Compiled by Patrick McCully of International Rivers Network from a variety of Indian government, World Bank, Narmada Bachao Andolan and independent sources. <&smiley;>
Hard copy versions are available for $4 from the address at the end of the report (checks payable to International Rivers Network).
Construction in 1994
World Bank Involvement
How to Meet Gujarat's Water Needs
Submergence and Displacement
Position of the Madhya Pradesh Government
Indian Review Committee
The first recorded proposal for damming the Narmada River and diverting its water to irrigate crops in Gujarat was made by a British entrepreneur in 1863. The first serious study of the development of the whole basin began in 1947. After independence these investigations were taken up by various government committees which proposed numerous dams on the Narmada and its tributaries. The first proposal for a dam at the Sardar Sarovar site was made in 1959 and preliminary construction began in 1961. Disagreements between the states through which the Narmada flows about how to share its water, however, led to the project being suspended. In 1969 the Narmada Water Disputes Tribunal (NWDT) was set up to decide the inter-state allocation of water and the costs of the dams and other infrastructure needed to exploit the river. The tribunal's decision or 'Award' was made in 1979. Full-scale construction of Sardar Sarovar began in 1987.
The Sardar Sarovar Dam is on the Narmada River in Gujarat state, 170 kilometres (106 miles) upstream from where the river flows into the Gulf of Khambhat in the Arabian Sea. The Narmada is the largest westward flowing river in India. A few kilometres downstream from the dam site on the north bank is Kevadia Colony, the town built to house the construction workers and related bureaucracy. Vadgam, the first village behind the dam, starts around one kilometre from the dam site and stretches out for several more kilometres along the north bank. About 15 km upstream on the south bank a small tributary running into the Narmada forms the Gujarat-Maharashtra border. On the eastern (Maharashtra) side of the creek is the village of Manibeli, a focus of resistance to the project where the Narmada Bachao Andolan (NBA -- Save the Narmada Movement) maintains an office.
The dam is a 1210 m (3970 feet) long wall of concrete across the valley. It is designed to impound a reservoir with a full level of 139 m (455 feet) above sea level (asl). The middle section of the dam is planned to reach a height of 146.5 m (481 feet) asl. The bed of the river at the dam site is at 17 m (56 feet) asl so the planned height of the dam above the river bed is 129.5 m (425 feet).
The main canal leading from the reservoir is scheduled to be 460 km (286 miles) long, eventually reaching the state of Rajasthan. It is 250 m (820 feet) wide at its head near the dam and planned to be 100 m (328 feet) wide at the Rajasthan border. A network of secondary canals totalling 75,000 km (46,600 miles) in length is planned to deliver the irrigation water to farmers. Large electric-powered pumping stations will need to be built to deliver water to the Saurashtra and Kutch branches of the canal system. A large powerhouse containing turbines and related machinery is being built at the dam and a smaller one at the head of the canal. A weir is to be built at Garudeshwar, around 16 kilometres downstream of the dam, with a capacity to store six hours of the maximum flow through the Sardar Sarovar turbines. This water can be pumped back into the reservoir at times of low daily electricity demand and then released through the turbines again to generate electricity at times of peak demand.
To oversee the implementation of the dams on the Narmada the NWDT set up the Narmada Control Authority (NCA) composed of senior representatives of the governments of Gujarat, Maharashtra, Madhya Pradesh (MP) and Rajasthan, and chaired by the Water Resources Secretary (the top civil servant in the central government Ministry of Water Resources). The NCA has established Environment and Rehabilitation Sub-Groups, chaired respectively by the Secretaries of the central government Ministry of Environment and Forests (MoEF) and the Ministry of Welfare. SSP is being implemented by the Sardar Sarovar Narmada Nigam Ltd (SSNNL or 'the Nigam'), a corporation wholly owned by the Government of Gujarat (GoG). The construction of the dam is contracted to Jay Prakash (J.P.) Associates, who have a virtual monopoly over major dam projects in India, and the construction of the canals to a number of smaller contractors.
Completion of the dam is scheduled for 1997. The canal network will not be finished until 2025 at the earliest. The annual construction schedule varies between different official sources. At the start of the monsoon in June 1993 - the end of the 1992/93 construction season - the lowest blocks of the dam were at 61 m asl (the sides of the dam are much higher). The lowest blocks are currently (mid-May, 1994) at 69-70 m with the rest of the middle section of the dam at 80 m.
According to a Supreme Court decision of August 1990, 'oustees' (the people to be displaced by the reservoir) should be properly resettled at least six months before submergence of their homes or lands. Concerns over the slow pace of resettlement and unfinished environmental studies led the NCA Environment Sub-Group to recommend in late 1993 that the dam should not be built above 67 m asl during the 1993/94 construction season and that the temporary construction sluice gates at the foot of the dam should not be closed. The NCA accepted this recommendation. The sluice gates allowed the river - outside of the monsoon - to flow under the dam and prevented any permanent impoundment of water. During the monsoon (generally early June to late August), the small temporary sluices could pass only a small fraction of the swollen river and the rest of the water would back up and flow over the top of the dam.
At the beginning of January 1994, work on raising the dam was suspended with the height of the lowest blocks already above the NCA's height limit, at around 69 m. On January 11, a statement issued after a meeting in New Delhi between Prime Minister Narasimha Rao and the Chief Ministers of Gujarat, Maharashtra and MP, said that there would be no further construction above 67m unless people were rehabilitated 'well ahead of time' (the statement falsely claimed that the dam was then at 61 m). On February 23, however, the dam authorities without notice closed the steel shutters on the ten temporary sluice gates and started to raise the dam wall again.
The day after the sluices were closed the Gujarat High Court stayed any further work on closure of the sluices, preventing the authorities from permanently filling the sluices with concrete. The authorities told the High Court that the chains on the steel shutters had been cut making the closure irreversible. Within two weeks of the closure the river rose to the height of the next set of openings in the dam, the 'river sluices' at 53 m asl. The level of the water behind the dam is now around 60 m. Houses and fields in four villages have been flooded and access roads cut off.
The number of families to be affected by submergence during the coming monsoon will depend upon how much further the dam is raised and how much rain falls. There are no gates in the dam which can be opened to let flood waters through. Official estimates of the number at risk during each monsoon are based on calculations of the height of water in a 1 in 100 year flood. The Government of Maharashtra (GoM) has cited figures for the number of families at risk in the state this year which range from 148 to over 2000. The NBA says that 500 families in Maharashtra are at risk of losing their houses and many more will lose crops. GoG claims that all its oustees have been resettled: the NBA says that 400-500 families in Gujarat are still in areas at risk this year. The Government of MP (GoMP) now says 15 villages are at risk, three of which could be totally submerged. In total, the NBA believes that at least 40 villages are at risk in the three states.
Shripad Dharmadhikary of the NBA wrote in early May about what is likely to happen during this monsoon:
"Most families would be stranded . . . or dumped by the government at inadequate, ill-prepared 'resettlement' sites . . . Many families would face a worse situation as the rising waters would fill up the numerous streams and gullies, cutting off access roads, and slowly turning the undulating region into a series of isolated islands . . . The houses could be marooned for as long as 4-6 months."
The families threatened are not only those who have refused to move because of their opposition to SSP, but also many who have accepted that the dam will be built and have asked for resettlement but who have not been given anywhere to go.
The World Bank agreed to lend $450 million for SSP in 1985. After years of criticism the Bank in 1991 commissioned a team of four independent experts to review the resettlement and environment components of the project. The Independent Review was chaired by an ex-head of the UN Development Programme, Bradford Morse. His deputy was Thomas Berger, a Canadian lawyer known for his work on human rights and environmental issues. Their report, released in June 1992, strongly criticised the project and the World Bank's involvement in it, concluding that:
". . . the Sardar Sarovar Projects as they stand are flawed, that resettlement and rehabilitation of all those displaced by the Projects is not possible under prevailing circumstances, and that the environmental impacts of the Projects have not been properly considered or adequately addressed. Moreover, we believe that the Bank shares responsibility with the borrower for the situation that has developed."
The international pressure on the World Bank to withdraw built up over the following months. Finally the Bank realized that the damaging publicity SSP was creating was going to get worse and that it needed to find a face-saving method of extricating itself from the project. It decided that the best course of action for both itself and the Indian government was for GoI to request the Bank to pull out. A deal was made and on March 30, 1993, India formally requested the Bank to cancel the $170 million remaining to be disbursed for the project. Around half the money spent on the project so far has come from the World Bank.
On the same day as the loan was cancelled the Bank's General Counsel, Mr Ibrahim Shihata, wrote a memorandum reminding the South Asia Department that notwithstanding the cancellation all the provisions of its 1985 loan agreements were still in place. The agreements contain several conditions on the resettlement and rehabilitation of the people to be displaced. The World Bank is thus still legally bound to ensure that the project authorities comply with these provisions. These conditions are being widely and comprehensively broken. Non-project specific loans from the World Bank to India may still be helping to fund SSP.
SSP's backers claim the project will irrigate a 'command area' of 1.8 million hectares (4.45m acres) in Gujarat and 75,000 hectares (185,000 acres) in Rajasthan; have an installed power generation capacity of 1450 megawatts; provide domestic water to over 2.35 million people in 8235 villages and 135 towns in Gujarat; and prevent flooding downstream.
"Irrigation efficiency in India has often been assumed at 60%, whereas a worldwide sample of irrigation commands indicates 37-40% efficiency in areas of low rainfall under reasonably good management, and in higher rainfall zones, an average of 23%. Most irrigation commands in India probably have an irrigation efficiency of 20-35%. If assumed efficiency is 60% and actual efficiency is 30%, actual water availability will be half the assumption at design."
According to the NWDT, Gujarat will get only 16% of the power from SSP, the rest being split between Maharashtra (27%) and MP (57%). Pumping water to the Kutch and Saurashtra branch canals would consume around 70 megawatts after allowing for the small amounts of power generated by turbines in the canal system. Large amounts of power would also be required to pump groundwater into the canals, an integral part of the irrigation plans; to drain the command area soils; and to operate the gates and other structures regulating the flow in the canals.
There are no firm estimates for the total financial cost of SSP. In 1983 the project authorities' submission to the World Bank estimated the cost at 42,040 million rupees in 1981-82 prices, including the canals but not the infrastructure for supplying drinking water. In 1985 the World Bank estimated the cost as Rs.136,400 million. In 1991, GoG revised its estimate upwards to Rs.90,000 million. In 1993 it was revealed that this figure does not include interest of over Rs.17,000 million. The NBA's detailed analysis estimates the total project cost at current prices including canals and water distribution as at least Rs.250,000 million ($8,300 million). A 1994 World Bank publication cited the project cost as $11,400 million (Rs.342,000 million).
GoG has unsuccessfully attempted to privatize the hydropower component of the project in the last few years. The privatization prospectus put the cost of the power component at Rs. 29,000 million - eleven times greater than the cost cited in the 1979 NWDT Award. The termination of Japanese and World Bank aid, and the large arrears in the mandatory financial contributions from Maharashtra and MP are putting further pressure on Gujarat.
In 1985 the World Bank calculated the values of different parameters at which the project's net financial benefit would become zero. Some of these are:
These are single parameter values, all of which will clearly be exceeded. Taking the combined effect of the changes in the parameters, the financial cost of the project must vastly exceed its potential financial benefit.
As the project is clearly not going to perform as claimed, what are needed are not 'alternatives' to SSP, but ways of solving Gujarat's water crisis, and especially the water shortages in Kutch and Saurashtra. Several plans have been developed by engineers and economists (with a minuscule fraction of the resources put into planning SSP) which show how Gujarat could fulfil the promised benefits of SSP without its massive financial, human, and environmental costs, and much more quickly. GoG's own water agencies have stated that it is possible to deliver water to Kutch and Saurashtra much more cheaply and quickly than could be possible with SSP.
Ashvin Shah, a Gujarati working with the American Society of Civil Engineers, notes that SSP is based on outmoded 1950s ideas of water development, and its planning has failed to benefit from the past four decades of experience with irrigation and water conservation schemes in India and elsewhere. With the large scale implementation of decentralized, small rainwater harvesting schemes, claims Shah, 21 MAF of rainwater could be collected within Gujarat each year, 50% more than the amount of water supposed to be made available by SSP. Shah's plan to solve Gujarat's water crisis is based on a more equitable sharing of the available water, water harvesting, water conservation, making existing water supply and irrigation schemes more efficient, the restoration of degraded watershed vegetation, and making farming practices less water and energy intensive. He also advocates the exploitation of Gujarat's wind, solar, tidal and biomass energy resources.
The NBA believe that over one million people will lose land or be otherwise severely affected by the various components of the project. As comprehensive surveys have not been completed by the dam authorities the following figures are all estimates.
The resettlement package differs between the three states. PAPs from Gujarat, or those from Maharashtra and MP willing to move to Gujarat, are eligible for a minimum of two hectares (five acres) of irrigable land in the command area of the project as well as house sites and some cash compensation. Major sons (those over 18) and landless families (many of the tribal families cultivate land to which they have no legal title) are also eligible for two hectares of land under Gujarat's policy. 'Landless' oustees and major sons settling in Maharashtra receive only one hectare; those in MP are not eligible for any land.
The authorities claim that around 7000 PAPs have been resettled in Gujarat and Maharashtra. No oustees have been resettled in MP. Those who have been resettled face a multitude of hardships and many have returned to their original villages. The stress and impoverishment caused by resettlement has increased death rates among the oustees, especially of children. The problems, which have been extensively noted by the official resettlement monitoring agencies and the World Bank's Independent Review include:
GoG has acquired under 14,000 hectares of land for the PAPs who have been resettled in the state, spread over approximately 400 different resettlement sites. There are no plans available describing where land will be found for the remaining 13,000 oustees expected to move to the state. The acquisition of such large areas of land combined with land speculation due to SSP has greatly increased land prices in the command area, inflating the cost to the government of acquiring land, and encouraging the government to buy land of increasingly inferior quality. GoM has only after great difficulty and controversy persuaded the MoEF to release 4,300 hectares of forest land for resettlement in Maharashtra. This is only just over half the land needed for oustees in the state. GoMP admits that it is unable to acquire any agricultural land for resettlement.
An important recent development has been that the new government in Madhya Pradesh, elected in November 1993, has admitted that MP cannot resettle the huge numbers of people in the state to be displaced by SSP as currently planned. A GoMP note circulated at an all-party meeting called by the MP Chief Minister Digvijay Singh in February 1994, recommended that the full reservoir level should be reduced from 455 feet (139 m) to 436 feet (133 m) asl which would spare from submergence over 8,100 hectares (20,000 acres) of land and 38,000 people in MP alone. The NWDT decided on the 455 feet level rather than 436 feet solely on the basis of the extra power it would provide. No water supply benefits would be lost by reducing the height. GoMP engineers have now calculated that as the Narmada flow is 17% less than the NWDT assumed, the reservoir level could be reduced to 422 feet and still supply the same amount of water to the Gujarat canals as a reservoir at 455 feet. Digvijay Singh (an engineer by training) wrote to Prime Minister Narasimha Rao in March 1994, saying that there was consensus among MP's political parties that the dam height should be reduced. GoMP has also said that it is willing to forego its 57% share of the electricity produced by SSP.
In August 1993 the central government set up a committee to 'look into all aspects of SSP'. The committee heard submissions from GoI, GoMP, GoM, the NBA, affected people and independent experts. GoG boycotted the review although Gujaratis close to the state government, including Mr C.C. Patel, an ex-Chairman of the Narmada Nigam who has been one of the key promoters of the project, did give evidence. The committee was originally due to submit its report to GoI by December 1993 but the date has been continually moved forward and it has still not been submitted. GoG has launched a legal action against the review in the Gujarat High Court, on the grounds that it is illegal under the terms of the NWDT Award. The committee's report cannot be submitted until the case goes to the High Court, which, under political pressure, has still not set a date for the hearing.