated the raising of awareness With regard to the institutional responsibility for the social about the condominial approach within the sewerage division. intermediation activities, CAESB initially employed firms with CAESB subsequently reverted to a pre-existing operational experience in handling participatory events of this kind, and structure, where water and sewerage operations were under subsequently created the conditions for a workers' cooperative to unified management and sub-divisions were based on geographic assume the task of meeting program requirements. At the same areas. This change shifted the balance of power within operations time, CAESB created a permanent in-house team which, during towards the larger and more established water supply teams, the implementation phase, was responsible for supervising the which had no prior experience with condominial systems. above initiatives and training professionals in this field of expertise. 20 II.8 Summary The Brasilia case illustrates the feasibility of applying the Middle-class household in Brasilia condominial approach on a large scale and obtaining substantial cost savings without any adverse consequences for subsequent operation and maintenance of the system. In its implementation of the model, Brasilia was evidently favored by an exceptionally orderly history of urban development. The successful application of the model also reflects a number of wise decisions taken by the local authorities. Moreover, the utility company made firm policy decisions with regard to technology choice, communicated these policies clearly to the public, adapted its internal structure accordingly, and applied the new model in a non-discriminatory manner in the entire service area. Furthermore, the utility company provided residents with clearly defined alternative system designs and ensured that the associated cost savings were adequately reflected in lower user fees. All of these factors undoubtedly contributed to increasing the acceptability of the model among the general public. Condominial Box (seen open in the right hand lower corner), Riacho Fundo, Brasilia 21 Condominial Branch (showing a "T" connection and an Inspection an Cleaning Access Point ­TIL in portuguese), Brasilia III. Salvador de Bahia: Large-scale Experimentation with Condominial Sewerage III.1 City Profile the World Bank, the Inter-American Development Bank, the Japan Bank for International Cooperation, Brazilian development Salvador, the capital of the State of Bahia, is the largest city in banks (CEF and BNDES) and the Bahia State Government. northeastern Brazil, with a population of approximately 2.5 million. This led to the creation of the Bahia Azul (Blue Bay) program, As Brazil's first capital, Salvador is also one of the country's oldest whose objective was to prevent the contamination of the cities, with an exceptionally rich cultural and historical heritage. Bahia de Todos os Santos by installing a city-wide sewerage The city is located in a beautiful natural setting, perched between network. Responsibility for implementation of this program fell the Atlantic Ocean and the Bahia de Todos os Santos (Bay of All to the Bahia Water and Sewerage Company (EMBASA), the Saints). Its cultural and geographic wealth has made it a major official concessionaire for virtually all water and sewerage tourist destination. systems in the state. Nevertheless, over half of the population lives in unsanitary urban Table III.1 demonstrates the size and basic features of the slums. These densely populated peri-urban settlements are program. The figures, taken from the EMBASA database, refer precariously built, often clinging to steep hillsides or occupying to the situation of the works in June 2000 in each of the 21 other unsuitable land. Many of the dwellings are built in a chaotic drainage basins where sewerage systems were installed. fashion, virtually on top of one another (on three or four levels) and However, these figures should be viewed with caution, particularly are packed together on all sides. Due to the spontaneous and when they refer to the condominial model since the technology disorderly development of these neighborhoods, little attention used differed from basin to basin, and since additional works was paid to the installation of sewerage systems. As a result, household sewage is typically discharged to stormwater drains in were undertaken in some of the basins both before and after the violation of all regulations on the separated systems. Blue Bay program. Moreover, the figures reflect an imbalance in the financial outlays in the different basins because the cost of Evidently, these types of neighborhoods present a major challenge large interceptor pipelines, used to concentrate sewerage flows for the development of sewerage systems. In particular, the at a single point in the city, are also included in the costs for application of conventional sewerage in such a setting appears to some of the basins. In order to overcome these deficiencies in be virtually impossible, providing a pretext for ignoring the problem. the aggregate information, subsequent discussion will focus on Nevertheless, by the early 1990s, the discharge of sewage into detailed information for two specific basins. the stormwater drainage system had led to serious environmental problems in the fragile ecosystem of the Bahia de Todos os In contrast to Brasilia, the adoption of the condominial system Santos. Once the system started to contaminate bathing beaches, in Salvador was not the product of conscious decision-making. it became a threat to the leisure and tourism industries central to Indeed, at the outset of the Blue Bay program, little was known the local economy, motivating local authorities to take decisive about the system. Moreover, there were many misconceptions action to resolve the problem. about the model, which was viewed as a "sewerage system built inside dwellings" or a "low quality system for the poor." III.2 Evolution of the Condominial Model As a result, the model was adopted gradually over a long period of time, and in response to specific constraints and challenges In response to this situation, the Bahia State Government experienced on the ground. Three distinct phases can be negotiated a US$400 million package of financial support from identified in its evolution. 23 Table III.1: Sewerage basins in Salvador Condominial Total Investment per Investment Localities Population Households Public network (2000) (projected) (kms) branches investment(*) household per person (kms) (US$ million) (US$/cx.) (US$/cap.) Alto Camurugipe 163,726 43,833 105.1 200.0 13.1 298 80 Alto Pituaçu 155,000 13,077 71.8 85.0 8.8 671 57 Aratu / Macacos 8,638 1,268 10.9 3.1 2.6 2,029 298 Baixo Camurugipe 100,000 13,465 47.3 91.7 8.0 594 80 Baixo Jaguaribe 20,815 1,000 12.2 0.0 7.5 7,522 361 Baixo Pituaçu 29,000 3,200 15.5 14.5 3.5 1,101 122 Calafate 82,000 4,139 34.3 51.8 6.4 1,546 78 Campinas 90,000 13,610 47.0 72.4 11.1 816 123 Cobre 80,000 8,608 43.6 50.8 9.1 1,052 113 Comércio 58,000 7,682 62.6 14.0 5.6 732 97 Itapoã 25,500 2,800 53.5 10.0 6.2 2,206 242 Lobato 64,000 2,867 45.1 47.2 6.7 2,337 105 Mangabeira 140,000 15,413 126.7 59.6 19.4 1,259 139 Médio Camurugipe 55,000 4,620 35.7 56.7 7.5 1,625 136 Médio Jaguaribe 16,504 2,100 13.3 14.0 2.8 1,344 171 Paripe 98,484 11,397 54.5 92.5 7.8 680 79 Península 115,700 7,495 35.5 99.0 14.8 1,975 128 Periperi 124,000 11,787 75.4 77.2 19.1 1,617 154 Pernambués 70,000 13,581 47.3 66.0 6.9 506 98 Saboeiro 99,686 10,300 14.9 11.6 9.8 947 98 Tripas 133,000 21,732 74.0 60.0 9.8 450 73 Total 1,729,053 213,974 1,026.1 1,177.1 186.3 871 108 Source: EMBASA (2000) Indicadores dos Sistemas de Esgotos - Programa Bahia Azul (*) Does not include sewerage treatment costs. 24 During the first phase, the condominial model served to fill gaps the full know-how to be transferred and absorbed. Instead, in places where urbanization patterns effectively ruled out the numerous amendments and adjustments had to be made on use of conventional systems, for example, along rough tracks a day-to-day basis as construction progressed. The various and up steep alleys and flights of steps rising between houses. contractors proceeded to make different changes where and when Nevertheless, the system, though inspired by the condominial they were considered necessary. While the main elements of the model, did not constitute an integral application of this model. condominial model were introduced in this piecemeal fashion, A key difference was the absence of any clear delineation between the overall result was that the procedures and standards forming the condominiums that form the fundamental building blocks of the backbone of the condominial system eventually prevailed. condominial networks. As a result, the condominial branches were almost always confused with the public network and there III.3 Engineering Aspects were either too many or too few of them used in the design of the systems. In addition, local community participation was confined The engineering designs used to adapt the condominial to specific localized agreements negotiated to allow pipes to be model to the challenging local conditions of Bahia showed laid on private property. considerable creativity and ingenuity, entailing new approaches such as pressurized sewerage networks, pipes that "cut across" The second phase began in 1994, when the Blue Bay program was residences, "over-ground" pipes and the use of large vertical drop implemented on a much larger scale. This period was characterized pipes (tubos de queda) that will be described in greater detail by the formal introduction of the condominial model, which below. However, the condominial principles were applied only at thereafter would be present to a greater or lesser extent throughout the local level, where challenging circumstances demanded, and the life of the program. Throughout this phase, employment of were not reflected in the macro-design of the project. The latter the condominial model was almost entirely confined to the highly was based on the idea of concentrating the entire effluent of the precarious urban slums, while use of the system in more typical city at a single point where it could be discharged via a major urbanized areas was limited. Moreover, the extent of adoption of condominal principles and practices varied considerably among the different contractors involved in the project. Household meter in Sao Caetano, Salvador de Bahia From 1997 onward, the condominial model effectively began to be applied throughout the city, regardless of the type of topography, style of urbanization or incomes of the populations benefiting from the program. Indeed, the model would eventually be applied in locations as diverse as the Bacia do Comércio (one of the oldest and most traditional parts of the city, incorporating the historic center) and the much poorer outlying peri-urban settlements. The essential dynamic was one of growing enthusiasm for the system as its advantages became apparent in the areas where it was first applied. As a result, there was insufficient time for 25 marine outfall, which would entail high transportation costs Figure III.2: Alto Camurujipe Basin to centralize the sewage. An alternative approach that was not considered --but which would have been more in line with condominial principles-- was to keep the effluent decentralized within the local basins and drain it into small-scale treatment plants. The following discussion presents an overview of the engineering designs, focusing on two specific and contrasting basins. The first is Bacia da Peninsula, which is one of the oldest slums in Bahia. It is characterized by low-lying land prone to flooding, and Figure III.1: Peninsula Basin is relatively sparsely populated by a diverse population with varying income levels. The second is Bacia do Alto Camurujipe, which is one of the largest and poorest unplanned urban settlements in the city (see photo next page). It is a complex urbanization occupying steep hillsides, with precarious structures, high population density and widespread poverty. Figures III.1 and III.2 show the locations of the two sub-basins in the city. (a) Public network Figure III.3 illustrates the design of the public network in Alto Camurujipe basin. It clearly illustrates the key characteristics of the system, with the sewers skirting each block of housing to provide a single sewage collection point. As a result, the density of pipes needed to serve the areas is low. In general, these networks were conceived and located in situ in accordance with condominial guidelines, but tended to employ larger-diameter 26 Illustration of Bacia do Alto Camarujipe pipes than would be strictly necessary for a condominial design sidewalks were exploited for the basic network, using maximum --a minimum of 150 millimeters as opposed to the 100 millimeters depths of one meter and minimum depths dictated by the outflows that would have been justified by the hydraulic conditions. It from the condominial branches. A key challenge was to protect is interesting to contrast some of the design features of the two the pipes of these shallow networks, which had to stretch from basins, which largely reflect the different topographical conditions one sidewalk to another under roads, that are normally subject of each. to heavy vehicle loads associated with traffic. Two alternative strategies were used. The first was to lower the depth of the Bacia da Peninsula pipes by one meter where possible before crossing the street and "recovering" the shallower depth with minimum slopes on In Bacia da Peninsula, the key engineering challenge was posed the opposite sidewalk. The second strategy was to use special by the flat terrain just above sea level, combined with a very protection for the pipes that had to be routed across the street shallow water table with much of the sub-surface already occupied at the shallower depth. When neither of these alternatives proved by other infrastructure. It was therefore critical to keep the network feasible, the maximum depth of three meters was assured by as near to the surface as possible. This was achieved by limiting using small subterranean automated pumping stations placed the depth of the installations to three meters. To this end, the on public lands to avoid expropriation issues. 27 Figure III.3: Design of public network for Bacia do Alto Camarujipe 28 Bacia do Alto Camajuripe Illustration of the area requiring pressurized sewers In Bacia do Alto Camajuripe, the key engineering challenge was posed by the very steep gradient and dense urbanization, which frequently occupied the banks of local rivers and creeks. One advantage of the topography was the possibility of installing long stretches of pipes under the sidewalks, where the shallower depth meant that standard manholes could be replaced by much lower cost inspection chambers. Nevertheless, the nature of the terrain posed many other difficulties, leading contractors to develop a range of innovative approaches, which --while they diverge from standard practice-- illustrate a fundamental commitment to Channel meeting the needs of the population under any circumstances. There are four particularly noteworthy examples, namely: pressurization of the sewerage system; large vertical pipes; surface routing of sewers; and aerial networks. Pressurized systems. The typical sub-basins in this area have a semi-circular shape, with dense urbanization occupying the Pressurized pipe natural drainage areas along local creeks, where trunk collectors would normally be routed (see photo). Wherever space constraints prevented a traditional gravity-based collector from being installed, Large vertical drop pipes. One of the sub-basins in this area pressurization of the sewerage system was used as an alternative. presented a further peculiarity calling for a unique solution. This led to the installation of two independent concentric collector This was the case of an area whose natural outlet consisted of systems within the micro-basin, separated by a strip of land. a ravine approximately 20 meters deep, the bottom of which The higher one, which accounted for at least 80% of the total load, was occupied by an informal settlement. The solution to this discharged into a pressurized interceptor located on the available problem was to use a vertical drop pipe consisting of a narrow banks, making it possible to transport the effluent at a 150-millimeter cast iron pipe fixed to the wall of the ravine minimal depth and slope. Further down, because of the narrow with clamps, protected by concrete casing and ending in a diameter of the condominial branches (almost exclusively 100 pressure-break box at the base, which also served as an initial millimeter), the remaining 20% of the load, flows entirely by gravity manhole. The sewerage network serves the informal settlement until the junction where the effluents are collected and redirected and the drop-pipe structure has functioned perfectly since it to the main system by a small pump. was installed in 2000. 29 Surface sewers. In other parts of the basin, the natural Illustration of surface sewers drainage channel between houses was so narrow that it could not contain normal stormwater flows without flooding the surrounding houses. The solution to this problem was to raise the level of the houses above the flooding level to permit construction of the collector above ground without disturbing the sewerage system of the abutting houses. The superficial Inspection sewer was protected by concrete encasement, which had Chamber the added advantage of providing a sidewalk to facilitate the movement of local residents (pictures of surface sewers). This Inspection Box approach proved to be much more cost-effective than burying the collector in the rocky terrain and has since been widely used Collector both in Salvador and other Brazilian cities with rocky terrain. Channel Illustration of surface sewers Aerial sewers. An extreme case of the aforementioned approach had to be used in cases where urban density was even higher and water runoff was not adequately covered or channeled. In these cases, flooding is usually prevented by placing the lower floors of the houses above the runoff level. This normally enables the Inspection Box installation of network branches bordering the houses, above the water runoff channels, either by using pipes attached to the house walls or supported on pillars. (b) Condominial branches The combined length of condominial branches installed in Salvador Channel had reached approximately 1,177 kilometers by 2000 (Table III.1) and may now be as much as 2,500 kilometers. Collector EMBASA established general standards for the design of condominial branches for the contractors executing the works. Rather than allowing the diameter of the pipes to be determined 30 by local hydraulic conditions in each case, EMBASA Figure III.4: Illustration of condominial branch design in limited the use of 100-millimeter pipes to the first 20 Bacia da Peninsula houses in any condominium, requiring diameters to increase to 150 millimeters for the next 30 houses, and 200 millimeters thereafter. These requirements were driven by the deep-rooted cultural tradition of combining stormwater and wastewater flows, which raised concerns about using narrower diameters for sewerage only. Furthermore, the general preference in Salvador was for PVC pipes. Notwithstanding these general guidelines, in practice there was considerable variation in the application of the model, reflecting in part the diversity of contractors executing the large-scale program as well as the nature of the challenges posed in the different areas of the city. In general, the depth of the condominial branches was the minimum compatible with discharge requirements from the dwellings, up to a maximum of one meter. Furthermore, inspection boxes were placed at every point of inflow, as well as at network junctions, and near fall pipes in cases where vertical drops exceeded 0.5 meters. A key issue for the communities was the adequate replacement of sidewalks and staircases that had been damaged during the course of the installations. Regarding the location of the condominial branches, there was general agreement that unconventional routings would be adopted wherever local conditions demanded. In some cases, this entailed routing under steps and narrow alleyways between houses, via the front or back yard, or even inside the houses. Given the numerous technical restrictions on the network design, the usual practice (observed in Brasilia) of allowing households to choose the preferred routing of the network had to be abandoned in Salvador because there was only one viable alternative in each sub-basin. 31 Figure III.5: Illustration of condominial branch design in Bacia do Alto Camajuripe 32 The other key consideration was the need to achieve (as far as possible) a complete separation of stormwater and wastewater Table III.2: Cost data for Bacias da Peninsula and do Alto Camarujipe systems, which went entirely against the local cultural tradition of combined networks. Alto Peninsula Camurujipe Figures III.4 and III.5 illustrate specific examples of design Total population served 115,700 163,726 issues for condominial branches. In the first example, taken Total number of dwellings served 23,217 43,833 from Bacia da Peninsula, the over-riding concern was to keep the network as shallow as possible to avoid the problems Total length (kms) posed by the high sea level and high water table and to · Overall length 218.8 394.3 prevent the precarious dwellings from collapsing. In the second example, taken from Bacia do Alto Camajuripe, the key issue · Public network 50.4 79.0 in this high-density area was to determine the actual layout of · Condominial branches 138.6 240.0 the pipes. Gradients did not represent a problem whatsoever, · Internal installations 29.8 75.3 since the natural slope was always greater than the established hydraulic minimum (0.005m/m) and constantly uniform pipe Average length (m) depth could be maintained. · Total per dwelling (m) 9.43 9.00 III.4 Financial Aspects · Network per dwelling (m) 2.17 1.80 · Branch per dwelling (m) 5.97 5.48 The engineering works for the two basins described in the · Internal installation per dwelling (m) 1.29 1.72 preceding section entailed the laying of 219 kilometers of pipes in the Bacia da Peninsula and 394 kilometers in the Total investment (US$ million) Bacia do Alto Camurujipe, including the public networks, · Overall investment 7.3 11.2 condominial branches and internal installations (Table III.2). Approximately three meters of condominial branches were laid · Public network 2.4 4.4 for every meter of public network and the public network was · Condominial branch 3.8 3.3 not much longer than the pipe required for internal installations. · Internal installations 1.1 3.5 This averages out to approximately nine meters of pipe overall for each dwelling served in both of these basins, of which only Breakdown of investments (US$) about two meters belongs to the public network and the · Overall total per dwelling 316 256 remainder to the condominial branches. · Per meter public network 48 56 The overall investment costs amounted to US$7.3 million for · Per meter condominial branch 27 14 Bacia da Peninsula and US$11.2 million for Bacia do Alto · Per meter internal installations 39 46 Camurujipe. This is equivalent to a cost per dwelling connected of US$316 and US$256, respectively, excluding the cost of Source: EMBASA, 2000 33 to facilitate this process, EMBASA offered two options: either Household meter in Sao Caetano, Salvador de Bahia pay the full sewerage tariff with no responsibility for maintenance of condominial branches or accept maintenance responsibilities in exchange for a discounted fee. The discount package proved very attractive to the local population and was widely accepted, although in practice the population has not always honored its maintenance commitments (see next section). III.5 Social Aspects The Blue Bay program required social mobilization on an unprecedented scale in Brazil. During the program as a whole, some 10,000 condominiums were established. A series of meetings was organized to address the different program issues. For example, given the urbanization density, the contract work downstream trunk collectors, which varies substantially among was often very disruptive, requiring careful negotiation with the basins, depending on their geographic location. These low unit community. Meetings also focused on securing the signing of costs were achieved by minimizing the length of the public network the terms of agreement document that governed the future versus the condominial branches as noted above.The cost per relationship between EMBASA and each respective condominium. meter of the public network (approximately US$50) is more than Nevertheless, in retrospect, the community intermediation activities twice as much as the cost per meter of the condominial branches were not intensive enough to prevent some serious problems (approximately US$20). arising with the subsequent operation of the system. These include With regard to the fees paid by the population, in the case of Steep Change of Condominial Branch Level Salvador, no connection fees were charged since the cost of the (pipe attached to wall), Sao Caetano, Salvador de Bahia investments was entirely subsidized by the Blue Bay program. Nevertheless, households were responsible for covering the costs of internal installations required to re-plumb their sanitary installations so that they drained into the new sewerage system instead of the old stormwater system. EMBASA usually did this work, and the beneficiary households and the utility company agreed on monthly payments. Prior to the Blue Bay program, most households had some type of drainage system since they were using the stormwater network free-of-charge. With the installation of the sewerage network, households had to pay for this service for the first time. In order 34 Children playing on the street, Sao Caetano, Salvador de Bahia the relatively low collection rate and the lack of adherence to the sewer, as well as the ongoing obligation to pay a monthly community maintenance obligations. sewerage bill. Moreover, since the resulting benefits took the form of general environmental improvements, they were not always First, as mentioned above, the local conditions prevailing in immediately tangible to the household. For all of these reasons, it Salvador made it particularly challenging to motivate household is not surprising that only around 30% of households in the Blue connection to the new sewerage system. Most households Bay program intervention area have connected to the sewerage in the affected areas already had internal sanitary installations network to date. Increasing this connection rate will require connected to the stormwater system, which was the source of renewed efforts to provide sanitary and environmental education to the environmental contamination observed in the Bahia de Todos the local communities. Additional financial incentives may also be os Santos. From the household's perspective, this arrangement needed. provided all the advantages of a sewerage system at zero cost. At the same time, switching to the new sewerage network entailed Second, as previously noted, many households opted to assume both monetary costs and disruption of service associated with maintenance responsibilities in return for a lower monthly re-routing of plumbing from existing sanitary installations to sewerage fee. However, in practice, these households were not 35 always willing or able to address the maintenance needs arising about their maintenance responsibilities. On the other hand, the in the condominial branches, and frequently requested company company could have been more consistent in switching users who maintenance crews even though they were theoretically not requested maintenance crews to the full sewerage tariff as they entitled to do so. In addition, the use of similar diameter pipes were imposing maintenance costs on the company. for condominial branches and public networks sometimes made it difficult for local residents to determine which segments of III.6 Operational Aspects the network were their responsibility. The situation could have been handled better in a number of ways. On the one hand, The gradual, unplanned approach that characterized the greater efforts could have been made to inform local residents adoption of the condominial model in the Blue Bay program was repeated when the newly constructed systems were handed Sewerage pipe attached to wall, Sao Caetano, over to EMBASA for ongoing operation and maintenance. There Salvador de Bahia were no special efforts to train operational staff on the use of the new system or to adapt existing maintenance procedures to the different characteristics of these networks. As a result, the traditional operational practices developed for conventional networks were also applied to the condominial systems. Gradually, as awareness of the new designs increased, adequate operation and maintenance procedures were adapted. The three main operational problems experienced with the condominial networks were essentially the same as those experienced with the city's conventional system, and reflect the customs and practices of the local population. First, owing to the history of combined wastewater and stormwater drainage, some parts of the sewerage network were erroneously connected to the stormwater network. This led to several problems, including overload of the network during storms, as well as unwanted litter and earth flushed into the pipes by rainfall and the sedimentation caused by heavier materials settling inside the pipes. Second, the fact that less than a third of the population had connected to the network meant that sewage flows were lower than originally envisaged and therefore not always adequate to transport solid residues. Third, the local population sometimes misuses sanitary facilities to discard solid items, revealing deficiencies in their sanitary education. 36 III.7 Institutional Aspects Gradients, Sao Caetano, Salvador de Bahia In contrast to the Brasilia case, the local utility EMBASA never made any formal decision to officially adopt the condominial model as a company-wide policy. Nevertheless, following the success of the Blue Bay program, the system has subsequently been adopted in dozens of small towns within the State of Bahia. The use of the model can thus be characterized as much more opportunistic than in Brasilia, and it has not been reflected in or absorbed into the institutional structure of the utility company. Regarding the implementation of the Blue Bay program itself, the large scale of the initiative, together with the five-year time frame, meant that many different contractors worked simultaneously in different parts of the city, including five or six of Brazil's largest construction firms. When the project began, none of these actors was familiar with the condominial model or with the socioeconomic context of peri-urban slums. This caused that the adoption of the model to be gradual and experimental, proceeding along different parallel lines in areas managed by different contractors. III.8 Summary The case of Salvador de Bahia is probably the largest single application of the condominial model. It is also noteworthy for the enormous engineering challenges posed by the city and the extraordinary ingenuity with which they were overcome. A key problem in implementation has been the difficulty in convincing the original rationale for the program. However, this problem is local residents to assume maintenance responsibilities associated not specifically related to the use of condominial technology, but with the networks. The main failing of the experience has been rather is indicative of a wider challenge affecting the expansion of the low rate of connection to the sewerage network, undermining sewerage networks. 37 Excavation works to install the Public Network, Brasilia IV. Parauapebas: Community Mobilization for Condominial Water IV.1 City Profile water from the local river, whereas another 15,000 people relied on contaminated water from wells, public fountains or municipal Parauapebas is located in the State of Pará, 650 kilometers from tankers. The sewerage network served only 6,400 residents in the the state capital, Belém. Established less than 20 years ago, the downtown area, whose untreated effluents were returned directly city owes its history and rapid growth to the development of rich to the river system. Elsewhere, residents developed their own iron ore deposits by the Companhia do Vale do Rio Doce (CVRD). on-site sanitation solutions, ranging from insanitary pits to septic Since its formal incorporation as a municipality in 1988, the tanks, which often overflowed onto streets and public areas of population has grown from 20,000 to current levels approaching the city. The precarious conditions of both services contributed to 100,000. Following its recent privatization, CVRD diversified its the proliferation of diseases, particularly gastrointestinal infections, mining activities to include copper, gold, nickel and manganese, despite major efforts in public health education. boosting the development of two neighboring towns (Canaã dos Carajás and Eldorado) and creating one of the most populated In 1993, a collaboration agreement was signed between CVRD mining centers in northern Brazil. Overall, the area can be and the Municipal Prefecture of Parauapebas (PMP) with a view characterized as a boom town with a frontier spirit. to securing a World Bank loan to finance the expansion of water and sewerage services in the city. A loan of US$7.8 million (later The City of Parauapebas is a rare example of a well organized increased to US$14.5 million) was disbursed to enable CVRD to urban area among the municipalities of north and northeastern finance the necessary works. In return, PMP agreed to earmark Brazil, on account of its careful urban design and well-planned 25% (later 27%) of its income from mining royalties to amortize infrastructure, including a modern road and street system, water the debt. The original project design entailed construction of a supply, sewerage services, stormwater drainage networks and river water intake of 230 liters per second connected to a garbage collection, as well as health and education facilities. This drinking water treatment plant and a 6,000 cubic meter reservoir. situation reflects the municipality's substantial financial resources Conventional water and sewerage networks would be developed obtained from mining royalties along with the effectiveness of the to serve 90% of the city's population, eventually channeling last two municipal administrations, which are generally credited effluents back into decentralized waste stabilization ponds in with having introduced modern and efficient management. five sub-basins of the city (Figures IV.1 and IV.2). IV.2 Adoption of the Condominial Model By 1996, the water production system had been completed. However, it became apparent that the remaining funding When CVRD was established in Parauapebas in the late 1970s, would not cover the cost of the planned conventional water the company invested in a water and sewerage system for and sewerage systems. Construction on the network therefore residents. There was no charge for the systems. Nevertheless, came to a standstill amid allegations of financial irregularities. the rapid growth of the city quickly rendered these initial systems As a result, CVRD undertook a feasibility study for switching obsolete. Thus, by the early 1990s, they were deteriorating and the network designs over to the condominial system with a served only a small area of the city. In terms of water supply, view to completing the project within the original budget. Once only 1,000 residents occupying CVRD company condominiums CVRD was convinced of the technical merits of the condominial had access to a regular supply. An additional 5,000 residents system, the decision to adopt one of the two systems depended of downtown neighborhoods had piped delivery of untreated on financial considerations. 39 The feasibility study for the condominial system was based on under half from the reduced use of materials, while in the case the same design parameters as the original (that is, 90% service of sewerage about one third of the savings would originate from coverage with a provision of 250 liters per capita per day) and the reduced excavation and the remaining two thirds from the compared the contract prices quoted for the original (unfinished) reduced use of materials. An additional advantage to the system project to equivalent expenses for a condominial design (Tables was that the cost savings enabled the financing of water services IV.1 and IV.2). This analysis predicted cost savings of approximately in the poorest neighborhood of the city (Bairro da Paz), which had 70% in the case condominial networks were used to supply water not been contemplated in the original project, and still achieve a (a unit cost of US$45 versus US$167), and approximately 40% cost reduction of 60% relative to the original scheme. Once these in the case of sewerage (a unit cost of US$56 versus US$94). It advantages had become apparent, the municipal council and also found that in the case of the water service, just over half of local community representatives approved the adoption of the the savings would result from the reduced excavation and just condominial model and the project was implemented. Table IV.1: Cost comparison for water systems Original conventional design Proposed condominial design Total cost Cost per connection Total cost Cost per connection Excavation 454,000 88 101,000 19 Pipes 407,000 79 129,000 25 Total 861,000 167 230,000 45 Table IV.2: Cost comparison for sewerage systems Original conventional design Proposed condominial design Total cost Cost per connection Total cost Cost per connection Excavation 263,000 39 186,000 28 Inspection chambers 181,000 27 85,000 13 Pipes 185,000 28 102,000 15 Total 629,000 94 373,000 56 Source for both tables: Condomimium, 1997 `Estudo de Aplicação do Modelo Condominial aos Projetos dos Sistemas de Abastecimento D´Água e Esgotamento Sanitário de Parauapebas-PA' 40 Figure IV.1: Original conventional water system design for Parauapebas 41 Figure IV.2: Original conventional sewerage system design for Parauapebas 42 Figure IV.3: Subsequent condominial water system design for Parauapebas 43 Figure IV.4: Subsequent condominial sewerage system design for Parauapebas NQ MC. 51º WGr. 44 Although the condominial model was applied to expand both authorities, who preferred the easier system operation associated the water and sewerage systems in Parauapebas, this case with a high-pressure network. study will focus on the study of the water system. The reason is that while there are other examples of condominial sewerage The resulting public distribution network covered 287 kilometers systems in Brazil (such as the preceding case studies of Brasilia of streets with a total length of just 43 kilometers. Given that a and Salvador), application of the model to the water network is conventional network would have been laid out along the full comparatively rare and thus of special interest. extension of the urban road system, this represents a reduction of 85% in the total length of water mains needed to complete the IV.3 Engineering Aspects public water supply network, resulting in major savings in terms of reduced excavation, breaking and resurfacing of sidewalks, fewer Following this decision-making process, the construction of the materials, as well as a more minor, briefer disruption of urban life in condominial water distribution system began. the execution of the works. The hydraulic capacity of this network complied with all Brazilian regulations applicable to networks of (a) Public network this kind. The same was true for the location of component parts and accessories, such as valves for cleansing outlets at the lowest By the time the construction of the condominial networks points of the network, as well as closing valves for interrupting the began in Parauapebas, the upstream construction associated water supply at predetermined points for maintenance purposes. with the town river-fed water supply reservoir was already complete. This sequencing of events prevented an integral More recently, the water distribution network has been further approach to the design of the water supply and distribution expanded to meet the demands of a growing population. In this systems. In effect, the prior decision to build a single large second round, it was possible to improve on the first-generation reservoir imposed a centralized structure on the design of the design, achieving additional cost savings as illustrated in Figure water distribution network, which was not necessarily the optimal one. For example, in some areas of the city that were not well Installation of micro-meter in household connection, placed with respect to the reservoir in terms of their relative Parauapebas gradients, the maintenance of prescribed pressure levels proved difficult. This is evident in the fact that the reservoir has a median level of 205 meters above sea level, while about half of the city is located below 165 meters above sea level. The problem could have been avoided by adopting a more decentralized design entailing a number of smaller reservoirs. Given that this possibility had already been ruled out, another alternative would have been to allow a lower pressure service in some areas of the city by introducing pressure-break valves or boxes, since this also would have helped to contain distribution losses caused by the highly-pressurized system. However, this solution was not acceptable to 45 IV.5. The new design involved an even lower pipe density areas, building densities, variation in consumption patterns, etc. than the first, with only 41 meters of distribution network per while maintaining the basic hydraulic requirements. In simple hectare, as opposed to 46. A key innovation was the use of a terms, given the new and different conditions associated with computer model of the water network that permitted automatic the introduction of this system, deficiencies in water supply determination of the accessories, valves and pipe diameters provision to one particular area could be immediately diagnosed necessary to modify the system operation and expansion in and corrective measures taken for restoring service standards response to events such as increases or decreases in occupied as necessary. Figure IV.5: Second-generation design of condominial water distribution network 46 (b) Condominial branches Intersection of Condominial Water Branch, Parauapebas The condominial branches for water were invariably routed along the sidewalks. This is in contrast to the sewerage networks described in the cases of Brasilia and Salvador, where branches were sometimes laid within lots. The main reason for this difference is the over-riding need in the case of the water service to ensure individual connections to each plot for subsequent metering of use. The condominial branches of the water network were placed at an average distance of one meter from the front of the plots and at a depth of approximately 0.40 meters. The branches were split at the double-tee junctions placed along the network at points coinciding with street intersections so that each was capable of supplying water to up to two condominial branches (that is, two urban housing blocks). Moreover, the design called for the installation of both block leading to a higher unit cost for the existing residents of the valves and macro-meters at the entrance to each condominium. condominium. It also raised the risk that the network capacity As discussed below, this feature of the design of condominial would have to be upgraded if plots were subsequently occupied water networks brings significant operational advantages in and urban density consequently increased. terms of facilitating leakage reduction initiatives and allowing operational problems to be isolated at the condominium level IV.4 Financial Aspects in order to reduce the impact of system maintenance on other users. There was no formal connection fee for the service; rather, all households were expected to purchase the materials required to Connections to each plot were made into the condominial complete the condominial branches and household connections, branches at the most convenient points in each case. The as well as to contribute their labor for the excavation of the diameter of the condominial branches was similar to those used trenches (see below for further details). in vertical buildings and conformed to the specificities of each case. A detailed survey of part of the network indicated that The fee system for the new condominial service was based approximately 55% of the branches used 32-millimeter pipes while on a series of studies conducted on the economic and financial the remaining 45% used 25 millimeter pipes. viability of the city's water supply and sewerage systems. In contrast to the sewerage systems considered above, residents One of the main problems encountered in the development of did not assume any maintenance responsibilities in the case the condominial branches was the relatively large number of of condominial water networks and hence there were no issues unoccupied plots. In cases where there were residents occupying of differentiation in service fees between conventional and the plots on either side, pipes had to be laid across empty plots, condominial systems. 47 The tariff structure consists of a fixed monthly fee equivalent to three decisions made at the outset of the process proved critical in US$2.78 and a social tariff of US$0.25 per cubic meter for the securing the commitment of local residents to this approach. first 10 cubic meters of monthly consumption. Residential consumption beyond the social tariff threshold, as well as all The first decision was to enter into a dialogue process with non-residential consumption, is charged at a rate US$0.99 per the community, whereby the rationale for the decision was cubic meter. In contrast with other parts of Brazil, a separate carefully explained and the rules governing the process publicly fee structure is applied for water and sewerage services. consulted. Residents were given exhaustive information about Nevertheless, the 38% of households that are not metered the condominial systems and the authorities' conclusion that its (because the metering program was discontinued for some of the adoption was essential for reaching universal access. Authorities new connections) are charged based on a monthly consumption made it clear that residents would have to pay no more for the of 10 cubic meters, making the adoption of metering financially development of the condominial branches than they would to unattractive to them. make a household connection to a conventional network. The consultation process was implemented by local grass IV.5 Social Aspects roots organizations, including churches, social clubs, trade unions, political parties and community associations. The adoption of the condominial model in Parauapebas was The second key decision was to develop workable rules for the motivated by the desire to reduce the project costs to make the implementation of the project, based on a fair and transparent goal of universal access feasible. Nevertheless, the price bid by division of labor between the community and the municipality. The the winning construction company for the development of the municipality agreed to assume all the more technically complicated condominial systems still proved to be prohibitively high, largely aspects of the works that would have been beyond the capability as a result of the cost associated with the development of the of the users to implement, such as determining the hydraulic condominial branches. In order to resolve this problem without capacity of the pipes as well as their installation. Moreover, all sacrificing the universal access goal, a proposal was made to urban water supply standards were met so that the community mobilize community labor for the construction of the condominial was assured of the company's obligation to keep the system in branches. In the original project design, the objective of the social good working order in return for payment of a reasonable fee. intermediation effort had been merely to educate the population regarding the proper use of the new water systems. Nevertheless, The third key decision was to implement a pilot project once an with this change of strategy, the community mobilization exercise initial level of support had been established. The pilot comprised took on the much more ambitious task of facilitating large-scale the planning and execution of the approach within a limited area community participation in the construction of the condominial of the city. This had an important demonstrative effect that helped branches. In that sense, the social component of the condominial to increase support for the project and provided a true test of system in Parauapebas was by far the most challenging of the acceptance. It also helped clarify the practical implications of the three cases considered here. construction process. The authorities were clearly concerned as to whether the The implementation period consisted of a number of distinct community would be willing to participate in this way. In retrospect, phases. The process began with the organization of at least one 48 percentage of household budgets. Once the condominium View of Condominial Water Branch, Parauapebas representative determined that all materials were ready, the route of the condominial branch and associated connecting pipes was marked out and the community excavated the trenches. The contractor then laid the pipes and registered new users. Finally, when the entire system was ready, the valve at the entrance to each condominium was opened, permitting the water to flow. Overall, the social mobilization process in Parauapebas involved 60,000 people organized into 800 condominiums. The provision of water service to the various condominiums did not reflect traditional political patronage patterns but merely the efficiency and efficacy with which each condominium organized itself and advanced through the various stages of the process. IV.6 Operational Aspects The operation of condominial water networks poses many of the same challenges as that of conventional networks. However, there are also a number of key operational differences between the two systems. First, the overall shorter length of the condominial network means less maintenance per meeting in each of the housing blocks, which were considered connection, and also reduces the volume of distribution losses condominiums for the purposes of the project. The meetings were that can take place in transit. Second, the architectural structure used as an opportunity for residents to clarify any doubts, discuss of the condominial network facilitates the introduction of the rules of access to the service and elect a condominium macro-metering and condominium-level meters, which both representative for the subsequent phases of the process. contribute to reducing leakage and allow smaller areas of the network to be isolated to minimize the disruption of service On the basis of the consensus reached at these meetings, all when repairs are made to the system. Third, the network design members of each condominium signed the terms of agreement reduces the number of easily accessible points to the network signaling their willingness to fulfill the commitments required to and pipes, such as junctions and others, thereby reducing initiate the construction phase. Contractors then analyzed the losses and discouraging illegal connections. hydraulic capacity of each condominial branch and made a list of the materials needed for construction. The community was In the case of Parauapebas, maintenance data are available responsible for purchasing these materials. This was undoubtedly on the number of breakages on the water distribution system the most delicate stage in the entire process, particularly in poorer during 2002 and 2003 (Table IV.3). This information indicates neighborhoods where such expenditures constituted a significant that breakages on the public network accounted for little 49 Table IV.3: Overview of breakages on water distribution system 2002 2003 Total number Percentage total Total number Percentage total 25 mm 180 24.2% 198 18.5% Condominial Branches 32 mm 174 23.4% 252 23.5% Condominial meter 337 45.3% 545 50.9% Sub-total 691 92.9% 995 92.9% 60 mm 19 2.6% 17 1.6% 75 mm 3 0.4% 9 0.8% 150 mm 3 0.4% 4 0.4% Public Network 250 mm 1 0.1% - 300 mm - 1 0.1% 600 mm 2 0.3% 2 0.2% Sub-total 28 3.8% 33 3.1% Household pipes 13 1.7% 27 2.5% Household connections Condominial valves 12 1.6% 16 1.5% Sub-total 25 3.4% 43 4.0% Total 744 100.0% 1071 100.0% Source: Condominial Operating Systems more than 3% of breakages on the system, and in absolute In order to measure the impact of maintenance activities on the terms amounted to no more than 0.7 maintenance incidents continuity of water supply in the city, a Supply Interruption Index per kilometer. More than half of these breakages occurred on was developed, which expresses the aggregate duration of the narrowest diameter (60 millimeter) segment of the public interruptions for all customers as a percentage of the maximum network. By contrast, approximately 93% of the breakages total hours of supply that would be possible under a continuous on the system were concentrated in the condominial 24-hour service. The Supply Interruption Index for 2001 oscillated branches. Almost half of these were related to the meter between 1% and 3% monthly. The higher values of the index serving the condominium. coincided with a major road asphalting program in the town, 50 Table IV.3: Overview of coverage, metering and distribution losses % 1998 1999 2000 2001 2002 2003 2004 Household coverage 18 56 74 74 79 82 82 Meter coverage 46 72 77 70 60 64 62 Distribution losses 56 44 43 51 62 65 66 Source: Condominium Operating Systems during which time the heavy machinery used caused significant With regard to metering, condominial systems allow for pipe breakages. For the rest of the year, the index averaged no conventional individual household meters and facilitate macro- more than 0.6%. Operational data from 2004 suggest that the metering as a result of the block structure of the network design. typical duration of a supply interruption was 15 to 30 minutes, The original design of the water system in Parauapebas envisaged while the number of affected connections was approximately 25, extensive macro- and micro-metering. However, this was not fully which is to say a single condominium. The very small scale of the implemented in practice. On the one hand, the authorities proved disruption illustrates the convenience of the valves that isolate unwilling to finance the modest budget associated with the macro- supply to each individual condominium. metering program (estimated at US$10,000). On the other hand, the population was not always willing to accept individual metering. As a result, meter coverage, which peaked at 77% in 2000, has Installation of the Public Water Network, Parauapebas subsequently declined to 62% of the total as overall coverage expanded (Table IV.3). For a small sample of the city (comprising 16 blocks), detailed information is available on household meter coverage and average monthly household consumption (as measured by the macro- meter at the entrance to each condominium). Analysis of this data indicates that the correlation coefficient between household meter coverage and average monthly household consumption is no more than 0.13, indicating a relatively weak effect of metering on household consumption decisions. Interestingly, distribution losses, which improved substantially from 56% in 1998 to 43% in 2000, subsequently rose to 66% in 2004 (Table IV.3). There are two likely explanations for this. The first is the 51 decline in meter coverage since 2000, which makes it more difficult to detect and eliminate leaks. The second is the growth in illegal Repaired Condominial Branch, Parauapebas connections after 2000, as unserved households in peripheral areas of the city began to tap into the system. As noted in Table IV.3, the water coverage rate in the city increased fourfold over the period 1998 to 2004, increasing from 18% to 82%. Nevertheless, these official figures do not necessarily capture the real situation on the ground. This is because clandestine connections are a major problem in Parauapebas. These have taken one of two forms. First, there is the issue of persistent non-payment by customers, leading to eventual disconnection by the utility company and often, the illegal reconnection to the system. This problem has been exacerbated by the lack of a clear legal framework on non-payment, as well as police reluctance to enforce existing regulations. Second, there is the issue of the unserved peripheries of the city. Those located close to the existing network have made clandestine connections to ensure access to the service. This problem can only be adequately resolved by IV.7 Institutional Aspects expanding the network to those areas. Prior to the development of the condominial system in Parauapebas, the limited service available had been provided Installed household meter, Parauapebas directly by CVRD to its employees. Following the expansion of the system, responsibility was transferred to the local municipality, which had no prior experience with the construction and operation of water and sewerage systems. An outside engineering firm was hired to build the system. However, there was considerable debate and indecision about the appropriate institutional model for the subsequent operation and maintenance of the system. The initial proposal was to delegate this function to a private contractor, but subsequently it was decided to create a Municipal Water and Sewerage Service within the local authority to assume responsibility for this function. However, when this proved to be problematic, management was eventually contracted out to a private operator. The Municipal Water and Sewerage Service continued to be responsible for regulating and overseeing the work of the contractor. 52 Considerable care was taken to develop an appropriate regulatory Installation of the Public Water Network, Parauapebas framework for the service, which was introduced as a Municipal Law following extensive debate and unanimous endorsement by the municipal council. This law incorporated principles for determining and updating water and sewerage fees, as well as service quality provisions that set a maximum of 5% for the Supply Interruption Index. Because the municipality has a weak institutional capacity, the regulatory framework has not always been adequately enforced and upheld, particularly with regard to tariff provisions and sanctions for non-payment of service. IV.8 Summary The case of Parauapebas is particularly noteworthy for a number of reasons. First, it is one of very few cases in which the condominial approach has been applied to water (as opposed to sewerage) networks. Second, it illustrates the challenges involved in constructing a city-wide system virtually from scratch and achieving a fourfold increase in coverage in the space of six years. Third, it is one of the most ambitious examples of community mobilization with 100% of residents both financing the materials for the condominial branches and contributing the labor to excavate the trenches. Fourth, it provides some insight into the potential operational advantages of condominial designs for water systems, even if these have not been fully realized in the case of Parauapebas. 53 Excavation works to install Public Network, Brasilia V. Conclusions This study described and compared the experience of the cost of connecting a household to the sewerage system in implementation of the condominial model in three Brazilian cities: Brasilia and Salvador was approximately US$300. The main Brasilia (the nation's capital); Salvador (a major, historic metropolis reasons behind this low unit cost are the reduced length of the in northeastern Brazil); and Parauapebas (a small but fast-growing public network (representing only 25% to 35% of the total length boom town). Together, the projects analyzed have succeeded of the system), reduced diameter (with a substantial proportion in providing condominial networks to 2.5 million urban residents of the sewer networks rendered in pipes measuring 100 to 150 in Brazil, with 214,000 connections organized in close to 16,000 millimeters in diameter), and reduced depth (with public networks condominiums (Table V.1). laid at a maximum of one meter and condominial branches laid at approximately one-half meter). The case studies differ in a number of ways. For instance, Parauapebas is the only case in which the condominial system The main lessons and conclusions from the case studies as a was applied to the water network. In Brasilia and Parauapebas, whole can be summarized as follows. the model was adopted in a conscious, planned manner, whereas in Salvador it was adopted gradually and opportunistically First, the implementation of condominial systems for both water in response to major engineering challenges. In Brasilia and and sewerage has proved feasible on a large scale in major urban Parauapebas, the implementation of the model was facilitated areas of Brazil, and at relatively modest cost. Moreover, such by relatively orderly urbanization, whereas in Salvador the systems systems have demonstrated their versatility in the most challenging had to be adapted to the demands of particularly dense, steep peri-urban environments, where conventional systems are simply and precarious urban slums. In Brasilia and Parauapebas, local not an option. residents covered the full cost of the condominial branches, whereas in Salvador these were constructed free-of-charge. Second, the large-scale social mobilization demanded by the In Brasilia and Salvador, the system was almost entirely condominial approach has also proven feasible in the Brazilian constructed by contractors, whereas in Parauapebas the context, and there have been no major difficulties for residents condominial branches were installed almost entirely by community to reach consensus about system design issues at the level of members. In Brasilia, community maintenance of condominial each condominium. The intensity of social mobilization efforts branches has not posed serious problems, whereas in Salvador needs to be considerably greater in cases where the community it does not seem to have functioned very effectively. These has chosen or is expected to contribute labor to the construction contrasts illustrate the considerable variety of forms that the of the system. However, even in other cases, it is important for condominial model can take. community mobilization efforts to include educational messages relating to proper system use and maintenance, the motivation for Notwithstanding, comparing some of the key parameters from and process of connecting to the network once available and the the case studies reveals basic underlying similarities. The average public health benefits associated with using the service. 55 Table V.1: Comparative overview of case studies Brasilia Salvador Parauapebas Overview · Years 1993/2001 1995/2000 1998/2004 · Service Sewerage Sewerage Water Scale · Population 700,000 1,730,000 60,000 · Households 188,000 214,000 11,000 · Condominiums 5,000 10,000 800 Distribution by type of system · Backyard 6% Na 0% · Front yard 43% Na 0% · Sidewalk 51% Na 100% Engineering Network length per connection (m) · Public network 3.30 2.00 3.80 · Condominial branch 7.20 5.75 10.00 Depth of trenches (m) · Public network 0.50 to 1.20 Na Na · Condominial branch 0.30 to 0.60 Na 0.40 Network pipe diameters · 100mm 56% 0% - · 150mm 29% Na - · 200mm 12% Na - · 250mm+ 3% Na - Branch pipe diameters · 100mm 100% Up to 20 households - · 150mm 0% 21 to 30 households - · 200mm 0% Over 30 households - · 250mm+ 0% - Financial Cost per connection (US$) 340 290 45 Cost per meter of network (US$) · Public network Na 52 Na · Condominial branch Na 21 Na Tariff discount for connection Backyard (45%) No connection fee Na Tariff discount for maintenance Yes (40%) Yes (44%) Na Social Participation in meetings 8% Na Na Self-construction of branches 1% 0% 100% Source: Table prepared by the author. 56 Third, the available evidence does not suggest that the operation genuinely assume maintenance responsibilities. and maintenance problems posed by condominial systems Fifth, public acceptance of the condominial approach is greatly for the utility company are significantly different or more severe aided by a decisive and coherent policy on the choice of than those posed by conventional systems. Relying on the technology for network expansion, which is communicated to the community to assume responsibility for maintenance has not public in a clear, comprehensible way. This was the case in Brasilia worked particularly well in Salvador, but has apparently been and Parauapebas, where the application of the same technology more successful in Brasilia. Community maintenance of networks in a non-discriminatory fashion throughout the service area greatly is an optional feature of condominial systems and as long as contributed to the credibility of the approach. local conditions allow for pipes to be routed along sidewalks, the system is perfectly compatible with traditional utility maintenance Finally, the condominial model also has implications for the design approaches. of upstream drinking water treatment and downstream sewage treatment facilities. Namely, it suggests that it may be more Fourth, it is important to ensure a fair fee structure for both feasible to decentralize these facilities in an effort to avoid the high connection to and use of condominial systems so that residents transportation costs associated with concentrating large volumes benefit from any cost savings associated with the choices they of fluids at single geographical points to feed large-scale plants. make. This entails lower connection costs for systems routed Nevertheless, this aspect of the condominial model was not widely through backyards and lower fees for cases in which residents applied in the cases considered. Excavation for the Public Network, Brasilia 57 Water and Sanitation Program Latin America Alvarez Calderón 185, San Isidro, Lima 27, Perú Phone: (511) 615-0685 Fax: (511) 615-0689 E-mail: wspandean@worldbank.org Website: http://www.wsp.org The World Bank 1818 H Street, NW Washington, DC 20433, USA Website: www.worldbank.org The World Bank