Dams Lite?
Published in World Rivers Review August 2001
Run-of-River Projects No Panacea
Are huge reservoirs a thing of the past? Today, a significant number of the world's proposed dams are being billed as "run of river," which feature smaller reservoirs and, presumably, smaller impacts. Can we believe the industry hype that run-of-river schemes leave rivers in a relatively natural state, or are such projects just “engineering gimmicks” as former WCD commissioner Ted Scudder has called them?
Run-of-river dams are marketed a bit like low-tar cigarettes or "lite" beer. They promise electricity, but with far fewer messy side effects. Even the term “run-of-river” evokes images of free flowing, cascades of water rushing unhindered to the sea. It seems to imply that the river is in charge, unaltered and uncontrolled. In fact, dam proponents often use the terms “run-of-river” and “minimal social and environmental impacts” interchangeably when promoting a project. Run-of-river projects can, however, have large dam walls, significant social and environmental impacts, and big price tags. In fact, although generally run-of-river means a smaller reservoir and more natural flows, the term is not clearly defined.
Most large dams are storage dams, meaning they store water during the wet season and release it during the dry season, or hold water when electricity demand is low and release it when it is high. Run-of-river projects, on the other hand, are theoretically designed to ensure that the amount of water flowing into them always equals the amount of water flowing out of them (even though many of them store water for hours, days, or even weeks at a time).
Proponents of run-of-river projects frequently claim they do not require a reservoir. In reality, all run-of-river dams impound water, either to create “head” for hydroelectricity generation, to raise water levels to the height of diversion intakes on water projects or to create enough depth for navigation. According to Patrick McCully, “In theory, no dam with gates should be considered run-of-river because closing and opening gates implies storing and releasing water. In practice, dams with some limited – but undefined – storage capacity are referred to as run-of-river. It's a sort of Alice in Wonderland, ‘it means exactly what I want it to mean’ term.” Thus, the 113-meter-high Pangue Dam in Chile qualifies as run-of-river, as does the 62-meter Bonneville Dam in Oregon, or the 1.45-kilometer-wide Cana Brava Dam in Brazil.
A key selling point of run-of-river dams is the claim that their "head ponds" are small in comparison to storage reservoirs. Large reservoirs result in physical and economic displacement, which adds considerably to project costs and often incites public opposition. Such opposition can result in costly delays and even project cancellations – all major concerns of the private investors and guarantee agencies who are playing a growing role in the damming of the world's rivers. Head ponds, on the other hand, are portrayed as small and innocuous – they're just ponds, the term seems to imply.
Recent dams described as run-of-river, however, blur the distinction between head pond and reservoir. The "head pond" behind Maheshwar Dam in India's Narmada Valley would force some 35,000 people from their homes. In Mozambique, the proposed Mepanda-Uncua Dam’s head pond would cover more than 100 square kilometers. In 1994, the Mekong River Secretariat proposed the construction of a cascade of nine “run-of-river” hydropower projects along the Mekong’s mainstem that would have inundated 1,000 square kilometers and displaced more than 60,000 people. Although the scheme would have turned the Mekong into a staircase of reservoirs, the project's documents never once used the words "dam" or "reservoir."
Though run-of-river projects are smaller than most storage projects, they can still have major impacts. As Shripad Dharmadhikary points out, “While there is some correlation between the size and the impact of a project, this is by no means a fixed equation.” For example, the proposed Bujagali Dam on the Nile River in Uganda is being promoted as run-of-river and would have only a 4.2km2 head pond, but it will deprive thousands of local people of their livelihoods, possibly drive some rare fish species to extinction, destroy a culturally important waterfall and wipe out a $4 million per year rafting industry. And the impacts of the tiny 3MW Babilonia HEP run-of-river dam in Honduras seem well out of proportion to its size (see article on page 10). More than 100 families will lose their coffee plantations to the head pond, a 1,500-foot waterfall will be destroyed, and communities living downstream will no longer be able to utilize the river through gravity-fed water systems.
Underpowered, Over-Hyped
One of the most fundamental problems with run-of-river projects is that they promise more electricity than they can actually deliver. The final report of the World Commission on Dams states, “Power benefits of run-of-the river hydropower plants may be much lower than could be expected from their installed capacity.” Because the electricity output of run-of-river projects is dependent on the amount of water flowing in the river at any given time, their power output may be insufficient and unpredictable, which may render them economically infeasible. During the rainy season, or when a river is in flood, run-of-river dams frequently cannot use all the water that flows through them to produce electricity. During the dry season, the runoff in the basin is often far too low to produce electricity at full capacity.
The Pak Mun Dam in Thailand exhibits both of these problems. It produces only 40MW of its 136MW capacity during the high demand period in April-May. In very wet years, it faces another obstacle to power generation. During some years, the Mekong River floods so severely that the river comes up to the toe of the dam. The elevated water levels at the dam’s base mean that there is not enough head pressure, lowering power generation, and even occasionally forcing the power plant to be closed.
Because run-of-river dams’ electricity production is based on the river’s flow rather than a reservoir's releases, construction of a stand-alone run-of-river dam may set off a spate of dam building upstream in order to maximize the efficiency of the run-of-river project. This occurred in Chile, where construction of the giant Ralco storage dam on the Biobío River was justified on the grounds that the run-of-river Pangue Dam downstream would consequently produce more electricity. Although Pangue was designed with Ralco in mind, project proponents did not publicly present the dams as a pair, thus smoothing the way for an easier approval process at the World Bank.
Environmental Impacts
Run-of-river dams’ impact on the environment may not be as great as that of storage dams, but they do indeed have environmental consequences, sometimes serious ones. The havoc wreaked on delicate gorge ecosystems by the Lower Kihansi Hydropower Project in Tanzania is a prime example of how run-of-river project can result in a loss in species diversity. Because the project diverted water from the 800-meter Kihansi waterfall, the endemic Kihansi Spray Toad, which survives on mist from the falls, has been brought to the verge of extinction. The river's wetlands are drying up, threatening rare plants. The US$275 million project also directly affected about 22,000 villagers.
Plant diversity is typically reduced along run-of-river head ponds, albeit less than along the banks of storage reservoirs. Researchers in Sweden noted that river reaches upstream of run-of-river projects hosted 15 percent fewer species after damming “because regulated shorelines were narrower than natural riparian zones.”
Fisheries are harmed just as easily by run-of-river projects as they are by storage dams. In both cases, the dam wall presents a barrier to migrating fish.
The Pak Mun Dam has wiped out 56 species of fish from the Mun River because the dam blocks their rainy season migration routes, the 60km2 head pond inundated their spawning grounds, and the species’ inability to jump has rendered the project's ill-conceived fish pass of little use. Some 100 additional species in the Mun were negatively affected by the dam’s construction, and the fish catch directly upstream of the dam has declined 60-80% since the project’s completion, resulting in an economic loss to villagers of about $1.4 million per year, according to the WCD. Affected villagers successfully lobbied their government to open the gates of the dam this year, which brought immediate improvements to their fish catches.
Other run of river schemes with serious fisheries impacts include the Theun-Hinboun Dam in Laos, which resulted in a serious reduction of local communities’ fish catch after it was constructed, and the proposed Bujagali Dam in Uganda, which will inundate a falls that harbors rare endemic fish species. The Ugandan government approved Bujagali contingent on the project including a fish pass. However, the project's Panel of Experts writes, "The Panel knows of no cases anywhere in Africa where fish ladders have been used successfully."
Clearly, run-of-river projects can have serious impacts, and must be given the same scrutiny as other dams. Run-of-river dams may be less destructive than storage dams, but they can also be just as capable of pushing species to extinction or families further into poverty. As with other large dams, run-of-river projects should be subject to review against the World Commission on Dams' guidelines.
A Run-of-River Sampler
The following are existing or proposed run-of-river projects whose impacts appear to outweigh their benefits.
Maheshwar Dam, India (Height: 35 meters)
This dam would displace 35,000 people and destroy the livelihoods of communities upstream. Rates of compensation are undervalued, and critics estimate that if compensation were tied to full replacement value it would render the project economically unviable.
Xeset Dam, Laos (10 meters)
This dam generates less than two thirds of its installed capacity and produces virtually no electricity during extended periods of the dry season. Driving output down further is the government’s commitment to maintain downstream fisheries by releasing water downstream during the dry season rather than diverting it.
Theun Hinboun Dam, Laos (25 meters)
In addition to the impacts on fisheries described in the article, this dam is generating far less electricity than promised.
Popa Falls Dam, Namibia (8 meters)
This recently revived 30MW scheme is located on the Okavango River, the source of the delicate Okavango Delta. During part of the year, it will leave the popular tourist destination at Popa Falls virtually without water. The 2.9km2 head pond will require the relocation of some 90 residences and increase evaporation by 1.5 million cubic meters annually, an impact that has unknown consequences for the fragile Okavango Delta ecosystem.
Cana Brava Dam, Brazil (66 meters)
This 1,450-meter long dam on the Tocantins River will, if completed, displace more than 1,000 families. Recent demonstrations concerning resettlement and compensation terms sparked confrontations with hundreds of police officers (see p. 9).
Bonneville Dam, USA (62 meters)
Constructed on the Columbia River in 1935, Bonneville seriously affected thriving salmon populations and other wildlife in the Columbia Basin. The dam also inundated a number of cultural sites important to Native Americans.
Chalillo Dam, Belize (30 meters)
The head pond behind this dam will inundate 1,000 hectares of riverine habitat, which is, according to some scientists, the only habitat of its type in all of Central America. The dam would further threaten the endangered Scarlet Macaw and other rare species.
Pangue Dam, Chile (113 meters)
This massive project displaced some 1,300 people, most of them from the indigenous Pehuenche group. Furthermore, the dam was designed such that only the completion of a storage dam upstream would maximize its efficiency.
