Abstract:
The National Water Development Agency (NWDA), Ministry of Water Resources, Govt,
of India has carried out studies on inter basin water transfers in India. It has identified 30
links for preparation of feasibility reports and has prepared feasibility reports of 6 such
links. The study got momentum due to a recent Supreme Court verdict directing the
Government of India to inter-link all the major rivers in India for inter basin transfer of
water. This entails construction of large river linking projects, which warrants sound
investigation, careful planning and huge expenditure. A faulty implementation of these
projects may be more harmful than doing nothing at all. The studies for most of these
rivers linking are at their initial stages. It is felt that the application of system analysis
techniques will help in better planning for these Herculean task. The proposed Parbati-
Kalisindh-Chambal link under the peninsular rivers development plan is considered for
this study. Three proposed reservoirs, namely, Patanpur in Parbati river, Mohanpura in
Newaz river and Kundaliya in Kalisindh river are proposed to transfer surplus waters of
Parbati and Kalisindh basins either to Upper Chambal basin (Gandhi Sagar reservoir) or to
Lower Chambal basin (Ranapratap Sagar). The Newaz river is a tributary of Kalisindh
river.
Optimal planning of a large-scale river basin as a unit of water resources system is
having a high priority in the economic development of a region. This has resulted in an
urgent need for accurate and efficient management of the water resources for its
conservation and use. System engineering provides methodologies for studying and
analyzing various aspects of a system and its response to various parameters by using
optimization and simulation techniques. Often these aspects are very complex with
different objectives, scopes, scales and timing considerations. In such cases there is
usually no unique model for the solution of the problems. A set of linked models may be
nested in these cases in such a fashion that outputs of one model are inputs to another or
two models are run in tandem. The answer to how the model links should be arranged is
problem specific, but such use of nested models maybe often quite useful.
IV
In India, as per the National Water Development Agency (NWDA) guidelines,
water can be transferred from a river basin to another, only when the exporting basin is
surplus in its surface water resources at 75% water year dependability. An assessment of
the annual water balance of each concerned river basin is carried out for determining
whether a river basin is surplus or deficit in its water resources (surface and ground
waters) in comparison to basin's annual future water demands. Here, the meaning of the
water balance is a comparison between annual water availability and annual water
demands of a river basin, and differs from the conventional meaning of water balance.
Annual water demands are calculated for different purposes like municipal, irrigation,
industrial, hydropower and salinity for the projection year of 2050 (with the expectation
that population would hopefully be stabilized by that time) and then compared with the
75% and 50% water year dependable flows. The annual availability of water consists of
two parts, viz., surface water and ground water. For the problem under study the annual
water balance assessments indicate severe water deficit situation exists in the Upper
Chambal basin. The Lower Chambal basin is marginally surplus in its surface water
resources only due to a committed amount ofimport water it is receiving from the Upper
Chambal basin. Therefore, water transfer options from Kalisindh and Parbati basins need
to be considered in order to reduce the imbalance caused due to the inequitable
distribution of water resources in comparison to waterdemands in the waterdeficit basins.
Yield model serves as an efficient preliminary screening model for reasonable
reservoir designs with release reliabilities near targets. This study extends the yield model
as available in the present form and presents an improved general-purpose yield model
(IGPYM) applicable to a multiple reservoirs system consisting of single purpose and
multipurpose reservoirs. The model is capable of considering more than two numbers of
water uses, different reliabilities for each water use, allows deficit in annual yields during
failure years, and redistribution of upstream regenerated flows in within the year periods.
The model can be applied to both compatible and incompatible water purposes, and
considers each purpose independently or in-group, depending on the total number of
purposes to be considered in a reservoir. It is found that the model offers better flexibility
in selecting reliabilities of water uses and deciding optimal yield failure fractions during
failure years for different water uses. The model can act as a better screening tool in
planning by providing outputs that can be very useful in improving the efficiency and
accuracy of models such as dynamic programming and detailed simulation.
The results of the yield model are approximate and require refinement. Dynamic
programming models are known to be efficient in resource allocation type of problem and
in this work it is decided to adopt DP models to find import water requirements, fixation
of design demands and for reservoir operation for all the reservoirs in the system. To
consider water transfer in a system of reservoirs (sites), it is important to look into two
aspects (a) excess water availability at a source (export) point and (b) annual water
demands at both source (export) and destination (import) points. To cover both the
aspects, initially it is assumed that at each reservoir all the known annual target water
demands have to be met completely. The available water at a reservoir may not be
sufficient to meet all its water demands and a DP model; namely, procurement problem
model (PPM) is formulated for such cases to calculate the import of water required by
each reservoir in a system facing shortage of water. The PPM assumes that unlimited
water is available at the upper mostexporting reservoir (starting pointof the water transfer
link) and hence all the annual target water demands can be met in the system. This
assumption is not practical, but the model is successful in giving the annual target water
export demands for all the water exporting reservoirs.
At this stage all the annual target water and energy demands are known for all the
reservoirs in the system. Another DP model, namely, controlled input model (CIM) is
formulated to fix the annual design demands for all the water needs that can be met with
prescribed annual reliabilities. If the annual target water demands cannot be met, the
model determines annual design water demands that can be met with prespecified annual
reliabilities for each water need. The CIM also does reservoir operation. Annual yields for
all the water needs are obtained at different reliabilities. Asimulation model is developed
to evaluate the anticipated performances of the system for the set ofdesign and operating
policy parameter values obtained through the application ofthe optimization models.
For the problem under study two alternative water transfer link proposals are
studied. Link-I assumes that water will be exported from Patanpur to Mohanpura,
Mohanpura to Kundaliya and Kundaliya to Gandhi Sagar; and Link-II assumes that the
water transfer will be done from Patanpur to Mohanpura, Mohanpura to Kundaliya and
VI
Kundaliya to Ranapratap Sagar. The IGPYM is applied to find (i) the maximum amount
of water that can be exported with design reservoir capacities after meeting their
respective annual target municipal water supply and irrigation demands at desired
reliabilities; (ii) the maximum reliabilities that can be achieved for irrigation, water export
and secondary energy generation; (111) to know the annual amount of water the reservoirs
are capable of supplying for each water use during afailure year; (iv) the maximum annual
firm and secondary energy generations; (v) the trade-offs between different reservoir
yields for known reservoir capacities; and (vi) the alternative reservoir capacities to derive
the same annual municipal water supply, irrigation and energy benefits as obtained from
the proposed reservoir capacities. Different cases are formulated depending on link
alternatives, alternative reservoir capacities and alternative link canal capacities for the DP
models. The PPM results present the amount of import water required for all the
concerned reservoirs to meet their respective target demands completely and the CIM
results present the design demands that can be met with specified reliabilities. Reservoir
operation results using CIM show the achieved annual yield for each water use
corresponding to different reliabilities by each reservoir in the system for all the cases.
Testing of the reservoir operation results are done by simulation and the most promising
cases under each link alternative are identified.
The developed models and their applications present asystems analysis application
methodology for planning and operation ofmultipurpose multireservoirs, involved in inter
basin water transfers. The results show that the Link-I is more promising compared to the
Link-II with respect to meeting their respective demands. The proposed capacity of
Kundaliya reservoir is high (around 350 MCM) and can be reduced substantially, by
marginal increase in the proposed reservoir capacities at Patanpur and Mohanpura (11
MCM and 12 MCM, respectively).