THE KOSI PROJECT: A Note on Myopia and Mismanagement

Contrary to popular Nepali perception, the Kosi Project may have proved to be a nightmare for those who live within the embankment area.

The Kosi floods of July 1991 have again highlighted the lack of consensus in Nepal and India over issues of common concern. In Nepal, news reports that a 400 metre breach in the western embankment at Hanuman Nagar had displaced 50,000 people in 12 villages in Saptari district, was followed with veiled insinuations that India was responsible for the calamity, and should, therefore, deal with it. Interestingly, the Nepali papers did not mention that 150,000 people in Saharsa and Madhubani districts of Bihar had also lost their homes in the same deluge.

The Kosi has been a focus of geo-political posturing by the two countries. India has maintained that Nepal has stymied all attempts at an exhaustive solution of the problems, while Nepal has argued that India has creamed away the benefits of the Kosi Project to the detriment of Nepal. It is our contention that the floods this year point to a deep-rooted malaise which will deteriorate rapidly if it is not addressed urgently. Focussing on an inter-related set of issues, this note aims at dispelling some of the prevailing misconceptions about the supposed 'benefits' of the Kosi Project to India, in the hope that it might promote a more comprehensive understanding of the issues.

HYDRO-GEOLOGY

The Kosi is 730 km long and originates in the mountains of Nepal and Tibet. It drains an area of 86,900 sq km between longitude 85'20' and 88 degrees, and has the largest catchment among the rivers of India measuring 61,440 sq km. Flowing through a gorge at Chatra, it enters Saharsa district of Bihar at Hanuman Nagar and almost immediately assumes an unstable deltaic form made up of intertwining and dividing streams, separated by shoals and bars. Further south, it is joined from the west by the Kamala Balan, the Bagmati, the Soni, the Bhati Balm, the Tiljuga, and the Sugarwe. The combined waters of the Kosi collect in a saucer-shaped depression near Kursela township and enter the Ganga at Karagola.

In the last 125 years, the Kosi has shifted westward by more than 115 km from its position just east of Pumea township, and has destroyed 8,000 sq km of land in Purnea, Saharsa, Darbhanga, Bhagalpur, and Monghyr districts. The rate of its westward drift has not been uniform. Between 1736 and 1922, it moved at the rate of one mile in four years. Since 1922, this movement has accelerated to over one mile per year. The 1961 Census of India had estimated that the Kosi has uprooted 6.5 million people and has inflicted a continuous annual damage of IRs 100 million.

There are several reasons for the instability of the Kosi. First, a flattening of the bed south of Chatra results in a drop in current velocity and despite a significant increase in depth, the river begins to meander. Second, it carries the largest sediment load among the rivers of the world amounting to 7,308 tons per square mile of its catchment most of which is deposited on the flood-plain. The consequent rise in the level of the river-bed forces the river to realign itself continually. Third, the seismicity of the region has enhanced the instability of the Kosi and contributed to its sediment load. Fourth, the forces of rotational deflection whereby surface current movements are influenced by the earth's axial rotation have facilitated the river's general westward tendency. Fifth, it has an unparalleled rate of erosion. In the period 1951 to 1953, the river eroded 10,000 feet of land in India between Belka and Hanuman Nagar. In its westward swing the Kosi has sabotaged its banks despite protective spurs at the vulnerable convex bends. Sixth, Glacial Lake Outburst Floods [GLOFs] in the mountain catchments of the Kosi may have unpredictable downstream consequences. A GLOF in Taplejung in 1980 resulted in floods 35 km south of Nepal which smashed through a spur, threatened the eastern embankment and the Bihar town of Supaul. Seventh, land degradation in the hilly catchments is suspected to have contributed to augmenting the sediment in the river.

FLOOD-CONTROL AND REHABILITATION

Although the Kosi had evoked serious official concern in India in the 19th century, it was only after Independence in 1947 that a flood-control project acquired concrete shape. In 1953, following a personal assessment of the Kosi floods by Prime Minister Jawaharlal Nehru, a three phase flood-control, irrigation, and diversion scheme was designed. Phase I consisted of a 3,770 ft barrage across the Kosi three miles north of Hanuman Nagar. On the east and the west, earthen dams measuring 6,220 ft and 12,800 ft respectively with afflux bunds 8 miles long were planned. Phase II envisaged two embankments running southward from the barrage for a length of 75 miles on the western side with ring bunds at vulnerable points, and embankments at the Tiljuga and Kamala Balan marginals. An eastern embankment 62 miles long, and a protective bank of 12 miles above the eastern afflux was recommended. Phase III consisted of a canals on the east to irrigate 0525 M hectares. A plan was also drawn up to rehabilitate the population inhabiting the region that would be embanked.

The Project was completed in the early 1970's and instead of ushering in prosperity has proved a nightmare for India in every possible way. The relocation scheme for 300 affected villages covering an area of 260,000 acres and inhabited by 115,000 people failed due to a lack of imagination, poor administration, and the sheer physical immensity of the task, and most of the rehabilitees returned to their original homes within two years. Today, there ate about 600,000 people in 321 villages within the embankments who live under the constant threat of floods.

Of the total sum of IRs 21.2 million allocated for rehabilitation in 1957, only IRs 3.69 million was received by the rehabilitees. Little is known about the balance. In 1978, the former Bihar Chief Minister Jagganath Mishra sum med up this misadventure by remarking that the only tangible impact of the Project was that it had exposed a population of 400,000 people in Saharsa to regular floods. Interestingly, the rehabilitation has been treated as a fair accompli, and in no official document is the existence of the embanked population acknowledged. On the other hand, the state does not fail to collect revenue from the embanked population for even barren and unculturable wastelands.

UNMITIGATED CORRUPTION

There is a saying in Bihar that while the rest of India has two seasons, in that state there are three, i.e. the Kharif, the Rabi, and the Relief. In the Kosi area during the monsoons, a total of two kilograms of grain per individual are distributed as relief to the marooned population by the Block administration. However, the disbursement of even this meagre support is fraught with unmitigated corruption. Surveys conducted by this writer have affirmed allegations by villagers within-the embankments that the bulk of the relief is diverted to the markets through the connivance of the officials, the village head-men, and the local traders.

The matter has remained largely invisible because political parties have quite successfully divided the population over the issue. The careers of several local politicians depend on resurrecting the 'Kosi Pirith' issue during elections, and distributing lucrative Project contracts to relatives and supporters thereafter. Electoral victories bequeath control over the spoils of the Kosi Project. There is an obvious political interest in keeping the problem alive and unresolved.

The Project has also imposed a formidable burden on the Indian exchequer. A project budgeted initially at around IRs 1,200 million, ultimately cost IRs 3,210 million due to the addition of secondary components. Subsequent allocations have also been extracted from the Centre by Bihar for the augmentation and maintenance of the embankments. Since the Fifth Five Year Plan, the Bihar government has habitually demanded between IRs 500 and 1000 millions per plan from the Centre for such purposes. In the absence of accountability, the bulk of this money has been siphoned off through well standardized routines. Kosi Project contractors usually start work just before the monsoons so that the floods destroy everything, and new contracts are awarded for the same work in the next financial year. The recurring burden is borne by the tax payer and the Kosi inhabitants.

IRRIGATION AND ECOLOGY

On the other hand, the Project has not brought any of the benefits associated with command area development. Of a total irrigation potential of 747,000 hectares created in Purnea, Saharsa, and Darbhanga districts, the actual utilisation has been of the order of 21.7 percent only. In India as a whole the utilisation of the irrigation potential is about 80 percent. The Kosi Project has with difficulty irrigated more than 30 percent of the estimated area in any given year. In the important spring and summer seasons, the utilisation is only 15 and 10 percent respectively. Unequivocally, the level of the realization of potential in this project is the lowest among the irrigation projects of India.

On the ecological front, the effect of the Project on the state of Bihar has been even less encouraging. Between 0.3 to 0.4 million acres constituting 25 to 30percent of the command area, is completely waterlogged as a result of seepage from unlined canals, further aggravated by poor horizontal and vertical drainage. Water has also seeped through the levees because of the rise in the river-bed and the water-table within, which is evident in continuous swamps along the entire length of both the embankments. Because of a neglect of parallel land-levelling exercises in the command area, the maladies have compounded. Instead of enhancing the yields of marginally productive lands, the Kosi Project has rendered large areas totally unfit for cultivation. Meanwhile, elaborate plans are being drawn at astronomical costs by the Bihar government for countering the adverse ecological effects of the Project.

UPSTREAM DIALOGUE

It is now universally acknowledged in India that the design and execution of the Kosi Project was seriously flawed. Apart from technical and institutional problems, officials assert that a lack of co-operation from Nepal has aggravated the situation, and that any ameliorative effort by India can achieve only limited success, as many problems must be addressed in the upper catchments. The officials in charge of flood forecasting in Khagauli, Bihar informed this writer of their inability to make accurate projections with data received only annually from their counterparts in Nepal. Consequently, they have no knowledge of floods until they actually enter Bihar. Requests by India for locating rain-gauges and wireless stations upstream in Nepal have been summarily rejected in the past. Indian officials consider progress in this sphere crucial for better over-all flood management.

Several years ago, Dr. Kanwar Sain, who as the Chairman of the Central Water and Power Commission in the early 1950s, designed the Kosi Project said that the Project had been conceived of as a temporary solution for 25 years only, and was critically dependent upon parallel secondary exercises. Among these, he assigned primacy to soil conservation and land management in the catchments of Nepal. For flood control, Dr. Sain had outlined five structural exercises. Of these, work on two critical components, namely channel improvement and soil conservation has been wholly unsatisfactory.

The effectiveness of the detention basins, the embankments, and the diversions depended on the performance of these tasks. Emphasising the unique sediment characteristics of the Kosi from the inception of the Project, Dr. Sain had recommended that both countries address this problem on a war footing. Reviewing the Project in 1980, Kamala Prasad, a former Kosi Commissioner stressed the need for immediate sediment control, failing which the life of the barrage originally placed at 100 years, would be drastically reduced.

EMBANKMENTS AND FLOOD-CONTROL

It should be noted that the utility of embankments for flood-control in the Kosi has been seriously questioned in the past. At the Patna Conference of 1937, called to discuss the floods in North Bihar, the Chief Engineer had recommended the removal of existing levees on the Kosi arguing that rather than reducing floods in that basin, the embankments had severely aggravated the problem. In this connection it would be pertinent to mention two outstanding examples which demonstrate both the inefficacy and the outright dangers posed by embankments. On the Hwang-Ho, levees were built in 603 B.C., and subsequently raised several times to counter the elevation of the bed due to silt. In 1897, the river which was by then flowing on a ridge, burst its banks and killed one million people. Similarly, the embankments on the Mississippi, were elevated by 17 feet between 1833 and 1927. Yet in 1882, the levees were breached on 284 occasions which caused widespread submergence of the surrounding countryside. In 1927, the damage caused by floods to the levees was so extensive that none of the embankments on the lower Mississippi are in their original positions today.

Owing to comparatively inefficient silt clearance in the Kosi, the prospects do not appear attractive. Apart from the fact that discharges can vary between 100,000 to 950,000 cusecs, a single flood season of the Kosi has been known to elevate limited areas in the basin by as much as seven feet. In the last 40 years, sediment deposited by the river in some sections has raised the flood-plain to a level higher than the embankments. The time-table for disaster is set.

The Project which was initially designed to regulate floods became unmanageable with the addition of irrigation and power as secondary components. The subsequent difficulty of inter-locking the complex range of imponderables in this scheme has underscored the importance of defining all project components from the start. Today, with the progressive intensification of the maladies in the Kosi basin, the redressal of the problems requires several levels of complex integrations that are not achieved easily. It should also be mentioned that despite initial indications favouring cheaper alternatives, costly structural methods were deliberately adopted.

The Kosi episode has been instructive in several ways although it is uncertain as to whether the lessons taught by that experience are being taken seriously. The Project was founded on incomplete data and even now a full survey of the entire basin has not been conducted. While Nepali studies have been confined to their own projects, the last major Indian investigation carried out in 1946 by the Geological Survey of India and the Central Water Irrigation and Navigation Commission, upto the Tamur confluence, was also incomplete and requires updating. To compound matters, neither side has volunteered any information to the other. Further, due to a palpable lack of co-operation, each side has viewed the Kosi from the vantage point of self-interest.

The benefits of experience come retrospectively, and often when it is too late. Earlier, if the search for common grounds had been pursued seriously by Nepal and India, solutions would have come more easily and would have cost less. Having paid an enormous price, both countries must now develop an agenda for the itemised basin by basin resolution of technical, institutional, and riparian issues as they relate to specific projects. Joint commissions on water resources where well-rehearsed routines are re-enacted by each side for the benefit of the other, have proved as ineffectual as political rhetoric. It does not help when discussions reflect neither a commitment to the responsibilities, nor a grasp of the problems. The first step towards the serious pursuit of a national policy and the formulation of a workable consensus is presaged by a better comprehension of the issues. Perhaps, it is still not too late to change.

I.J. Thappa, Ph.D., is a McNamara Fellow currently researching ecological adaptation processes in the Kosi area of Bihar.