RESERVOIR SIMULATION OF THE MULTIPURPOSE PROJECT UNDER CHANGING CLIMATE: A CASE STUDY OF KANKAI MULTIPURPOSE PROJECT, NEPAL

Abstract

Climate Change has stood as a burning issue since last two decades because of its potential impacts on the environment. Its impact has been observed predominantly in the hydrological cycle in global, regional as well as local scale. As water has stood as one of the major commodity for sustaining a life, its spatial and temporal variation can be a major hindrance in almost all aspects of natural life ecosystem and can have a severe impact on it. In order to develop the resilience against climate change, there stands a challenge to minimize the temporal and spatial variation of water resources. Reservoir system is an artificial system composed of several physical infrastructures which serves single or multiple purposes by regulating the natural flow of water as per the demand downstream and hence to cope up with the temporal variation of water resources. This study has been carried out in order to predict the impact of climate change in the reservoir operation for Kankai Multipurpose Project (KMP), Nepal. Kankai Basin is one of the five medium River Basins in Nepal originating in the Mahabharata range and draining towards south in Mahananda River Basin in India. The catchment of Kankai River at the dam axis of KMP is 1164 sq. km. To carry out this study, four specific objectives have been set. The first specific objectives was the analysis of trend of temperature and precipitation of the stations in/ around Kankai Basin and analysis of trend of Kankai stream flow using Mann Kendall and Sen’s Slope Estimate. It has been observed that annual maximum mean temperature of two of the stations are in rising trend whereas annual minimum mean temperature for one of the stations is in falling trend. There was no significant trend observed for the other stations. Even though there is not much significant trend observed in precipitation, there has been observed significant increasing trend for stream flow. After having the preliminary idea about the trend of climatic parameters, the next objective was the hydrological modeling of the Basin using Soil and Water Assessment Tool (SWAT). Window of nine years for climatic data (1990-1998) along with the other required input parameters were prepared for the model run (3 yrs warm-up, 4 years calibration period and 2 years validation period). SWAT- CUP has been used for calibrating the model for the daily data. 15 parameters were identified to be sensitive from global and one at time sensitivity analysis. The values of NS coefficient and the Coefficient of determination for the calibration period were 0.6 and 0.64 respectively and for the validation period, the values were 0.69 and 0.70 respectively. v The third objective of simulating the reservoir using the historical data has been accomplished by the use of HEC Res Sim software using the observed time series of 2004-2011. The simulation has been carried out, prioritizing the environmental release and the Irrigation water demand in each alternatives/ scenarios formulated for the simulation. 14 scenarios (S1-S14) scenarios have been formulated for the simulation using historical time series. Simulations has been carried out in daily time step of various installed capacity from 60MW to 130 MW and then in hourly time step for 60, 90 and 120 MW with various operation patterns. The result of these simulations with reference to the energy generated, uncontrolled spill and power duration curves suggested 90 MW installed capacity would be the best option among all. This came up with 309.6 GWHr annual energy generation with the average annual uncontrolled spill of 15.6 cumec and 30 MW of energy generation at 75% exceedence. Next objective was the simulation of the reservoir for the predicted future flows. For this the CNRM-CM5 model data were extracted for temperature and precipitation from CMIP5 database for the mid century (2061-2070) for two scenarios RCP 4.5 and RCP 8.5. Extracted data were bias corrected using the linear bias correction technique. The flows for both the scenarios have been generated using the calibrated SWAT model. The flows generated for future, for both the scenarios, were compared to the historical observed as well as the historical SWAT simulated flows. It was found that the future flow had been deviated considerably from the observed flow, but had maintained the agreement with the SWAT simulated flow. 3 scenarios (S15-S17) were formulated to see the impact of climate change in reservoir simulation. Two of them corresponded to the future flow (RCP 4.5 and RCP 8.5) and next one corresponded to the SWAT simulated flow. On comparing the reservoir simulation result with the SWAT simulated historical flow, the energy generated is expected to increase by 13.8% and 28.9% for RCP 4.5 and RCP 8.5 respectively. The uncontrolled spill showed no definite pattern of change.