OPTIMAL WATER UTILIZATION AND MANAGEMENT OF AN IRRIGATION SYSTEM

Abstract

The comprehensive planning of irrigation projects considering the various aspects related to them is of primary concern for an overall increase in the efficiencies of these projects which have been subjected to much criticism. Planning and management of the water resources and then efficient distribution of the available water in the fields for an increased agricultural production are the two major objectives which should be met to make an irrigation project successful. A number of simulation and optimization models have been developed and applied to derive planning and operating strategies for large irrigation reservoir projects over longer intervals of time (generally monthly). However less attention has been paid to develop demand based water releasing strategies for shorter intervals of time (weekly, bi-weekly, ten-daily periods) for which the canals are operated leading to wide gaps between irrigation water supplies and demands over shorter intervals of time. Thus there exists a need to develop planning models for reservoir release that could optimize water releases according to the canal operation days. To obtain reservoir yields for shorter time intervals, a series of historic data of reservoir inflows for shorter time intervals are required which are not generally available for a longer time period and need to be generated. After planning for the short term reservoir releases, these releases should be efficiently distributed to the irrigation system so that adequate and timely supplies are provided to the farmers by developing appropriate canal delivery schedules. Studies regarding the assessment of water deliveries in irrigation systems reveal that there is non-uniform and inadequate distribution of irrigation water to the farms (Clemmens and Molden 2007) leading to low productivity of irrigation projects. iv Hence the present study has been undertaken for optimal water utilization and management of an irrigation system under the Harabhangi irrigation project, an interbasin irrigation project in Orissa state. Under this project, a reservoir has been built having a live storage capacity of about 100 Mm3 for supplying irrigation water to a command area of 9150 ha. For releasing water into the main canal an operational plan has been provided by the authorities of the project. A rotational system of irrigation is practiced in the commandarea. In an earlier work on this project (Jena 2004), optimal monthly water releases from the reservoir were determined using monthly river flows at the dam site. Whereas the optimal releases from a reservoir should be obtained on the basis of actual canal operation days (considering irrigation and non-irrigation days) specified for the particular system under study which requires generation of daily inflow data. Planning of releases in these short time periods require estimation of daily crop water requirements. Further canal schedules need to be developed to distribute the available water under the rotational water delivery system considering the multiobjectives of an irrigation system. Scheduling of canal deliveries for a rotational system is a complex task if the canals are large in number and vary in design discharge, length and command area. Special skills are required to lay down priorities for allocating the water with a defined set of objectives. To obtain canal schedules quickly for any irrigation system with rotational irrigation there is a need to develop a software for decision making for the water managers. To accomplish the above mentioned objectives for Harabhangi irrigation project the following studies were carried out. For the existing catchment area of the Harabhangi reservoir, 8 years of daily rainfall/runoff data have been calibrated and validated using a hydrological simulation model based on modified SCS-CN method by dividing it into two parts and different cases are taken up. Daily crop water requirements are estimated using the Penman Monteith method. Optimal annual reservoir yields and within year yields for irrigation from the Harabhangi reservoir have been determined at 75 percent annual project dependability, with a provision to supply a predetermined partial amount of irrigation water during failure years to take care of the minimum food requirements of the people in that area. Implicit stochastic linear programming models, i.e., yield model and an integrated reservoir yield model have been used for this purpose. The optimal within year yields are determined on monthlybasis and on the basis of project specified canal operation days using both the models. An optimal cropping pattern is also simultaneously determined incase of integrated reservoir yield model. To efficiently distribute the water available at the head, schedules for canal water delivery on the basis of rotational irrigation are developed using a multi-criteria mathematical model based on the objectives or targets of irrigation. Some modifications in the model are suggested to remove the drawbacks of untimely water supplies and inconveniences of the gate operator. In the present study this proposed model is designated as the improved planning model for canal scheduling. On the basis of this proposed model, a software has been developed in a spreadsheet environment by developing macros in MS Excel's internal language: Visual Basic for Applications (VBA), to obtain canal delivery schedules for rotational irrigation. Based on the above studies, the following are the findings in respect of the Harabhangi project. Using the hydrological simulation model, for one of the cases the model performance is best; with efficiency under model calibration to be 81% and under model validation to be 69%. The calibrated model parameters obtained for this vi case are subsequently used to generate daily runoff data of 8 years. In this way 16 years of daily inflow data is made available for reservoir yield analysis. The optimal annual reservoir yield obtained according to the project specified canal operation days using the IRYM is 204.82 Mm3 and would be the optimal target reservoir yield which may be considered for planning reservoir releases in the actual canal operation days according to the optimal cropping pattern. Actual releases in the 43 irrigation and non-irrigation periods are obtained for 16 years by carrying out simulations using project base cropping pattern and optimal cropping pattern. A canal operational schedule is recommended for the whole irrigation distribution system for the Kharif season using the proposed model for canal scheduling which is more effective in fulfilling the multiple objectives