Few will dispute that after air, water is the most essential element to life. One can live without food for days but no water means early death. The people of India have given to themselves a Constitution that guarantees the right to life, and based upon this fundamental right, the Water (Prevention and Control of Pollution) Act, 1974, serves to protect water and water sources for people. All potable water is drawn from rivers, lakes or aquifers and the cost involved is only that of extraction since water itself is "free". Water sources are often outside urban areas, and water is pumped to urban areas and purified and distributed through a piping system. Water that was hitherto free of charge is priced in urban areas to pay for the infrastructure that has to be constructed and maintained and operated to collect, convey, treat and distribute water. In rural areas, the cost of water is due to financial investment in the canal network or from consumption of electric power for IP sets.

Water needs to be "used where it falls" or transported in necessary quantities from one place to another. Availability of water is the ease or facility with which an individual has access to adequate water of the right quality. It depends upon rainfall in the area, population, water usage (domestic, agricultural, industrial), facility for water storage and treatment, and social ambience (e.g., denial of use of water source due to caste, community or religion). In the present circumstance of almost nation-wide water shortage and poor availability of water, the huge quantities of surface and/or ground water that are drawn for industrial use (mostly in urban areas) is at the expense of availability for domestic consumption. This is because the law grants individual rights to corporate bodies, which are very wealthy and influential, making it unfair on the poorer sections of society who are the vast majority of unempowered individuals.

The approach of planners has been to use capital-intensive and technology-intensive methods to supply huge quantities of water making water a marketable commodity, but without paying adequate attention to its distribution. Merely supplying huge quantities of water does not necessarily improve availability of water, and when supply is privatized, water becomes available only to those who can afford to pay, thus violating the fundamental right to life that the Constitution guarantees.

Fresh water supply from nature by rainfall is a distributed phenomenon and at the same time, water is essentially a decentralized requirement because millions of different users need water at millions of places in different quantities at different times for various uses. Centralizing water collection and storage, and then conveying the huge requirements of millions of different kinds of users over long distances and distributing it to them is less efficient in terms of finance, energy requirement and materials than a relatively decentralized system. Where large numbers of people live in close proximity as in urban areas, the present centralized systems are not able to cope and hence decentralizing to the extent possible by rooftop rainwater harvesting is coming into use.

However, water for agricultural purposes needs to be decentralized to avoid conveying huge quantities of water over long distances for intensive (flood) irrigation. This is quite apart from the fact that intensive irrigation has resulted in salination of the soil over the years. For example, about 20% of irrigated land in Punjab is salinated and unfit for cultivation, as also the entire Pakistani Punjab districts of Lyallpur and Sargodha. It has been conclusively demonstrated that adequate availability of both surface and ground water for agriculture is possible especially in arid and semi-arid areas by construction of check dams, each of which cost only a few lakhs or, with participatory labour of those who benefit from it, even less. There are successful working examples of such decentralized systems in various states all over India.

Interlinking of rivers and mass transfer of water represents major centralization of planning, finance, etc., and therefore benefits sectors that are centralized in nature, namely, urbanized industrial areas. The rural-agricultural sector being decentralized, it is capable of meeting its own needs of water by decentralized, low-cost means. This should particularly be viewed in light of the fact that the rural-agricultural sector uses 70% of total fresh water.

Interlinking of national rivers: options and comments
Interlinking or networking of rivers entails construction of dams and canals and other connected hydraulic engineering works for mass transfer of water across river basins. Basically, the scheme is to convey floodwater in the Ganga and Brahmaputra river basins to the arid and semi-arid areas of Rajasthan and Madhya Pradesh, and to the peninsular rivers of south India. To achieve this there are basically three "options". They are the "canal option" to construct lengthy canals, the "tunnel option" to convey water under mountains, and the "pumping option" to pump water over mountains. A combination of these options is possible, but such combinations would not be clear alternatives to the general solution of mass transfer of water across river basins.

One typical proposed scheme of mass transfer of water envisages construction of a network of contour canals 120 meters (m) wide and 10 m deep, with an aggregate length of 14,900 kilometers (km) to form navigable waterways at an estimated cost in 2002, of Rs.4 lakh crores:

(a) Himalayan waterways of 4,500 km length at 500 m contour connecting all tributaries of Ganga and Brahmaputra.
(b) Central Waterways of 5,750 km length at 300 m contour connecting southern tributaries of Ganga with Mahanadi, Narmada, Tapti, etc.
(c) Southern Waterways of 4,650 km length at 300 m contour connecting the Godavari, Krishna, Kaveri and some west flowing rivers of the western coastal strip.

The scheme with an annual impounding capacity of 15,000 thousand million cubic feet (15,000 TMC) offers benefits of year-round inland water transport by small ships (that consume only about 20% of fuel of road transport to save fuel oil), extension of irrigation to 150 million acres, and generation of 60,000 MW of electric power. It proposes to retrieve floodwater now "going waste" to the sea by impounding, transmitting and distributing it to water-scarce areas.

A canal that runs along a topographical contour necessarily has a level bed, and water will not flow in such a canal. But if there is even a slight bed slope in one direction, flow of water will be unidirectional, making one river basin always the donor and another always the recipient of water. This will surely not be always acceptable to the farmers in the donor river basin. Inter-State transfer of water of the same river is problematic enough, and inter-river transfer of water across States will not be easily resolved. Also the very slow rate of flow can never transfer the huge quantities of water envisaged. Further, generation of electric power from water that does not flow or flows only slowly, is impossible or uneconomical.

The canals will interfere with the natural drainage pattern of the land, causing rainfall run-off on the uphill or higher side of the canal to flow alongside the canal embankment and not pass over to the lower side. But if the uphill water is admitted into the canal, it will function as a catchwater drain and get filled with silt, reducing its capacity. With a bed slope less than that which will provide a natural "scouring" velocity of water flow, the canal will get silted up and call for expensive and recurring de-silting operations.

The canals will cause a sociological division between the uphill and the downhill sides since people will have to walk tens of kilometers to a bridge merely to cross over to the other side to graze livestock or for trade or social interactions. The cost of constructing bridges on existing rail and road routes (National Highways, State Highways and Main District Roads at the very least, neglecting Village Roads) will be enormous and will add to the cost of the project, affecting its financial viability apart from recurring maintenance costs.

By diverting substantial quantities of water from the higher reaches of a river (say 300 m elevation), there will be reduced water availability in the delta region where maximum agricultural operations are carried out, thereby causing reduction in total production and also causing permanent degradation of fertile land by sea water ingress. Particularly in the absence of a coherent agriculture policy concerning irrigated (water-intensive) or rain-fed crops, food crops or cash crops and assured prices and markets for agricultural produce, water diversion will cause destitution among many farmers.

The land required for 200 m wide canals with total length of 14,900 km will be 2,980 square kilometers (sq km). The excavated earth, not less than about 1,000 cu m per meter length for a 10 m deep canal, will have to be dumped on one or both sides of the canal, and the land to be acquired will be not less than 300 m wide. This will make the land requirement not less than 4,000 sq km. To this figure, add the land that will be lost due to inundation behind the numerous dams, and the total land required could be about 8,000 sq km. By present standards, acquisition of such a huge area of land will take decades if at all it can be done, and inevitably result in time and cost overruns in the project.

Apart from the longer-term ecological effects, the enormous environmental ill effects of change of land use of about 8,000 sq km of land in forest as well as inhabited areas can only be imagined. Admittedly, it is easy for Government of India and the State Governments to pass legislation to take this project out of the purview of the Environment Protection Act, 1986, that mandates Environmental Impact Assessment, but this will be extremely short-sighted because it will not in any manner lessen the impact.

The cost of acquisition of 8,000 sq km of land besides land for re-settlement/rehabilitation also needs to be considered. Madhya Pradesh has already stated that there is no land available for resettlement of the oustees of Sardar Sarovar Project who are still to be resettled. Other States are not very different in respect of land availability - the Kabini dam (Karnataka) oustees of 25 years ago still await compensation while the oustees of the Damodar Valley Project languish after nearly 50 years. Considering exceedingly poor past performance by various states in resettlement and rehabilitation of oustees for huge dam projects or paying compensation, the current mammoth project has the potential to dwarf all previous records of human rights violations. Severe resistance to the project is sure to show up in many places, causing more delay and uncertainty, both of which will render planning and management of finance, machines and materials extremely difficult and tenuous, to say the least.

Of course, it may be possible for Governments to force their way to complete projects with the help of Police force, but the social effects of doing this are already apparent almost all over India with rising disaffection among the rural and tribal populations. As Kancha Illiya says, theirs is today not a cry for pity but a scream of anger. Such reaction is a concomitant of centralized planning without consulting the people.

The operation and maintenance of the canals (de-silting, repair, etc.) and the appurtenant structures and facilities (lock gates, power generation plants, buildings to house services and personnel) and their security is likely to cost at least 5% annually, that is Rs.20,000 crores, apart from establishment costs. Compare this with the national annual revenue from direct taxes - about Rs.90,000 crores. The economic feasibility of the project not only in terms of initial cost but also operational and maintenance costs needs to be carefully examined.

Pumping for networking river basins
Water from the Ganga-Brahmaputra basin can be conveyed to peninsular India either by pumping it over the Vindhya mountain range or passing it through large tunnels through the mountains. These options have been considered earlier at Government level and rejected as being financially or technically not feasible. Nevertheless, the magnitude of the problems and the practicability of the options are mentioned briefly. From any school atlas it may be verified that the altitude of the Ganga-Brahmaputra flood plains is about 100 metres (m) above MSL, the Vindhya mountain range that separates the Ganga basin from peninsular India is about 300 m above MSL, and the Deccan plateau is at about 250 m above MSL. The electric power required to pump such huge quantities of water at usable rates over such heights is very close to that presently generated nation-wide, and the pumping option is therefore impractical, besides the environmental and cost effects of such power generation. Cost-wise, tunnelling for tens if not hundreds of kilometers is not going to be very much more afforable than the pumping options.

What needs to be considered is whether existing incomplete river projects (Rs.80,000 crores more is needed as on today to complete them) are to be abandoned and if so, how undelivered benefits of the projects can offset the displacement of people and financial expenditure that has already taken place. The interlinking scheme itself is very likely to be subject to time and cost overruns besides shortfalls in quality that lead to poor serviceability and increased expenditure for both rectification and maintenance. What if the financial costs mount up, work does not progress and returns from the project are delayed? What if land is acquired and not used (for whatever reason) for the purpose? Would the people whose land is acquired become encroachers? These questions need to be asked and answered considering the track record of financial, technical and administrative management of major projects all over India.

Next: Alternative to interlinking rivers and Cost-Benefit Analyses