Oswaldo Sevá
Saturday, April 30, 2005

Alerts regarding the consequences of hydroelectric projects planned for the Xingu River, Brazilian Amazon

Publication of the panel of experts on the Xingu hydroelectric complex, May 2005

Executive Summary

1. Summary of the plans for hydroelectric dams on the Xingu River

An obsession of global engineering is the fixed idea to dam all rivers that have significant hydroelectric potential. Within this vision, rivers are no longer viewed as living flows of water, bio–chemical media which maintain life in equilibrium, including the lives of riverbank dwellers and migratory animals. Instead, rivers are examined using calculators, as if they only existed to be blocked by a giant wall to have their energy exploited.

The Xingu River, as the proponents of hydroelectric dams would say, is "a good potential". However, one of its most important characteristics, certainly known to its indigenous and riverbank dwelling inhabitants, is that its flow of water is extremely variable over the course of a year, and even over a period of weeks or days.

The river fills quickly and in great volume, since its basin is very large. The volume of water drained per kilometer is larger than many other important rivers in Brazil, more than 17 liters per second per square kilometer. The Araguaia/Tocantins basin registers 14–16 l/s per km2, and the Paraná 11 l/s.

But, the Xingu is also a river that dries quickly, remaining at low flow levels for about four months out of the year. At the city of Altamira, in Pará state, on the lower Xingu.

  • Low flow periods average less than 1,000 cubic meters of water per second
  • Minimum flow periods are 450 – 500 m3 /sec. in September and October
  • Average high flow periods are above 25,000 m3 / sec.
  • Peaks of high flow are registered above 30,000 m3 / sec.

Knowing the flow volume of the river, we can gauge the mechanical potential of the river which can exploited by building hydroelectric dams, and whose equation will also depend on the vertical drop (hydraulic head). According to the document "Hydroelectric Inventory Studies of the Xingu River Basin", by the consulting firm CNEC – Camargo Corrêa in 1980, the "best" alternative for the complete hydroelectric exploitation of the Xingu basin (alternative A of the studies carried out) would be that between 281 meters above sea level, in the upper Xingu basin, near the BR 80 highway (either in the Kapoto–Jarina indigenous reserve, or in the northern part of the Xingu Indigenous Park) and an altitude of 6 meters above sea level, just below the village of Belo Monte do Pontal and just above Vitória do Xingu, Pará:

  • five dams would be built on the Xingu (at the points denominated Jarina, Kokraimoro, Ipixuna, Babaquara and Kararaô) and one dam on the Iriri river, the Xingu’s largest tributary on its left bank (at a site called Cachoeira Seca).

The reservoirs of these six hypothetical dams would flood islands and forests, many according to the CNEC study still virgin growth, drowning an area of nearly 20,000 square kilometers, the equivalent of half the areas flooded by all dams built in Brazil until now. A good part of these two million hectares of land are riverine ecosystems, and included are a significant group of indigenous territories, many officially protected and others still not officially recognized.

The Babaquara dam, alone, could flood more than 6,500 km2, creating Brazil’s largest reservoir and the world’s second largest (Akosombo reservoir, on the lower–middle Volta river, floods more than 8,000 km2, cutting the country of Ghana in western Africa in half). Brazil ’s largest reservoir is Sobradinho, on the São Francisco river in Bahia state, which floods 4,200 km2, and its second largest is Tucuruí, on the Tocantins river, with 2,860 km2.

But the combined impact of these projects would go beyond the area flooded. The destructive consequences of the projects would also result from new roads built for dam construction, the cutting of swathes for transmission lines, and clearing for construction sites.

The map produced by the geo–processing laboratory of the Socioambiental Institute (ISA), included in this volume, shows the numerous interferences and overlapping territories which imply a diverse series of consequences, including a complex mosaic consisting of:

  • extensive federal and state public lands, which are being licensed for use by private interests, or which are the target of claim jumpers and professional land invaders;
  • protected areas such as Biological Reserves, and areas destined for sustainable management, such as National Forests;
  • reserves set aside for artisanal mining, and other areas which have been invaded by miners, and
  • immense areas with a great number of authorizations for prospecting and extracting mineral deposits.

The loss of extensive land areas, islands, rich forests and farm lands, and territorial modification caused by relocation of roads and ports will affect claims on these lands. Structures, plantations, and other "improvements" by natives, or immigrant settlers on public and private lands, as well as in forested areas will also give rise to claims for compensation.

In Chapter 1, written by professor Oswaldo Sevá, we see some of the characteristics of each stretch of the Xingu valley threatened by the potential consequences of each of the six dams being planned. We register the focal points of the conflicts which characterize the recent occupation of the region by non–indigenous residents, and by economic activities (such as soy cultivation, cattle ranching, timber extraction, and gold mining) in this region where until recently, only indigenous peoples lived.

The initial proposal to dam the Xingu River gave rise to a strong opposition by indigenous peoples and a broad group of environmentalists and social movements. The actions by indigenous leaders, including international trips and meetings with NGOs and international financial institutions culminated in the "Encounter of Indigenous Peoples in Altamira" in February, 1989, which had a great impact, burying for a period the Kararaô dam project, the first part of Eletronorte’s plans for the complete exploitation of the hydroelectric potential of the Xingu River valley.

2. The second frustrated attempt to dam the Xingu River

Until 1999, the company was, generally in a very discrete manner, intensifying its plans for the implantation of the project. Relevant geographic and technical modifications were made, and the project was re–baptized for the second time, this time as the Belo Monte Hydroelectric Complex, referring only to the engineering works for the first dam on the Big Bend of the lower Xingu. Eletronorte argued that Belo Monte would be feasible even as an "isolated" dam on the Xingu.

Around 1999, Eletronorte, defeated ten years before, re–organized its efforts. It became active as a political actor in Altamira, on the Transamazônica, a fact registered in the testimony and information presented in this book by local leaders Antonia Melo and Tarcísio Feitosa da Silva.

Still, Eletronorte suffered as a result of continuing errors in the handling of problems and required actions at Tucuruí, its emblematic example of how not to build a mega–dam in the Amazon. Much has been written during the two decades since Tucuruí was built regarding problems affecting the reservoir area and downstream ecosystems and populations. The social impacts of this hydroelectric investment continued to gain notice and have wider implications, both because new evidence of the scope of the disaster continued to be produced, as well as because of the fact that the project’s so–called second phase, involving construction of another powerhouse raised the reservoir, increasing the flooded area by more than 400 km2. Social movements of dam–affected communities continued to criticize the project’s effects. Anthropologist Sonia Magalhães explains in her chapter in this book, based upon her long accompaniment of the situation at the ground level, how the social dynamic and the political life of the country and region help determine the dimension of the social effects of large dams.

At various times during the period 2001–2004, the presidents of Eletronorte referred to plans for the building of a second dam just upstream from Belo Monte, before called Babaquara and now called Altamira.1 The Mines and Energy Ministry, in its public presentations on plans for expansion of the electric sector in the Amazon region, shows Altamira along with Belo Monte.2

With plans for damming the Xingu advancing, the destiny of the energy that would be generated by Belo Monte continued to be unclear, above all because of the intense criticism of the case of Tucuruí which was built to provide subsidized energy for the aluminum plants which had been built near Belém and in São Luís, and also due to the losses suffered by Brazil as a result of the renegotiation of these energy contracts.

In February and March, 2001, following a rather dry summer, it became clear that the Southeastern – Central Western region and the Northeastern region of Brazil ’s interconnected electrical distribution network was at risk due to low water levels in many of the largest hydroelectric reservoirs in the Paraná and São Francisco basins. A crisis in electricity supply followed, worsened by weak interconnections between regions. It was at this time that dam proponents re–floated Belo Monte as Brazil’s "salvation", and demanded publicly that obstacles be removed so that construction of the project could get underway as soon as possible.

These maneuvers to hide the motives behind the dam’s construction, to create rational arguments to support the strategy of making Belo Monte a "done deal", to shuffle or camouflage alternatives, are thoroughly analyzed in the chapters of this book written by the journalist Lúcio Flávio Pinto, who details the manipulation of economic – financial numbers, and clearly exposes the grandeur and misery of the state of Pará – exporter of minerals and energy and center of poverty in the Amazon; and by engineer and anthropologist Diana Antonaz, who interviews prominent intellectual figures in the "electric" and "petroleum" spheres who today occupy central positions in the government’s energy sector, analyzing their discourse and logic. She finds that the notion of development which these technical experts defend is aimed at benefiting an abstract population, a mass without identities or cultures, which fails to consider the concrete needs of people with flesh and blood.

At the end of 2000, Eletronorte signed a contract with a foundation called FADESP, linked to the Federal University of Pará state, through which teams of researchers were formed to carry out an Environmental Impact Study. The conditions of this contract procedure, carried out without competitive bidding, and the attempt to obtain an environmental license for Belo Monte at the state, rather than at the federal level, led the Federal Public Ministry (Attorney General’s office) to file suit. Also questioned were Eletronorte’s failure to consult with indigenous populations who would be affected by the project and the need for a special act of Congress to authorize such studies in projects affecting indigenous peoples, both constitutional requirements (article 231). Eletronorte had attempted to dodge these requirements when it redesigned the project, siting the principal dam on the da Serra and Pimentel islands, some 50 km upstream from the prior location, which had been downstream from the Jericoá rapids. Eletronorte argued that the indigenous populations would no longer be affected by the area flooded by the dam. While it is true that the Paquissamba indigenous reserve would no longer be submerged, the reserve would be seriously affected by the drying of the river bed, since flows downstream from the dam would be restricted to a small percentage of the historic average.

The number of indigenous people in the region who would be directly affected by Belo Monte dam is far greater than the 50 or so Juruna residents of the Paquissamba reserve. The Indigenist Missionary Council (CIMI) has documented more than 400 indigenous people of the Xipaia, Kuruaia, Arara, Juruna and Kaiapó who live downstream from the Pimental island, and whose survival would be left high and dry by the damming of the Xingu. There is much contact by riverine communities with Altamira, and the drying out of the river downstream would mean these communities would also have difficulties in traveling to the city.

Eletronorte admitted as much (and then later denied it) when they chose the Kararaô alternative in 1988 over another alternative being studied (KararaoIII/KoatinemaII) which was very similar to the current plan for Belo Monte. They said it would "cause indirect impacts of greater proportions, due to the interruption of the flow of water through the Big Bend of the Xingu, which would interfere with aquatic and riverbank ecosystems and the riverine and indigenous populations living there..." and admitted there were "344 (indigenous) individuals who would be directly affected" (Usina Hidrelétrica Kararaô, Efeitos e Programas Ambientais: Síntese, Eletronorte/CNEC, Oct 1988)

Such facts and a discussion of who and how many would be affected, as well as details of their ways of life are researched and related by anthropologist Antonio Carlos Magalhães, and by geographer Reinaldo Corrêa Costa, in two other chapters of this book.

The legal decisions in 2001 and 2002, resulting from the lawsuit by the Public Ministry resulted in the suspension of the EIA and the licensing process. These topics are discussed in detail in the chapters written by lawyer Raul Silva Telles do Valle, of the Socioambiental Institute in Brasília, and by the Federal Prosecutor in Belém, Felício Pontes Jr. with anthropologist Jane Beltrão of the Federal University of Pará.

3. Summary of the Dimensions of the Belo Monte project, 2004 version

  • The total installed capacity in the second version of the project, which was disseminated from 1998 until mid–2003 was 11,182 Megawatts, of which 11,000 MW at the principal powerhouse (Belo Monte) and 182 MW would be in a complementary powerhouse located in the principal dam at Pimental island. This is the same installed capacity as the prior version of the project, which in 1988 was called Kararaô, and is greater than the capacity of 8,400 MW indicated in the Xingu Hydroelectric Inventory (CNEC, Eletronorte, 1980). In 2003, Eletronorte began stating publicly that it was analyzing alternatives with only 10 turbines of 550 MW each in what it termed an "initial stage" of the project, for a total installed capacity of 5,500 MW, or another configuration with 14 turbines totaling 7,700 MW of installed capacity.

  • The range of variations in the flow of the Xingu River is extremely great, and the two reservoirs being planned for Belo Monte would have limited capacity to store water. The powerhouse would operate near or at its maximum capacity only during three months annually in the best of hypotheses (in March, April, May when the average flow exceeds the maximum capacity of the turbines of 13,900 m3/s). The feasibility study delivered to the National Electrical Energy Agency (ANEEL) indicates "firm energy" of 4,700 MW (a plant factor of 42%, near that of many other hydroelectric dams in Brazil), implying that the dam would generate at least this amount of energy permanently, even in critical low–water periods.

  • The total areas to be flooded for reservoirs and for the canals carrying water to the powerhouse would be between 440 km2 and 590 km2 – a fourth of this area would be in the reservoir formed on land drained by four creeks, and three–fourths would be along the Xingu River itself. The previous project projected a reservoir of 1100 km2 in area.

  • The project is designed without ability to regulate the flow of the river. The water storage capacity of the reservoir would be about 3.8 billion m3 of water (by comparison, Tucuruí has storage capacity of 50 billion m3, more than 13 times Belo Monte’s capacity, and Balbina 17.5 billion m3, or 4.6 times that of Belo Monte. The average depth of the two reservoirs and the canal systems would be between 6 – 8 meters (figure 2.3–1 of the Feasibility study). The flow of the Xingu would be diverted into three canals and a shallow reservoir with little accumulated volume. The turbines would essentially receive a volume of water comparable to that diverted at the initial dam, meaning that in electrical engineering jargon, the dam would be "run of river". It is important to note that, even if the project was constructed in two phases, and even if only one canal was constructed initially (that of the Maria e Gaioso Creek), with only half of the powerhouse installed, this would not diminish the project’s environmental or social impacts, and the three massive structures projected for the project would still need to be built: 1) the dam and floodgates on Pimental island, diverting the flow from the Big Bend and housing the complementary powerhouse of 182 MW 2) the structure housing the complementary floodgates below the Jericoá rapids, on the Xingu’s left bank; 3) the final wall of the second dam which would house the principal powerhouse, and the release for the turbinated waters near the Santo Antonio Creek.

4. Simulation of the hydraulic potential of the Xingu River, if the dams had been in operation since 1931

Methodology: The simulation was carried out using the model Hydrolab (Cicogna and Soares Filho., 2003, FEEC, Unicamp) using data from SIPOT (System of Information of Hydroelectric Potential, of Eletrobrás) on the water flow of the Xingu river in Altamira, measured on–site or estimated from 1931 to 1996. We also noted results from the period 1949 to 1956, considered that of highest rainfall benefiting hydroelectric generating systems on Brazilian rivers in the Southern hemisphere.

The model does not attempt to estimate generation based upon installed capacity, nor in terms of what the dams would be able to generate in the future, but rather to calculate how they would have functioned in the past, if they had been constructed on the points of these rivers where the river flows were measured.

The results of these simulations are noted for three types of hypothetical situations:

      if only one hypothetical dam, Belo Monte, had functioned generating electrical energy to the national interconnected system between 1931 – 1996

The maximum secure energy generation would have been 1,356 MW
(that is, that if during that period, the demand for energy from Belo Monte had exceeded 1,356 MW, the flow available to the turbines would not have been sufficient to guarantee this generation, and energy demand would have had to be attended by some other energy source on the same system)

      if only two hypothetical dams, Belo Monte and Babaquara functioned together during the period 1931–1996

The maximum secure energy generation of the two dams would have been 7,950 MW

Dividing the generation of the two dams, supposing the total exploitation of water by both (without releases of water which could be turbined), we would have:

  • 31% of the energy generated by Babaquara 3,078 MW
  • 69% of the energy generated by Belo Monte 4,872 MW

By comparison, projections of the total installed capacity of the two dams were 17,772 MW with the most recent version of Belo Monte and with a complementary powerhouse or 12, 090 MW in the version announced in October 2003 with half of the installed capacity in Belo Monte’s principal powerhouse.

The evident conclusion is that only with the two hypothetical dams, Belo Monte and Babaquara (Altamira) functioning, does the operational and economic situation become acceptable, because for an installed capacity of 17,772 MW or 12,090 MW, the minimum assured would be nearly 8,000 MW. By its original design, Babaquara would have an enormous reservoir, projected to flood an area of more than 6,000 km2., a situation which would cause enormous social and environmental impacts.

    if the six hypothetical dams had functioned together during the period 1931–1996

(Jarina, Kokraimoro, Ipixuna, Iriri + Babaquara and Belo Monte)

The minimum secure generation in the six dams would have been 12,806 MW

By comparison, the total installed capacity of the six dams, according to the guidelines of the "Integral Hydroelectric Exploitation" of the Xingu, (IHX, CNEC, Eletronorte, 1980) and registered on the SIPOT database would be:

  1. Jarina 620 MW
  2. Kokraimoro 1,490 MW
  3. Ipixuna 1,900 MW
  4. Iriri 770 MW
  5. Babaquara 6,590 MW
  6. Belo Monte* 11,000 MW

or alternatively: * in the reduced version of the project announced in Oct. 2003 5.500 MW

Total capacity originally planned for installation: 22,370 MW

Or with Belo Monte in its reduced version: 16,870 MW

5. Possible markets for Belo Monte’s electricity and the (ir)rationality of electrical planning

The destination of the electricity that would be generated is unclear, nor are there firm commitments for the energy it would generate. Logically, there are only two possibilities:

  1. to meet the demand of other regions and/or

  2. to meet demands in the Amazon region; where energy flows could be allotted towards

    • Attending conventional urban and rural markets in the region and/or
    • Attending electro–intensive consumers already in operation and/or
    • Attending new electro–intensive industries to be installed in the region.

Arguments and promises to meet demand in the Center–Southern or Northeastern Brazil using energy generated by Belo Monte are frequently made in the suspended EIA, in the feasibility study presented to ANEEL, and in the discourse of many economic authorities and the electric sector, both in prior governments as well within the Lula government.

To supply the grid reliable energy year–round, using only additional energy from Belo Monte is difficult, perhaps impossible to justify. In our simulation of past conditions, the power plant would have furnished little more than 1,300 MW during the most favorable dry months over the past seven decades.

However, in months with greater stream flow, and in more favorable years, Belo Monte would be able to send excess seasonal energy to the Northeast or the Central–Southern region, but this would depend on how much energy Tucuruí was generating, and on the operational transmission capacity of the North–South I and II lines. As to Belo Monte working as a complement in terms of its seasonality with dams on rivers in the Southeast and Northeast, this would be a technical feat, because there would only be an offset of 40 – 50 days between the peak flow in the Paraná basin, for example, at the end of January or beginning of February, and that of the Xingu in March or April.

The transmission lines required to carry this energy from the Xingu to the North–South trunk line would be very expensive and would increase the cost of generation by 60–70%; the cost of these transmission lines alone was estimated in 2001 to raise the cost of energy by $12/Mwh. In the chapter of this book written by Andre Saraiva de Paula, the imprecision of cost estimates is summarized as being on the order of several billion dollars, according to which source of information you believe, regarding potential investments in the transmission system associated with Belo Monte dam.

The company could apparently lower these costs, since in the future additional transmission corridors are projected to reinforce the North–South connection and its integration with the hypothetical transmission system originating in the Big Bend of the Xingu. But, this would be a mere maneuver since these expenditures all would have the same final objective, as described below.

Conventional local markets in Pará, Maranhão, and Tocantins states are already well–supplied with energy, with no motive for crises or deficits. Even if programs of rural electrification are carried out, given the population and economic dimension of these states, the additional consumption would be modest. In fact, rural communities and those isolated in the Amazon forest, or along river banks stand a greater chance to be attended by electricity obtained through the use of photovoltaic panels, micro–hydroelectric dams, or moto–generators burning diesel oil and eventually biodiesel.

Meanwhile, large clients (metallurgical industries and the CVRD mining company) are for now guaranteed of an energy supply through the second phase of Tucuruí, now nearly completed, via contracts (signed initially in 1984) which were recently revised or substituted.

The only other option which could explain the decision to construct and install a hydroelectric complex of this size on the Xingu – besides regional energy interchange – is that the additional electricity to be dispatched by Belo Monte would make new mining or metallurgical activities in the region, or the future expansion of current operations feasible.

For example, an additional 1,000 – 2,000 MW guaranteed would be a good reinforcement in transmission for Vila do Conde, Pará, and for São Luís, Maranhão, where aluminum foundries are located, or perhaps for Açailândia, Maranhão (pigiron plants) or to expand Companhia Vale do Rio Doce’s Carajás iron ore and manganese mines and its new plant for concentrating copper, inaugurated in 2004 by President Lula and the Director President of CVRD, Roger Agnelli.

Certainly, even the pig iron and other metallurgical plants could also be expanded, and new plants could be built in the region, including the often cited steel works in São Luís. The U.S. company Alcoa is evaluating the implantation of a bauxite mine and alumina refinery in Juriti Velho, in the region of Santarém, Pará, and has publicly expressed its interest in becoming a partner in the Belo Monte mega–project.

This "option" for the consumption of future electricity from the Xingu in mining and metals processing appears officially as one of several alternatives, always lumped together in the basket of options. However, plans for future transmission lines in the region published between 1999–2002 make this option clearer.

However, for the project proponents and the federal government, a repetition of Tucuruí, which was built to power aluminum foundries is not politically attractive, and few are willing to publicly admit that this high–cost and high–impact electricity would be exclusively or principally to make export–oriented mining and metallurgical plants feasible.

6. Summary of the local consequences of the hypothetical works for Belo Monte dam

The territories that would mobilized for this complex of engineering works, and others that would be directly affected by the flooding of the reservoir and by radical changes in local conditions include:

A large land area of the Big Bend of the Xingu between the river and the Transamazon highway, in an area of colonization projects and ranches between Altamira and the Belo Monte do Pontal ferry landing in addition to river bank lands and islands along the Xingu for 200 km, in two totally distinct stretches.

  1. along the first stretch of 80–90 km in extension, river banks and adjacent lands, and islands would be covered by water to the level of 97 meters above sea level (some documents say the level would be 98 m) forming the "river bed" reservoir. The land near various igarapés (creeks) which flow into the Xingu River would be flooded, and in the city of Altamira, low–lying areas along the Ambé, Altamira, and Panelas igarapés would be affected (details to follow);

  2. and the second stretch, 110 km long through the Big Bend of the Xingu downstream to the tailrace, where the water which passes through the plant’s turbines will re–encounter the Xingu, the natural riverbed will permanently be reduced to a flow well under the historical minima (details to follow);

More than 2,000 families from the urban low–lying areas will be forced to move, in addition to 800 residents of the rural area and 400 families of riverbank dwellers.3

In all, the total would be 3,200 families, approximately 16,000 people, most of whom have very little reliable information about the project and the consequences in terms of their expulsion from their homes and lands. In Eletronorte’s version, all are currently miserable, living very poorly, without minimum public services, and therefore would be much better off after receiving compensation or in the new communities that the company would generously offer them.

Robert Goodland’s article presents an international standard for analyzing the impacts of large dams and the recognition of the rights of dam–affected populations, which could indicate more adequate procedures for the planning of large dams in the future. The use of strategic environmental evaluation would make it more possible to evaluate the relative impacts and benefits of various options for regional development projects. The recognition of the right to prior informed consent is perhaps the only known way of guaranteeing that those affected by projects of the electric sector can participate, and be active in the determination of their own future.

Resuming the hypothetical impacts of the Belo Monte dam in Altamira :

Project maps permit us to deduce that Altamira will be cut by the stagnant waters of igarapés which will be dammed along their descent toward the Xingu, along its left bank:

Igarapé Ambé. The work areas and ovens of ceramic workers and the area of the São Francisco bathing area, along the highway bridge linking the city to the Transamazon. Various residences will be affected, and perhaps the highway itself will have to be raised, and a new bridge built. Also affected will be a saw mill, and a neighborhood of fishermen and horsecart transporters.

Igarapé Altamira. The homes built on stilts will be flooded, and a series of streets and bridges will be flooded. The São Sebastião neighborhood where Xipaia and Arara indigenous families, as well as non–indigenous residents live, will also be affected by the flooding.

Igarapé Panelas. Lands and ovens of ceramic workers will be flooded, and possibly also stretches of the road which links the city with the airport, including the bridge. The Pajé Beach, along with its archaeological site indicating the ancient presence of indigenous peoples, will be lost.

Main river bank walkway. The dammed water will rise to the wall on João Pessoa avenue, two meters below the walkway. All river bank residences from the BIS to the Xingu Club will have to be removed, and the "ports" for river boats and speed boats will have to be moved, for example on the "Seis" ramp where various houses will be below the 97 meter level.

Rainwater drainage pipes emptying into the river, as well perhaps as some clandestine sewer outlets will be affected. The Arapujá island in front of Altamira will be nearly totally submerged, with only the tallest trees still appearing.

Probable effects near the city. Beaches will disappear or be greatly reduced in size; islands will have most of their surface area reduced, with water permanently reaching the trunks of their trees. The ferry linking the city and the "Trans–Assurini" highway will have its operation altered. Most operations for taking sand from the river bed will be affected.

Consequences in the river bank region of the Big Bend of the Xingu

River islands, from upstream of Altamira to the Pimental and da Serra islands (80 km by 8–20 km wide) will disappear under water, and with them the vegetation found there. The tallest trees – samaúmas and Brazil nut trees, will take awhile to die.

The Gaioso and da Maria igarapés will be artificially widened, reaching up to 500 meters in width, and will have their beds reinforced with concrete, and their banks with tall dikes to carry water from the river to the inland reservoir. The smaller tributaries of these igarapés will continue outside the dikes, and will form flooded areas during the rainy season, drying up during the summer, affecting rural communities along the 27 and 45 road spurs.

This entire area will be crossed by the principal service road to the Pimental dam and work camp, and will also be crossed by 69kV transmission lines to power the work site. During the dam’s operation, 230 kV lines from the complementary power house will cross this area.

Most of the river’s flow will be diverted through the canals to the dam, returning to the Xingu only after passing through the turbines in the principal power house at Santo Antonio do Belo Monte.

The descent through the rapids of the Volta Grande, about 150 km long will be flooded in its initial third, and will be permanently drier in its last 100 km, with lower flow levels than the historical lows.

This, even if the operational scheme is indeed followed (generally dam operators choose to prioritize maximizing water available to the turbines, rather than downstream flows):

  • During the rainy season, lowest flows are in March, with 9,561 m3/second and in April, 9,817 m3/s – according to the EIA, a minimum of 15.7 % and 20.4 % respectively of these stream flows will be liberated – 1,500 m3/ s in March and 2,000 m3/s in April.
  • During the dry season, the lowest monthly flows on the Xingu River were 908 m3/s in August – when only 250 m3/s, some 27 % will be permitted to pass the dam; and 477 m3/s in September – when only 225 m3/s will be liberated. In October, where the lowest river levels are measured with 444 m3/s, only 200 m3/s will be permitted through the Pimentel floodgates.

That is, during the critical dry months, only 45% – 47% of the historic minimums registered will be permitted to flow through the Big Bend.

Navigation, always difficult in the dry season, will become impossible. The river bed, cut by islands and rocks, will practically turn into a series of shallow, stagnant pools during the dry season, with hot water during the day, and lukewarm water at night.

How will the fish, trapped in the pools without a chance to swim, or to flow upstream fare? And the cari ornamental fish that are caught and sold as aquarium fish? Will they survive? The biting horse flies that are a plague in the region will proliferate, since they always increase in number during the dry season. Will the small shellfish found in sand banks survive? What will the birds and waterfowl eat? How about the snakes and turtles? And the bees that depend on the flowering bushes along the shore?

Obviously, the lowering of the water level, facilitating access to the river bed will benefit manual gold mining, but drinking water and water to rinse their equipment will be more difficult to find, as will access to mining sites because of difficulties in river navigation. The wastes of these gold mines will pile up as eternal pyramids along the dried–up river bed.

One of the main problems of river bank dwellers will be access to drinking and domestic water. Some water wells should also dry up, because the water table will tend to drop.

At the confluence of the Bacajá river with the Xingu, the meeting of the waters of the two rivers produces a typically Amazonian phenomenon. During the dry season, the Bacajá has a low flow into the Xingu, whose flow is also reduced. During the rainy season, the Xingu can flow with such great force that it "invades" the mouth of the Bacajá, forming a "remanso", a hydro–dynamic barrier which permits the Bacajá to retain water and which floods lands in the lower part of the river. With the reduced flow of the Xingu following the construction of Belo Monte, the Bacajá should flow directly into the Xingu, and its greater force will increase erosion along the lower stretch of the river.

The principal rapids along the Big Bend, beginning with Jericoá, will dry out, and the extensions of sand banks and islands will increase. The gallery forests and mangroves along the lower stretch of the river will tend to die, because they will pass a significant period without being flooded by the water which nourishes and fertilizes them. In addition, their roots will no longer reach the declining watertable.

The Xingu River below the Jericoá beach will drain the four igarapés – the Paquiçamba, Ticaruca, Cajueiro, and igarapé Cobal – that are formed near the farms on the Transamazon highway, and the 45 and 55 roads. These will become totally dry at their source, reconstituting apart of their usual flow only during the rainy season.

7. Summary of the environmental impacts and other risks of the Belo Monte and Babaquara projects on the regional and planetary scale

The "new" hydrological system "hydroelectric reservoir" replaces the river ecosystems that preceded it. Existing habitats are destroyed, either entirely or in part, and other habitats will be created by the reservoir and the new relief and interfaces it defines. The first project, Belo Monte, will cover more than 400 km2, or 40,000 ha in two separate reservoirs. If the second dam, Altamira is built – and our simulation shows that only with the two dams operating will the Xingu "complex" be feasible – more than 6,000 km2 or 600,000 hectares will be flooded. Among the changes that can be predicted:

  • stratifications of temperature and light by layers – the deeper, the colder and darker it will get;

  • drowning and rotting vegetation, humus, and organic residues from the soil on the bottom of the reservoir, with resulting emissions of organic volatile or gaseous acids, hydrocarbons, carbon gases, and sometimes volatile or gaseous sulfides;

  • slow decay of dying tree trunks, and the slow putrification of brush which forms along the shallow banks and stagnant pools in the reservoir;

  • accumulation of sediments carried by the river and tributaries of the reservoir and the retention of a part of these sediments by aquatic plants;

  • evaporation of surface water, water passing through the floodgates and the tailrace; proliferation of some aquatic plants;

  • forced selection of microfauna species, of creatures living in the sediments, and of fish, crustaceans, mollusks, frogs and other reptiles that will survive in the reservoir;

  • blocking or making difficult the migratory routes for aquatic species; new stopover points for migratory birds and peri–aquatic animals; proliferation of insects which thrive on stagnant water (reservoir) and turbulent waters (dam spillways).

Each new ecosystem will permit species reproduction as long as biochemical conditions remain within certain limits, or if discontinuities in the food chain, or changes in the oxygen level of the reservoir are not too great. Other limits to reproduction include water velocity or stagnation, acidity, temperature, and concentration of metallic ions or toxic organic compounds.

Dams are also subject to the effects of degradation upstream, affecting tributaries of the Xingu, and lands in the basin – most common are increases in sedimentation caused by erosion, accumulation of untreated sewage and industrial wastes, contamination from agro–chemical use, and fermentation of excess organic materials, consuming a part of the oxygen in the water.

With the intensification of upstream agriculture and cattle ranching, there will always be the risk of accumulations of excess nutrients (nitrates, phosphates) and ammonia dissolved in water and sediments. As sewage from Altamira will also be able to accumulate in certain stretches of the reservoir, there can be a proliferation of algas and plankton, for example of cyanobacteria and others which cause adverse reactions in fish and humans. The process of eutrofication of the water body caused by the decomposition of the exposed portion of vegetation drowned in the reservoir emits carbon gas (CO 2 ). At the bottom of reservoirs, there is no oxygen, and the decomposition of material produces methane gas (CH 4).

In the initial years following the formation of the reservoir, the methane is produced by layers of forest leaves, humus, and from a part of the carbon in the soil; the gas continues to be produced in subsequent years by the decomposition of herbacious plants that grow, each year, in areas temporarily exposed during low–water periods. The water which passes through the turbines comes from deeper levels of the reservoirs, where methane is more concentrated.

The article by researcher Philip Fearnside carefully analyzes this process of emission of carbon gases, which contribute to the increase in the greenhouse effect on the global level, affecting the world’s climate. Considering the hypothesis of the construction of the two dams (Belo Monte and Babaquara), a part of the methane gas produced in the enormous Babaquara reservoir would be liberated in the reservoir itself and in the dam’s spillway and turbines, while another part would be passed downstream to Belo Monte, increasing its emissions level. The two dams, together, would have a negative result, in terms of greenhouse gas emissions, when compared to a thermoelectric plant generating the same quantity of electrical energy, during at least the initial 39 years following the closing of the floodgates on the first dam.

Besides this, there would be an increase in the reservoirs of organic acids (acetic, formic) and eventually of sulfides, and the progressive acidification of the water, with proven consequences for animal and human health, as well as for the mechanical structures of the dam itself. Damages from accelerated corrosion of metallic parts of equipment in contact with water have already affected Eletronorte’s Balbina dam (Amazonas state) and Celpa’s Curuá–Uma dam (near Santarém, Pará), both in the Amazon.

The acidity will permit greater solubility of heavy metals in the soil and compounds carried in sediments and in the turbulent waters, as well as wastes from activities such as mercury used in manual gold mining. The bio–accumulation of these metals via the food chain could affect aquatic and peri–aquatic animals, and eventually humans (as in the Minamata syndrome).

Enormous reservoirs such as Babaquara could also provoke seismic events. Belo Monte and Babaquara, which will be formed on crystalline rocky reads with natural fractures and caverns, could also run the risk of seepage into neighboring basins. This could also take place as percolation through the walls and lateral dikes in these dams (Belo Monte will have 30 such structures, and Babaquara even more).

Finally, construction of the Xingu Hydroelectric complex would mean the destruction of one of Brazil’s fluvial monuments, the Big Bend of the Xingu, for which compensation or mitigation of impacts is impossible. This story is told within the context of a survey of mega–hydroelectric dams by professor Oswaldo Sevá.

8. The third try by the dam–builders and electro–intensive industries, from 2003

During the 1990’s and the beginning of this decade, political and party polarization took place in Altamira and in Pará regarding this mega–project, generally found left–wing political parties (PT, PCdoB, PSB) in opposition to Belo Monte, with local and regional allies of the state and federal government (PMDB and PSDB) in favor.

In 2001 and 2002, all who followed the situation felt that a victory by Lula could bury the project once and for all, along with a host of other projects promoted by the Cardoso government. But this was not to be.

One of the reasons was that during the initial months of the new government, in 2003, Senator José Sarney, a political ally of Lula’s, convinced the federal government of the importance of and the opportunity presented by Belo Monte. In 2004, Sarney showed that he still commanded his fiefdom in the federal machine, by causing the changing of the president of Eletrobrás, principal shareholder in Eletronorte as well as other state electric companies.

The new directorship marked the return of engineer Muniz and his team to the forefront. Only this time, they had to perform contortions to diminish the size of the projected investments, recognizing that the company did not have the means to raise such a huge quantity of money. Seeing that investors would be required, they convinced the president of the state bank BNDES, even without the necessary technical–economic analysis to make a verbal commitment to finance a part of the project. In addition, a group of powerful companies was brought together, who would be capable of providing leverage for ensuring financing in Brazil and internationally, as well as guaranteeing the purchase of significant amounts of the energy that Belo Monte would generate. These included three state power companies, construction companies led by Camargo Correa, heavy equipment manufacturers such as ABB and Voith–Siemens, and electro–intensive industries, led by Alcoa, CVRD, and BhpBilliton.

Eletronorte would probably be only a minority partner in the so–called "Brazil Consortium", performing the functions of local and national political lobbying. Perhaps Eletronorte would also administer a fund for "regional insertion", allocating future royalties through a "special purpose company", and maintaining its power as a state within the State of Pará.

Given this new alignment, it becomes clearer that Eletronorte would not be the only operator of the dam, nor the principal partner in its ownership, and it is also clear that if it were built one day, the energy would not go to "the rest of the country", nor the Northeast on the edge of an energy crisis, but rather to electro–intensive industries in the Amazonian state of Pará.

9. Re–thinking mega–dams

Mega–hydroelectric dams are now seen in a different light than they were during the hey–day of large dam construction in the late 20th century. The social and environmental impacts of large dams are now well–known, and doubts have been placed on their overall effectiveness as an energy source.

Patrick McCully writes that the dams industry is fighting back, attempting to market dams as "renewable" energy sources. The recommendations of the independent World Commission on Dams provide a beacon toward a new paradigm where large dams are viewed only as a last resort for meeting water and energy needs.

Finally, the determination and activism of grassroots organizations in the Transamazon region is shown by the reprinting of letters through which they voice their opposition to Belo Monte and other dams on the Xingu to political leaders.

We trust this book will be a tool to promote increased, informed debate regarding the potential benefits and impacts of hydroelectric dams on the Xingu.


  1. In 2001, the then–President of Eletronorte, José Muniz Lopes, in an interview with the newspaper O Liberal (Belo Monte entusiasma a Eletronorte por Sônia Zaghetto, 15/07/2001), affirmed "Within the electric sector’s planning for the period 2010/2020, we’re looking at three dams – Marabá (Tocantins river), Altamira (previously called Babaquara, Xingu River) and Itaituba (São Luís do Tapajós). Some journalists say that we are not talking about these dams because we’re trying to hide them. It’s just that their time has not yet come. We’re now asking for authorization to intensify our studies for these dams. Brazil would be greatly benefited if we could follow Belo Monte with Marabá, then Altamira and Itaituba."

  2. In 2004, on the site of the Ministry of Mines and Energy, an official presentation was posted showing the Xingu projects within the mega–program called the South American Regional Infrastructure Initiatve (IIRSA) at www.caf.com.

  3. Eletronorte, EIA/RIMA, 2002