Credit for climate mitigation by Amazonian dams: loopholes and impacts illustrated by Brazil’s Jirau Hydroelectric Project

Abstract

Tropical hydroelectric dams are now one of the main destinations for funds under the Kyoto Protocol’s Clean Development Mechanism (CDM), with 1482 dams approved for credit and 840 dams in the CDM ‘pipeline’ awaiting approval. Thousands of dams are being built by countries such as China, India and Brazil, irrespective of any additional subsidy based on mitigation of climate change. Carbon credit granted to projects that would occur anyway allows the countries purchasing the credit to emit greenhouse gases that are not offset. Damage to global climate is further increased by CDM accounting procedures that undercount the greenhouse gases emitted by tropical dams. Still more damage stems from the limited mitigation funds being squandered on ‘nonadditional’ projects such as dams. An example indicating the need to eliminate credit for hydroelectric dams is provided by the Jirau Dam, now nearing completion on the Madeira River near Brazil’s border with Bolivia. The dam has severe impacts in addition to climate change. The project was approved (registered) by the CDM Executive Board on 17 May 2013.

Figure 1.  Location of the Jirau Dam and other sites mentioned.

On 17 May 2013, the Clean Development Mechanism (CDM) Executive Board approved the Jirau project for carbon credit, making it the largest ‘renewable energy’ CDM project to date in terms of tons (t) of carbon dioxide equivalent (CO₂-e) potentially mitigated [101]. This makes examination of this example especially relevant, because problems in carbon claims cannot be dismissed as aberrations that would be eliminated by the CDM’s validation and review process. In fact, the majority of proposals are eventually approved and it can be claimed that the CDM’s review process does not result in a portfolio of approved projects that has been purged of its many problems. The purpose of the present article is a questioning of the CDM guidelines themselves, not whether or not this particular project meets the current CDM criteria.

The Kyoto Protocol was negotiated in 1997 and entered into effect in 2005 for the purpose of reducing net global greenhouse gas (GHG) emissions, thereby helping to keep atmospheric concentration below levels that would be ‘dangerous’ for the climate system. The Protocol’s first commitment period was from 2008 to 2012, and the mitigation activities in future Kyoto Protocol commitment periods or under replacement agreements will need to benefit from the experience gained so far. An important feature of the Kyoto Protocol is the CDM, which allows the countries that have accepted a maximum limit or ‘cap’ on their national emissions (Annex I countries) to meet part of their emissions-reduction commitments through projects in countries without a national cap (non-Annex I countries). Because the host countries for these projects have no national caps, there is no safeguard at the national level accounting for GHG emissions if the benefits claimed at the project level do not reflect true reductions in net emissions. The Kyoto Protocol therefore requires that all CDM emissions reductions be ‘additional’ to those that would have occurred in the absence of the project. This gives rise to the term ‘additionality’ to describe the degree to which emissions reductions would not occur without this subsidy.

It is important to maintain a clear distinction between what is ‘additional’ in the spirit of the Kyoto Protocol, and what is classed as such by applying the operating procedures that have been adopted in implementing the CDM. These are emphatically not the same thing. Meeting the objective of the Kyoto Protocol in reducing global net emissions of GHGs requires that the CDM projects financed would, in fact, only take place because of the revenue from sale of carbon credits. This is very different from justifying a classification as ‘additional’ by providing values for the internal rate of return (IRR) that, after a variety of adjustments permitted by the CDM, imply that the dams would otherwise not be profitable. The current article uses the term ‘additional’ in the sense of the spirit of the Kyoto Protocol, unless otherwise specified. The focus of the present article is on the adequacy or inadequacy of the CDM’s activities in subsidizing tropical hydroelectric dams as a means of combating climate change – not on whether or not the claim for carbon credit by the example case (Brazil’s Jirau dam) conforms to the current regulations of the CDM. Hydroelectric dams, which have become one of the dominant forms of mitigation in the CDM, are rarely additional, as they have been, and continue to be, built by the host countries without recourse to added income from carbon credits.

The CDM has a three-step process. In the first step, the CDM project team contracts a Designated Operational Entity (DOE), which is an independent entity designated by the United Nations Framework Convention on Climate Change (UNFCCC) CDM Executive Board to perform validation work. The second step is the ‘validation’ by the DOE, including site visits and interviews. The progress of the project can be viewed as part of the ‘pipeline’ of projects that are tracked by the UN Environmental Program’s Risoe Center [102]. This tracking is based on information collected from validators and is independent of the CDM Executive Board. The third step is the project developers’ request to the Executive Board for ‘registration’, and subsequent review and approval by the Board. Once registered, the project is eligible to receive carbon credit. The actual granting of carbon credit (Certified Emissions Reductions [CERs]) will be done after each project year (or other time period), at which time another DOE will verify whether the emission reductions have been achieved and, if the DOE finds that they have, the CDM Executive Board will grant CERs as the resulting tradable product.

As of 1 December 2012, the CDM had issued credit to 586 hydroelectric projects. The cumulative number of hydropower projects registered increased from 1225 to 1482 between 3 June and 1 December 2012, which represents an approval rate of almost two dams per day. Another 896 projects had been registered but not yet issued credit, and 840 dams were in the ‘pipeline’ awaiting completion of the approval process for CDM funding [102]. A large majority of projects entering the pipeline are eventually approved: the overall rejection rate for the CDM is 19.8% through the validation phase, after which 4.2% are rejected by the Executive Board [102]. For hydro projects listed by the Risoe Center as having had a decision (i.e., as either having been rejected or having passed the registration phase), only 14.2% had been rejected [102].

For the period ending in 2012, the CERs issued for hydropower totaled 112.7 million t of CO₂-e (or the sum of all GHGs, such as CH₄, expressed as the amount of CO₂ that would have the same impact on global warming over a 100-year period). However, for all approved and global ‘pipeline’ hydropower projects, the annual total number CERs for the period through 2012 is expected to be 331.2 million t of CO₂-e [102], which is equivalent to 90.3 million t of carbon per year. The number of CERs expected for the 2013–2020 period from hydro projects approved through 1 December 2012 totals over 7 billion t of CO₂-e [102]. For comparison, the total emissions from fossil fuels in the USA in 2010 is estimated at 6.8 billion t of CO₂-e [1].

Little of the carbon represented by hydropower CERs is additional to that which would have been emitted without the projects [2–4]. The countries that purchase the CERs, therefore, emit carbon to the atmosphere without any real offsetting of the emissions by the CDM projects. Hydroelectric projects in the CDM also consume a substantial share of the money that the world has for combating climate change; the amount of credit through 2012 (331.2 million t CO₂-e annual total for the approved and ‘pipeline’ projects) would be worth US$8.5 billion considering the 2008 CER price (˜$25/t CO₂-e) used in the Jirau Project Design Document (PDD) ([103]; page 52). Note, however, that CER prices are highly volatile; they have crashed since 2008, but would increase dramatically if international negotiations achieve major commitments to reducing emissions. Allocating a portion of the available mitigation funds to nonadditional projects means that projects with real climate benefits are not undertaken. Hydropower proposals represent 26% of the total carbon credits in the CDM pipeline [102]. The Jirau Dam on the Madeira River in Brazilian Amazonia provides an example illustrating these problems.

Jirau Dam & the CDM

The Kyoto Protocol’s CDM is intended to provide a means by which projects in developing countries can be funded through the sale of carbon credits to developed countries (Annex I countries); thus, allowing the developed countries to meet their Kyoto Protocol emission quotas (assigned amounts) more economically, while at the same time helping the developing countries to achieve ‘sustainable development’. One of the most controversial parts of the CDM has been projects for hydroelectric dams, especially in tropical areas such as the Brazilian Amazonia [5,6]. The CDM Executive Board’s registration of the CDM project for Jirau on 17 May 2013 is effective retroactively to 26 December 2012, thus allowing the project to sell carbon credit to the European Union Emissions Trading Scheme [101].

The Madeira River Dams (Jirau and the adjacent Santo Antônio Dam) are both now nearing completion. The Madeira River is a major Amazon tributary that drains parts of Brazil, Bolivia and Peru (Figure 1). Its average flow at Jirau (17,686 m³/s) is 24% greater than that of China’s Yangze River at the Three Gorges Dam. The Madeira River dams have, for many years, been the subject of intense opposition by groups concerned with the environment and human rights [104]. These dams would certainly not be considered to represent ‘sustainable development’ by most people’s understanding of that very flexible term, but the Kyoto Protocol’s requirement that all CDM projects contribute to sustainable development ([7]; Article 12, Paragraph 2) has been effectively neutralized by the decision that each country decides for itself what sustainable development is, and any project submitted to the CDM by the host country’s Designated National Authority is automatically assumed to represent sustainable development. The Jirau project has now passed through the various steps in the CDM’s approval process, culminating with the completion of an ‘official assessment report’ on 17 May 2013 [8], and approval of the project on the same day by the CDM Executive Board. The report was sharply criticized during the public comment period [9], as was the Jirau project’s PDD [105,106]. An examination of the Jirau case reveals issues in the ability of the current project evaluation system of the CDM to prevent approval of projects that contradict the overall purpose of the Kyoto Protocol and the UNFCCC by doing harm to the global climate, in addition to causing notable social and environmental impacts in the host country (and, in this case, two neighboring countries as well).

Located in Brazil’s state of Rondônia near the border with Bolivia (9° 15’ 17.96” S, 64° 38’ 40.13” W), the Jirau Dam is being built by the French multinational GDF Suez together with Energia Sustentável do Brasil S.A., which is a consortium composed of GDF Suez (60%), Eletrosul (20%) and Chesf (20%). On 2 July 2013, the Brazilian government approved a proposed sale of a 20% share by GDF Suez to the Japanese firm Mitsui. The beginning of commercial energy production was delayed several times and began in September 2013; the full 3750 MW of installed capacity is expected come online by 2015. The PDD for the carbon credit proposal [103] is similar to other CDM proposals for dams. The Jirau Hydropower Project completed the final version of its PDD on 12 April 2012 and the validation process began on 24 April 2012. It is emphasized again that the present article is intended to examine the rationale for the current CDM rules for granting carbon credit to hydroelectric dams (using Jirau as an example), not whether the Jirau carbon project conforms to the current CDM rules. Problems include lack of additionality, underestimation of GHG emissions by the dam itself, emissions from deforestation induced by the dam, and many noncarbon environmental and social impacts. The author would like to highlight that many of the problems reviewed here also apply to the other three large hydroelectric dams in Brazil’s Amazon region that have submitted carbon projects to the CDM (Santo Antônio, Teles Pires and Dardanelos). These problems, including lack of true additionality, appear to be common to many dam projects throughout the world (e.g., [107]).

Environmental & social impacts

▪ Jirau & sustainable development

The Jirau PDD states that “The Brazilian Energy Research Company (EPE), part of the Ministry of Mines and Energy, … Considers economic and social as well as environmental aspects … In this context, EPE also considers Brazil’s insertion in international agreements … This view implies that all external costs and benefits are adequately assessed to identify most beneficial projects and to maximize their socioeconomic benefits and minimize their impacts” ([103], page 5). The irony of this portrayal is apparent, given the enormous accumulation of criticism from civil society of the Brazilian Energy Research Company plans, and of the Madeira River dams in particular, based on their socioeconomic and environmental impacts [10–12,104,108]. A 2010 letter to the president of GDF Suez from 16 environmental groups describes multiple deficiencies in dealing with environmental and social impacts at Jirau [109]. These problems led to GDF Suez being elected a 2010 finalist for the ‘public eye award’ that is given annually to the world’s most irresponsible company [110].

The PDD asserts that “…the Jirau HPP [hydropower plant project] is being developed according to the best technical and environmental practices and standards” ([103], page 5). The local people and the civil society groups mentioned above would contest this claim. As the Folha de São Paulo described it, the licensing process was a ‘collection of errors’ [13]. Jirau’s environmental impact assessment (EIA) was done jointly with the neighboring Santo Antônio Dam [111]. The preliminary license and the installation license were granted overriding the technical staff of the federal environment agency, the Brazilian Institute for the Environment and Renewable Natural Resources (Instituto Brasileiro do Meio Ambiente e dos Recursos Naturais Renováveis [IBAMA]) [14,15,112].

All CDM projects must contribute to ‘sustainable development’ ([7]; Article 12, Paragraph 2). Sustainable development is generally considered to have three pillars: social, environmental and economic. The PDD stresses the social benefits of jobs provided by the construction project, pointing out that “more than 70% of workers are hired locally” ([103], page 6). However, this apparently must mean that workers ‘hired locally’ include migrant laborers who have traveled from other parts of Brazil to the construction site of their own accord before being hired. A parliamentary commission of the Brazilian National Congress is holding hearings to determine whether human trafficking was employed to bring ‘slave’ labor to Jirau from other states [113]. The social problems caused by massive migration to the construction site and surrounding area would be better described as an impact rather than a benefit of the project [16,17]. An indication that conditions at the construction site are less than ideal is given by persistent labor unrest at Jirau, including two major incidents of criminal arson [18]. Although labor discontent is a common phenomenon, the uprisings at Jirau stand out among all of the hundreds of construction sites that have been established under the Program for the Acceleration of Growth.

▪ Blockage of fish migration

One of the major impacts not mentioned in the PDD is blockage of fish migration. The only aspect of this mentioned is that the bulb turbines used will kill less descending fish larvae than other types of turbines would ([103], page 5). Readers of the PDD will have no idea of the dam’s major expected impact (together with the Santo Antônio Dam under construction immediately downstream of Jirau) in eliminating the fishing of the giant Madeira River catfish (Brachyplatatystoma rouxeauxii and Brachyplatystoma platynemum), not only in the Brazilian portion of the Madeira but also in Peru and Bolivia [19–21]. These catfish have, until now, been a major source of income and food for the local people in the Madeira Basin in these three countries, and the dam projects do not even admit, let alone compensate for, the loss of livelihoods.

▪ Flooding in Bolivia

Flooding in Bolivia is an important part of the controversy surrounding the Jirau Dam. Because the upper end of the reservoir is planned to terminate precisely at the Brazil/Bolivia border (at the town of Abunã), any raising of the water level would flood land in Bolivia. The dam can cause the water level to rise in the river upstream of Abunã in two ways. First, by a deliberate increase in water management, where the level at the dam is increased thereby extending the reservoir proper into Bolivia. Second, by the expected formation of a ‘backwater stretch’, where coarse sediment accumulates at the upper end of the reservoir, and this mound of sediment impedes water flow and causes the water level to rise above its natural level in the river stretch upstream of what is officially defined as the ‘reservoir’.

Flooding in Bolivia is not mentioned as an impact. The PDD states that water levels at Abunã are to “follow their natural seasonal variation” ([103], page 8), as required by a 2006 decision of Brazil’s National Water Agency (Agência Nacional das Águas, resolution 555/2006). The claim that no flooding would occur in Bolivia has been contested multiple times [15,20,22,23,114]. The issue is a delicate one diplomatically. Sedimentation and consequent raising of water levels in the ‘backwater stretch’ is expected to result in flooding in Bolivian territory, both along the Madeira River and the Abunã River (a tributary of the Madeira that also forms part of the Brazil/Bolivia border). Impacts of flooding would include effects on a protected area on the Bolivian side of the Madeira above the town of Abunã. A detailed review of the Madeira River ‘sediment controversy’ is available in [15] and additional information in [115].

The EIA considers the reservoir to be shorter in length at any given water level during the high-water period than during the low-water period ([111]; Tomo A, pages VII–8). To prevent flooding in Bolivia, the planned operation of the dam in the PDD would keep the water level at the dam at 90 m above sea level for 4 months (January–April), at 85 m for 4 months (May, June, November and December) and at 82.5 m for 4 months (July–October) ([103]; page 9). The calculated effect is only for the length of the reservoir per se, not for the backwater stretch. The planned management of the water level will not prevent a mound of sediment from forming at the top of the reservoir, and consequent flooding in Bolivia by the water that is held back by this impediment. The Madeira River has one of the highest sediment loads in the world, accounting for approximately half of all of the sediment in the Amazon River [24].

The water management plan presented in the PDD to avoid flooding in Bolivia has lower water levels in 6 months of the year ([103]; page 9), as compared with the 2005 EIA ([111]; Tomo A, pages VII–13), and there are no months in which the PDD plan calls for higher levels (Table 1). Aside from indicating the inaccuracy of the 2005 claim in the EIA that the plan presented in that document would cause no flooding in Bolivia ([111]; Tomo A, pages VII–16), the lower levels imply less power generation and an increase in the calculated ‘additionality’ of the carbon claims. More importantly, either one of the operation plans makes clear the very substantial impact on Jirau’s operation from the existence of the international border with Bolivia at the upper end of the reservoir. Operating the reservoir for much of the year at a water level below the ‘maximum normal’ elevation of 90 m implies lower electricity generation due to the lower head at the dam. Although the Jirau reservoir was originally expected to operate at the 90 m reservoir elevation throughout the year ([25]; pages 2.155–2.167), the plan was changed in 2004 to adopt a variable reservoir level in order to avoid flooding in Bolivia ([26]; Vol. 1, page 2.2). This raises the possibility that the consortium may be expecting that, at some future date, Brazil could reach an agreement with Bolivia to allow the water level to be raised to 90 m (or perhaps even higher) throughout the year. Brazil is negotiating with Bolivia over the planned binational Guajirá-Mirim Dam (also known as the Cachoeira Riberão Dam or as the ‘Binacional’ Dam) upstream of Jirau [27–29]. In addition to adding to flooding by the Jirau reservoir, raising the water level beyond that specified in the PDD could make the Jirau Dam’s carbon credit claims considerably less ‘additional’ than the calculations presented to the CDM imply.

On 2 July 2013, Brazil’s National Agency for Electrical Energy (Agência Nacional de Energia Elétrica [ANEEL]) requested that the Ministry of Mines and Energy negotiate with Bolivia to allow the increase in the water level at Jirau [29]. It is intriguing that ANEEL made this request only 46 days after the CDM Executive Board approved the Jirau project for carbon credit. Revenue from the extra power that will be generated if Bolivia allows its territory to be flooded by the reservoir proper would make the CDM project’s claim that the dam would be unprofitable without carbon credit even less credible. It might be argued that the PDD should have covered this possibility both in the investment analysis and in the description of project impacts, given that the request to Bolivia appears to indicate the intention of the project developers to operate the dam at the higher water level.

▪ Other impacts

Various other impacts of the dam are unmentioned in the PDD, such as mercury methylation in bays along the river edges [30], and a variety of social and biodiversity impacts. One unmentioned impact with relevance to GHG emissions is the dam’s providing of a link in a series of planned industrial waterways (hidrovias) that would promote the advance of soybeans into rainforest areas in Brazil, and especially in Bolivia, thus causing emissions and other impacts from deforestation [31–33]. The benefits of the waterways are emphasized in the Jirau Dam’s viability study ([26]; Vol. 1, pages 1.21–1.24). Current deforestation rates have increased sharply in the immediate vicinity of Jirau and the adjacent Santo Antônio Dam [34]. This area was the biggest hotspot of deforestation in Amazonia in December 2010 [35,36]. Substantial future deforestation can be expected upstream for soy in areas to be served by the planned waterways [37]. The PDD, however, claims that the dam will have no ‘leakage’ causing emissions outside of the project area ([103]; page 59). Another deforestation impact was a degazetting in 2009 of 140,000 ha of the Rio Verde State Reserve to accommodate Jirau; the Rio Pardo Federal Reserve was increased in area by the same amount to compensate for this, but only half of the area added to the reserve had forest – the rest had been cleared and was occupied by 5000 families [112,116]. This implies a net loss of 70,000 ha of protected forest due to Jirau.

GHG emissions

▪ CDM loopholes

The PDD classifies the dam as only a ‘minor emissions source’ of methane (CH₄) ([103]; page 13), but makes clear that officially the emissions are zero, and that no measurements or monitoring are required ([103]; page 58). No technical studies are cited to substantiate the claim that the dam would only be a ‘minor’ source of CH₄. The claim rests on the loophole in the CDM’s regulations classifying dams by power density, or the ratio of installed capacity to reservoir area. Dams with small reservoirs and large installed capacities are allowed to claim that they have no emissions. In reality, having a small reservoir reduces, but does not eliminate, emissions from the reservoir surface (in addition to those resulting from emissions from other sources, including deforestation).

The loophole in the CDM regulations that allows dams with high power densities to claim zero emissions specifies 10 W/m² as the limit for this complete exemption [2]. Jirau claims a power density of 18.05 W/m², allowing the PDD to conclude that “Therefore, … project emissions can be neglected” ([103]; page 7). CDM regulations allow the power density to be calculated, not in the normal way as a simple ratio of installed capacity to reservoir area, but rather using the ‘reservoir area increase’ as the denominator. This means not counting the natural river bed, which the CDM proposals take to include not only the river channel, but also the wide area that is temporarily flooded at the peak of the high-water period in Amazonian rivers. These areas support ‘Várzea’ (floodplain) forests that are adapted to survive flooding for several months, but not permanent flooding as occurs when they become part of a reservoir. The Jirau PDD uses a ‘reservoir area increase’ of 207.74 km² ([103]; page 8), rather than the full reservoir area of 361.60 km² ([103]; page 9). If the power density were calculated using the full reservoir area and the current configuration of 3750 MW, the result would be 10.4 W/m², or barely above the 10 W/m² cutoff for making use of the loophole for considering the dam’s emissions to be zero. If the 3300 MW installed capacity originally planned were used, the power density would have been 9.1 W/m², making the project ineligible for claiming zero emissions from the dam. Adding the six extra turbines after the dam location was changed in 2009 gave a double assurance of being able to take advantage of the loophole.

The CDM regulation’s not counting the full reservoir area is apparently based on the mistaken assumption that the water over the natural river bed does not emit CH₄[2]. The CDM’s regulation allowing dams with power densities over 10 W/m² to claim zero emissions cites as its justification a submission by Marco Aurelio dos Santos and Luiz Pinguelli Rosa that is not available on the CDM website or under the same title on other websites [38]. This submission is also the origin of the low emission of 100 g CO₂/kWh attributed to dams in the 4–10 W/m² range (note, the lower limit of 4 W/m² was established in 2007, previously the limit was 5 W/m²). These assumed emission values ignore the main sources of CH₄ release in most dams (the turbines and spillways), as well as underestimating reservoir surface emissions due to a series of mathematical errors [6,39]. The 10 W/m² cutoff stems from a suggestion by Luiz Pinguelli Rosa, former head of Eletrobras, from 1996 (i.e., before the Kyoto Protocol) regarding supposed climate benefits of Brazil’s proposed Belo Monte Dam ([40]; contested by [41]). This remains an important controversy to this day [42–47].

▪ Jirau emissions

The PDD repeatedly cites official documents ([103]; pages 18–19) referring to hydropower as ‘clean’ or as a ‘non-emitting source’ [48,49]. The assumption that hydroelectricity is clean energy has been contested, especially for Amazonian dams [6]. Although the claim that hydropower is clean has been repeated many times, such claims have been scientifically untenable for some decades [50]. Multiple studies indicate large emissions in tropical dams [51–57]. The high water flow rate through the reservoir at Jirau will result in lower emissions than at other Amazonian dams, but emissions will not be zero.

The question of whether the water in the reservoir will stratify is important to the potential for emissions. In large reservoirs, such as that of the Tucuruí Dam on the Tocantins River in Brazil’s state of Pará, the water divides into layers, with warmer water in a surface layer 2–10 m thick (the epilimnion) that is in contact with the air and contains oxygen, and cold water in a bottom layer (the hypolimnion) where oxygen is almost completely absent. The anoxic water at the bottom means that CO₂ cannot form, and all decomposition of organic matter in the sediments produces CH₄ instead. Multiple studies have shown high CH₄ emissions in stratified tropical reservoirs [6].

The EIA for Jirau calculates that the water in the reservoir will not be stratified based on the turnover time and on the Froude density equation that relates stratification to water velocity ([111]; Vol. 7, page 3.8). The one-dimensional models used in the EIA have been criticized by Forsberg and Kemenes as inadequate to model stratification in the irregularly shaped reservoir, and these authors expect stratification along the reservoir’s edges [30]. Stratification would occur in the bays and other features where water velocities are much lower than the all-reservoir averages used in the EIA calculations. Stratification with anoxic bottom water can be expected in the flooded mouths of tributaries that enter the reservoir, as shown by simulations performed by the proponents at the request of IBAMA ([58]; Annex V). The anoxic sediments in these edge areas can be expected to produce CH₄, some of which would be emitted through the surface. However, unlike reservoirs with more widespread stratification, most of the portion of the dissolved CH₄ that does not reach the surface as bubbles will be prevented from reaching the turbines. This is due to the presence of oxygenated water in the main channel where the water is moving faster. However, recent measurements in the very similar Santo Antônio reservoir immediately downstream of Jirau indicate high fluxes of CH₄ from the water surface in tributaries ([59]; page 25), suggesting stratification, while a high CH₄ concentration in water immediately below the Santo Antônio Dam (60]; page 28), indicates that not all CH₄ is oxidized before reaching the turbines.

Jirau will have four ‘big pockets’ (bolsões) or shallow bays along the eastern shore of the reservoir ([58]; Annex V). These will be associated with two tributary streams (Jirau and Caiçara), the Mutum-Paraná River (which is the largest tributary entering the reservoir) and the ‘Bolsão do Mutum’ formed by the Cotia River (a tributary of the Mutum-Paraná). At the 90 m water level, the areas of these bays will be 9.84, 17.84, 22.27 and 18.24 km², respectively ([58]; pages 124–125). Since the Jirau Reservoir has only been filled in 2013, no flux measurements are yet available. However, flux measurements were made in February 2012 in two tributaries to the very similar Santo Antônio Reservoir; the tributaries emitted 16 and 39.6 mmol/m²/day of CH₄, respectively ([59]; page 25). In the main river the fluxes were minimal: 0–0.5 mmol/m²/day. This can at least give an idea of the magnitude of the flux at Jirau, although uncertainty is obviously very great. Two of the four bays at Jirau are expected to stratify when the water level is at 90 m: those associated with the Jirau and Caiçara streams ([58]; page 148). The ‘Bolsão do Mutum’ would have almost no oxygen due to the very slow turnover time, but is too shallow to stratify ([57]; page 235). Conservatively, considering only the Jirau and Caiçara streams and the mean of the fluxes from the two streams measured at Santo Antônio, the surface emissions from these Jirau tributaries would total 4494 t of CH₄/year, equivalent to 94,372 t CO₂-e considering the global warming potential (GWP) of 21 for CH₄ still used by the CDM (more recent GWP values, such as the value of 28–34 for CH₄ in the Fifth Assessment Report of the IPCC for the same 100-year time horizon, imply greater impact of dams [61]CO₂-e considering the global warming potential (GWP) of 21 for CH₄ still used by the CDM (more recent GWP values, such as the value of 28–34 for CH₄ in the Fifth Assessment Report of the IPCC for the same 100-year time horizon, imply greater impact of dams [61]). The Fifth Assessment Report also presents a GWP value of 86 for a 20-year time horizon [61], which is more relevant for policies to avoid exceeding a 2°C temperature increase; this implies a quadrupling of the impact of dams. If the ‘Bolsão do Mutum’ emits at the same level, these emissions would increase by approximately 50%. These emission values assume the year-round 90 m water level that would apply if negotiations with Bolivia are successful, otherwise these tributaries would be stratified completely for 8 months of the year and ‘intermittently’ for an additional 3 months ([58]; page 148). Year-round operation at the 90 m level might result in stratification of the Mutum-Paraná River during the period of low stream flow in this tributary (July–December). The danger of stratification of the Mutum-Paraná was highlighted by Forsberg and Kemenes in the context of mercury methylation [30]. The consortium was sufficiently concerned about GHG emissions from the Mutum-Paraná that it bulldozed part of the area to be flooded and buried the biomass in shallow pits ([8]; pages 186–187).

CO₂ will be released by aerobic decomposition of vegetation flooded by the Jirau Reservoir, including the killing of the Várzea forest. Considering the areas of each vegetation type (with the smaller reservoir that would have been created with the dam in its originally planned location) and the biomass estimates (above-ground only) for each vegetation type presented in the EIA ([111]; Tomo B, Vol. 7, Annex 2), decay of this biomass represents 2.7 million t of CO₂ emission. Of course, the deforestation stimulated elsewhere by the hydroelectric project and by the associated waterways represents a still greater emission of CO₂.

In summary, the stratified tributaries and other sources will result in GHG emissions. While the amounts may appear substantial, they are modest as compared with most other hydroelectric projects in Amazonia.

Additionality & carbon credit

▪ Internal rates of return

The Kyoto Protocol requires that any emission reductions claimed for CDM credit must be “additional to any that would occur in the absence of the certified project activity” (Article 12, paragraph 5). Establishing ‘additionality’ requires a hypothetical baseline scenario representing what would have happened without the mitigation project and demonstrating that the project (in this case, construction of the dam) would not have taken place without CDM funding. The burden of proof for establishing additionality rests with the project developers.

CDM regulations allow projects to calculate an IRR and compare it to a benchmark IRR value, in order to establish that CDM funds are needed to make the dam profitable. The IRR represents the highest discount rate at which an investment of capital would be considered profitable. A discount rate is the percentage by which costs and benefits are devalued for each future year in translating future values into their equivalent in current terms (i.e., net present value). There are many possible adjustments to the IRR values used, with major implications for investment decisions.

The benchmark IRR is calculated based on the expected costs and revenues from power generation. In the case of the Jirau carbon project, these expected values were based on the dam’s official design at the project start date (22 July 2008 according to the PDD, [103]; page 24). The consortium led by GDF Suez won the bidding for Jirau on 19 May 2008 and, shortly thereafter, announced the plan to move the dam 9.2 km downstream to Ilha do Padre where construction costs would be lower, reportedly by R$1 billion, or approximately US$500 million [62]. The river at Ilha do Padre is wider than at the original location at Cachoeira Jirau, allowing additional turbines to raise the installed capacity. The original plan for the dam was to have 44 turbines totaling 3300 MW (the ‘base case’ used for the calculation of carbon benefits in the original PDD, later revised to 46 turbines in the current one). In March 2010, the plans at Ilha do Padre evolved to a design with 50 turbines totaling 3750 MW ([103]; page 26). It is evident that the plan for the dam with at least 46 turbines at the downstream location was the real scenario envisaged by the developers prior to the official start date for the carbon project, and it appears that this was also the plan prior to the bidding, thus allowing the consortium led by GDF Suez to win the contract with a bid for the electricity tariff 21% lower than the bid by the rival consortium led by Odebrecht [63]. The National Council for the Environment (Conselho Nacional do Meio Ambiente) questioned the move to Ilha do Padre in its 17–18 June 2008 meeting, given that the change was to occur without any additional environmental studies or consultations with local residents, but IBAMA nevertheless approved the preliminary installation license for Jirau on 14 November 2008.

Later, in response to criticisms of the carbon project, the consortium voluntarily made a calculation of the IRR with the six additional turbines to be installed at Ilha do Padre in order to show that the carbon project would still be considered additional under CDM regulations ([64]; page 62). This new 50-turbine version of the ‘optimized scenario’ (the previous version was with 46 turbines) was presented in a revised version of the PDD (version 4 of 18 April 2012), which calculated an equity IRR of 7.5%, as compared with 6.8% in a revised ‘base case’ with 46 turbines ([103]; page 50). No such re-calculation is known with the increased water level that is expected, assuming that permission to allow the reservoir proper to flood in Bolivia is obtained. However, the 2004 viability study had calculated that operating the dam year-round at the higher (90 m) water level would increase generation by 12% ([26]; Tomo 1, Vol. 1, page 1.1). The 12% figure is an underestimate for the effect of the current change, since it was based on the water management plan in the 2004 viability study rather than the lower water levels proposed in the PDD, in addition to being based on the smaller 44-turbine installed capacity (3300 MW). Permission to flood in Bolivia has been seen as a likely diplomatic outcome since a meeting of the presidents of Brazil and Bolivia in January 2008, 6 months before the start date of the Jirau carbon project [65]. This would make the Jirau Dam substantially more profitable and less likely to be considered additional.

The PDD emphasizes a comparison of the 6.8 and 7.5% IRR values (for the 46- and 50- turbine configurations, respectively) with a project-specific IRR value of 15.7%, implying that the dam would be unprofitable by an ample margin in either case ([103]; page 50). However, the CDM’s decision to grant carbon credit was apparently based instead on comparison of a calculated IRR of 10.9% with a “standard benchmark for baseline conditions” of only 12.46% ([64]; page 60). This benchmark IRR was calculated using a value of 50% for the financial leverage, rather than the 70% leverage used for the much higher project-specific IRR value; the standard financial leverage of 50% is required by Investment Analysis Guidance 18 ([66]; Annex 5) and resulted in revision of the IRR values in 2012 at the request of the project validator ([64]; page 61). The result is that Jirau is no longer indicated as unprofitable by a wide margin, and the dam would probably be classed as nonadditional (and therefore ineligible for any carbon credit) if the IRRs were calculated with 50 turbines and the year-round water level at 90 m above sea level. CDM regulations ([67], pages 38–39) indicate that, as a registered project, Jirau is already required to submit a new PDD with the 50-turbine configuration and that in the future it would have to submit another revision if the higher water level is approved; the revised additionality calculations could result in suspension of CERs.

CDM projects can choose between two types of IRR for their calculations to demonstrate additionality. The Jirau project developers used a Capital Assessment Pricing Model to calculate a benchmark IRR in terms of return on equity. Equity IRR is different from the project IRR calculation based on Weighted Average Cost of Capital, the option used, for example, in the CDM projects for the Teles Pires and Santo Antônio Dams. Equity IRR is from the viewpoint of a shareholder, and is generally higher than project IRR, which is from the perspective of the project as a whole. The two cannot be directly compared.

Project proponents have wide latitude in choosing an IRR value to use as a benchmark. Obviously, there is an inherent interest in picking a high value so that the dam project will be classified as unprofitable when compared with the benchmark, thus making it ‘additional’ and eligible for CDM credit. The 15.7% ‘project-specific’ benchmark that the PDD emphasizes for Jirau is calculated in the PDD based on various correction factors ([103]; page 35), including country risk (2.73%), equity risk premium (6.20%) and a multiplier of 1.60 for adjusted industry Beta. The principal justification given for this benchmark is not the rationale behind the corrections applied, but rather citation of confirmation from an almost identical value in a report by the World Bank [68], which refers to Brazil’s ANEEL as having said (without referencing any document) that “investors are prepared to invest in electricity generation only when rates of return are approximately 15%” ([103]; page 34). This figure is repeatedly referred to as a World Bank estimate, and the ‘prestige and experience’ of the Bank are mentioned explicitly to increase the credibility of this IRR value ([103]; page 34). Nevertheless, the origin of the number is an unreferenced statement by an anonymous ANEEL employee.

▪ Displaced emissions

The PDD states that “the electricity displacement will occur at the system’s margin, i.e., mainly fossil fuelled thermal plant generation will be substituted” ([103]; page 15). However, Brazil plans to build dozens of other dams over the next decade, and a CDM-subsidized dam could really be displacing another less-profitable dam (rather than fossil fuel), especially if dam building is as unprofitable in the absence of subsidies as the PDD claims. Of course, Brazil’s future priorities regarding its energy matrix could be different than they are today: this dam could just as well be displacing a future energy-efficiency program, or a wind, solar or tidal energy source instead, in which case the justification for claiming carbon credit for Jirau would evaporate even if the dam really were additional.

On the other hand, electricity displacement may not occur at all. The fundamental baseline is often conceived as one in which more expensive sources of power will be ‘switched off’ as more CDM-financed hydropower comes on line. But in the case of Brazil, nothing will be ‘switched off’. Instead, power generation facilities for both hydropower and gas are expected to expand enormously over the next decade [49]. Rather than switching off some existing plant, the new facilities will continue to be added to the existing generation system. The additional help from the CDM merely subsidizes the growth that is already planned.

Because CDM power projects are often built to extend the capacity of the power grid instead of replacing existing capacity, the CDM Executive Board has two baseline methods: one is ‘operating margin’, in case a CDM project replaces existing fossil-fuel capacity; and the other is ‘build margin’, to be used when a project ‘replaces’ capacity that would have been added to the grid in absence of the project. In Brazil, the ‘build margin’ is largely hydropower. Nevertheless, the threat of increasing fossil-fuel generation is often used by authorities as an argument for relaxing environmental licensing hurdles for building dams [117]. In the case of the Jirau and Santo Antônio Dams, in 2007 then-president Luiz Ignácio Lula da Silva argued that the government would build nuclear power plants instead if then-environment minister Marina Silva did not approve the licenses for constructing the dams [118]. As a country where over 70% of the electricity comes from hydropower, and where many undammed sites with hydroelectric potential still exist, other dams represent the most likely alternative. This would be the case on a time scale of years even if thermoelectric generation is temporarily increased to compensate for individual generation shortfalls and delays in dam construction. Even if nuclear plants were really the alternative to dams, as President Lula claimed, the ‘build margin’ would not be fossil-fuel combustion. Needless to say, the choice of the 100% fossil-fuel ‘operating margin’ instead of the ‘build margin’ for the PDD’s calculations of additionality implies much more carbon credit and financial return for the Jirau CDM project.

▪ The effect of subsidies: ‘E minus policies’

Classification as an ‘E minus policy’ allows the effect of subsidies to be removed from the calculation of the project’s expected IRR, making the project appear to be less profitable and therefore more likely to be considered additional. The CDM Executive Board has ruled that “A baseline scenario shall be established taking into account relevant national and/or sectoral policies and circumstances, such as … power sector expansion plans” ([69]; Annex 3, paragraph 4). The Brazilian government has large expansion plans for the power sector [49], which is reflected in subsidized credit from the government’s development bank (National Bank for Economic and Social Development [Banco Nacional do Desenvolvimento Econônico e Social (BNDES)]). As compared with an IRR calculated with existing subsidies, a hypothetical IRR calculated without subsidies would be much lower. The question of whether these subsidies can be excused from consideration depends on whether they are motivated by climate concerns, as the CDM Executive Board created an exception for policies that encourage less emissions-intensive technologies (‘type E-minus policies’ or ‘E- policies’) that were implemented after 11 November 2001 ([69]; Annex 3, paragraphs 6 and 7[b]). The subparagraph of the regulation specific to E-minus policies does not include the phrase ‘motivated by’ ([69]; Annex 3, paragraph 6), but in the original decision creating this category ([69]; Annex 3, paragraph 1) as one of four types, the term ‘motivated by’ is explicitly used to distinguish two of them from policies that are implemented for other reasons and are only ‘incidentally’ beneficial to climate change, making clear that climate motivation was the guiding factor in creating the exceptions. Executive Board decisions are supposed to reflect the intent of the Kyoto Protocol and the Protocol is clear in ruling out CDM projects for developments that would happen anyway. This makes the ‘motivated by’ criterion essential to maintaining the integrity of global efforts to mitigate climate change.

The Jirau PDD cites mention of hydropower’s benefits for climate in recent official documents as evidence that the BNDES dam subsidies are motivated by climate concerns, making them E-minus policies and, therefore, excused from inclusion in the baseline IRR. Brazil’s 2008 National Plan for Climate Change [48] is given as the justification for considering preferential financing for dams as motivated by climate concerns ([103]; pages 46–47). However, Brazil has been heavily subsidizing dams for many years before 2008 through a continually evolving series of measures. The practice also extends to long before the adoption of the Marrakesh Accords on 11 November 2001, which is the cutoff date for subsidies qualifying as E-minus policies [69]. Classification as an E-minus policy is supposed to mean that the policy, in this case BNDES subsidized financing for dams, is motivated by reducing emissions ([70]; Paragraph 1). The notion that the Brazilian Government’s support for its large dam-building program in Amazonia, including Jirau, is primarily motivated by concern for GHG emissions stretches the limits of this author’s credulity, but apparently not that of the consultants who drafted the PDD and the Official Assessment Report.

Whether or not the intent of the CDM has been reinterpreted over the years is of little importance, what is apparent is that dam projects are being approved for carbon credit on a large scale despite the fact that they are subsidized by most governments for motives unrelated to climate change, such as increasing energy independence and promoting industrialization. It is safe to say that most of these dams are being built by countries to increase their total supply of electricity, not to replace existing fossil-fuel plants. Because electricity from hydropower in Brazil is much cheaper than power from thermoelectric plants (provided the environmental and social costs of dams are ignored), expansion of hydropower would continue at the maximum rate practicable without recourse to a subsidy based on alleged climate benefits.

Hydropower receives favorable financing terms from BNDES. These include lower interest rates, partly from a 2007 change in BNDES policy to offer large hydropower projects rates calculated from a ‘basic spread’ of only 0.5%, versus 1.8% for fossil fuels ([103]; page 47). ‘Basic spread’ refers to the difference between a bank’s interest rates for lending and for deposits. It is ironic that the 0.5% ‘basic spread’ value charged to large hydroelectric projects is much lower than the 0.9% charged to wind power ([103]; page 47), raising doubt as to whether the generous concession to hydropower is really motivated by climate-mitigation concerns. The greater subsidy for hydropower as compared with wind is supposedly due to the greater interest burden for hydropower that results from a longer construction time; however, if wind power can be made viable with less subsidy from the government, then why not promote wind in preference to hydropower? Another bonus for hydropower is a provision in effect since 2006 for a 20-year amortization period, as compared with 14 years for gas-fired power plants ([103]; page 48).

Brazil’s offering of more favorable loan terms from its government development bank (BNDES) to hydropower as opposed to gas-fueled power plants ([103]; pages 45–47) is not motivated by concern for climate change alone. In 2006, the Evo Morales government in Bolivia took over Petrobrás facilities in Bolivia and sent shockwaves through the Brazilian government [71]. Brazil imported 50% of its gas from Bolivia in 2006, a percentage the country has now managed to reduce to 36% [72]. The Bolivian crisis motivated Brazil to institute a strong program to promote domestic energy production through hydropower development for a reason that is independent of any alleged benefit of hydropower for climate change. Of course, additional factors could also be favoring dam projects, since the construction industry represents one of the largest sources of contributions to political campaigns [73,74]. Brazil is currently engaged in a massive effort to build dams, with an average of one large dam to be completed in the country’s Legal Amazon region every 4 months for the next decade ([49]; page 285). The portrayal of the government’s subsidy for this as a selfless contribution to global efforts to combat climate change stretches the limits of credibility. In other words, the applicability of the CDM regulation allowing a benchmark IRR value to be used based on a hypothetical scenario without the subsidies from E-minus policies is open to question.

▪ Investor behavior & additionality

The simple fact that the Jirau Dam is already under construction is strong evidence that it is not additional in the sense intended by the Kyoto Protocol (as distinct from current CDM rules). Starting construction before CDM project submission does not violate current CDM rules, as retroactive crediting is allowed for initiatives that were set up as CDM projects. Nevertheless, the basic fact remains that the dam is under construction and the firms would not abandon the project in the absence of CDM funds. The PDD calculates an equity IRR of 9.7% after taxes and in real terms (i.e., after discounting inflation) under the project’s actual loan terms from BNDES (i.e., not under a hypothetical scenario) and without CDM credit ([103]; page 53). This is not a bad deal as compared with most investments. While an IRR of 9.7% was calculated for the case without CDM credit, the case with credit would raise the return to 16.7% ([103]; page 53). The difference represents a huge windfall profit for GDF Suez and could hardly be considered a wise use of mitigation funds.

It is possible that hydroelectric project developers may count on future carbon revenues if the risk of the CDM project failing to be approved is perceived as small. However, from the perspective of policy setting for the CDM and for mitigation in general, it would be unwise to assume that hydroelectric project decisions are based on carbon credits. This means that decisions by the CDM governing body should not ignore the most obvious physical manifestation of the dam project developers’ actual decision; namely that the dams are under construction with large investments before carbon credit is requested, let alone granted. It is the purpose of the Kyoto Protocol in combating climate change that is of importance, not the vast labyrinth of executive board decisions, guidances and clarifications that has been erected since.

‘Hot air’ & climate change

The amount of ‘hot air’ (carbon credit granted without a real climate benefit) to be generated by the project is large. As an annual average, emissions reductions claimed at Jirau are 6,180,620 t CO₂-e/yr, equivalent to 1.69 million t C/yr. Over the 7-year project, they will total 43.3 million t of CO₂-e or 11.8 million t of carbon, which is approximately equal to 1 year’s emissions for the city of São Paulo. This amount of carbon will be emitted somewhere else in the world, in the countries that purchase the CERs awarded by the CDM. Since the dam would have been built anyway, no real mitigation takes place to offset the emissions authorized by the credits. The Jirau project’s allowing a total of 43 million t of ‘hot air’ to be emitted over 7 years by the purchasing countries is a significant impact in and of itself; for comparison, all of the gasoline used in Brazil in 2005 emitted 39.1 million t of CO₂ ([75]; page 159). The global importance of carbon credit for hydropower is much greater, with 331 million t of CO₂-e in the CDM pipeline [102].

CER revenues from Jirau are expected to average R$250 million per year at the exchange rate on the 22 July 2008 project start date, or US$158 million per year ([103]; page 52) . Over 7 years, a total of US$1.11 billion would be siphoned off from real mitigation. Note that CER prices (in US$) declined by approximately 90% between June 2008 and December 2012, which would lower the 7-year total to US$110 million. However, as is the case for any commodity, carbon prices are governed by the equilibrium between supply and demand; it can be can be assumed that the demand (and consequently the price) would increase very substantially if the countries of the world commit themselves to major cuts in emissions. If climate change is to be contained, it will be necessary to make much greater cuts in global emissions than international negotiators have contemplated to date [76]. The measures needed will be very expensive, therefore it is important not to waste money intended for this purpose.

The initial paragraphs of the PDD describe GDF Suez as dedicated to ‘responsible growth’ and ‘respecting the environment’ ([103]; page 3). Unfortunately, in addition to the many other impacts of the dam [104], this author believes that the Jirau CDM project represents a blow to global efforts to contain climate change. The contrast between these impacts and the company statements dramatizes the need for basic reorientation, not only of the companies, but also of the Brazilian designated national authority, the CDM, the CDM Executive Board and the UNFCCC as a whole. The problems of the Jirau CDM proposal are, in essence, common to many other carbon proposals for dams. The lesson this implies is that funds for mitigating climate change should be used for other types of projects – not dams.

Future perspective

This article illustrates how carbon credit for hydroelectric projects can fail to be additional to that which would occur without funding from the Kyoto Protocol’s CDM. The author believes that the financial calculations in the PDD arguing that the dam is additional under CDM regulations instead indicate that the CDM’s current regulations are harming the climate and should be changed. Credit granted to dams that would be built anyway allows the countries purchasing the credit to emit GHGs without a corresponding real offset. Further damage is done by tropical hydroelectric dams emitting more GHGs than are recognized in CDM procedures. The problem of undeserved (nonadditional) credit for hydroelectric projects is sufficiently general that the only practical solution is to completely eliminate credit for hydropower under the CDM. The theoretical possibility of credit being denied to some rare individual project that is actually additional is insignificant as compared with the damage that hydropower credit is doing to global efforts to combat climate change.

The purpose of the CDM and of the climate convention as a whole is to help contain climate change. It is not to distribute subsidies that countries and corporations may have come to view as if they were entitlements. The logic of judging mitigation projects is different from, for example, a court of law. In a legal system, those on trial are assumed to be innocent until proven guilty beyond a reasonable doubt, and the corollary that accused criminals are frequently acquitted even though they are, in fact, guilty, is seen as a very reasonable price for society to pay as compared with the alternative of unjustly punishing some of those who are innocent. For climate mitigation it is the reverse, projects must be assumed to be nonadditional until they prove themselves to be additional. This author believes that the result that some truly additional projects are refused in no way justifies any other procedure. The same logic that applies to individual projects applies to whole classes of projects (such as dams); each class of projects needs to have a net effect that is additional, which can only be achieved if the approval of credit for nonadditional projects is very rare. In the case of hydroelectric dams, even though there may be some dams that are, in fact, additional, a system that approves large numbers of nonadditional dams has very grave consequences for the world as a whole and, therefore, this author believes that it should not be allowed to continue.

Table 1.  Jirau reservoir water management plans.

Month	Mean monthly streamflow (m^3/s)	2005 EIA water level (m above msl)	2012 PDD water level (m above msl)	Difference (m)
January	23,900	90.0	90.0	0.0
February	29,100	90.0	90.0	0.0
March	33,600	90.0	90.0	0.0
April	30,200	90.0	90.0	0.0
May	22,700	89.5	85.0	4.5
June	15,900	87.0	85.0	2.0
July	10,600	85.0	82.5	2.5
August	6800	83.0	82.5	0.5
September	5600	82.5	82.5	0.0
October	6800	83.0	82.5	0.5
November	10,400	85.0	85.0	0.0
December	10,600	87.5	85.0	2.5

EIA: Environmental impact assessment; msl: Mean sea level; PDD: Project Design Document.

‘Additionality’

The carbon benefit as calculated by subtracting the observed emissions from the baseline emissions. This difference is considered ‘additional’ to that which would have occurred without the mitigation project.

Internal rate of return

Highest discount rate at which an investment of capital would be considered profitable.

Project Design Document

A proposal for carbon credit submitted to the Clean Development Mechanism.

Power Density

The ratio of a dam’s installed capacity to its reservoir area, in watts per square meter (W/m²). How a reservoir area is defined is a matter of controversy.

Baseline

Hypothetical scenario for calculating future emissions without a mitigation project or activity. The emissions in this scenario will be compared with the actual emissions determined from monitoring the mitigation project.

Benchmark internal rate of return

The value used in order to establish that Clean Development Mechanism funds are needed to make the project profitable.

‘Hot air’

Carbon credit granted without a real climate benefit. The term derives from the English-language pun, where it implies both higher temperature from global warming and the chatter of meaningless talk.

Executive summary

Dams & carbon credit

▪ As of 1 December 2012, Clean Development Mechanism (CDM) carbon credit had been awarded to 586 hydroelectric projects totaling 112.7 million tons of CO2 equivalent. At the same time there were 896 dams that had been ‘approved’ (registered) but not yet awarded credit, plus 840 dams in the ‘pipeline’, awaiting completion of the approval process of the CDM. Very little of this carbon is additional to that which would have been emitted without the projects.

The Jirau dam

▪ The Jirau dam, which is nearing completion on the Madeira River, has been approved (registered) for carbon credit under the CDM. The proposal is similar to other CDM proposals for dams and serves as a warning of the inadvisability of carbon credit for hydroelectric dams in general.

Environmental & social impacts

▪ The dam has numerous environmental and social impacts that have provoked intense controversy in Brazil, as well as in Bolivia and Peru where impacts will also be felt. None of this is mentioned in the carbon credit proposal.

Greenhouse gas emissions

▪ Multiple studies indicate large emissions from tropical dams. A loophole in the CDM’s regulations allows dams with high power densities to claim zero emissions. The high water flow rate through the reservoir at Jirau will result in lower emissions than at other Amazonian dams, but emissions will not be zero.

Additionality & carbon credit

▪ The fact that the Jirau Dam is already under construction does not support the project’s claim of additionality, even though under CDM rules retroactive crediting is allowed for initiatives that were set up as CDM projects. The reason is that the Project Design Document does not convincingly demonstrate that the project could only be initiated due to the CDM crediting. CDM regulations allow projects to calculate an internal rate of return and compare it to a ‘benchmark’ internal rate of return value, in order to establish that CDM funds are needed to make the dam profitable. Figures are chosen in various ways to justify a scenario of dependence on the CDM.

‘Hot air’ & climate change

▪ As an annual average, emissions reductions claimed are equivalent to 1.69 million tons of carbon per year. Over the 7-year project, this totals 11.8 million tons (approximately 1 year’s emissions for the city of São Paulo). Since the dam would have been built anyway, no real mitigation takes place to offset the emissions authorized by the credits.

Acknowledgements

The author thanks A Pereira for assistance during a visit to the Jirau Dam site in August 2006, and the staff of Energia Sustentável do Brasil, SA and the Universidade Federal de Rondônia for information during visits to Porto Velho in January 2008, May 2008, and in August and November 2012. The Secretaria do Meio Ambiente do Porto Velho provided assistance during a visit to deforestation hotspots near Jirau in May 2012. The author also thanks the Energia Sustentável do Brasil staff, Estudo de Impacto Ambiental – Relatório de Impacto no Meio Ambiente authors, the Brazilian Institute for the Environment and Renewable Natural Resources officials, and the many academics and NGO personnel who contributed information and insights on the project over the years. The author especially thanks the local people for their assistance. PMLA Graça and the four anonymous reviewers provided useful comments.

Financial & competing interests disclosure

This research was supported by Conselho Nacional do Desenvolvimento Científico e Tecnológico (Brasília, Brazil: Proc. 305880/2007–1; 304020/2010–9; 573810/2008–7; 575853/2008–5) and Instituto Nacional de Pesquisas da Amazônia (Amazonas, Brazil: PRJ13.03). The author has no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.