Optimization of Release from Gheshlagh ReservoirBased on Hydrological Uncertainty Conditions

Dam reservoirs supplies water resources for drinking, agriculture, and industry almost in all parts of Iran. Lack of proper utilization rules and unbalanced demands versus available water has led to many problems regarding to ineffective use of the water resources. Precipitation and flow rate are stochastic phenomena. Therefore, multi-step processes related to the water use efficiency should be considered during the decision-making systems. The aim of this study was to optimize the maximum allowable release from Gheshlagh dam reservoir using linear and dynamic programming models in a one-year time scale. Therefore, Linear and dynamic programming models were applied in Gheshlagh dam reservoir, which supplies the drinking water of the Sanandaj city. Results indicated that the two-parametric log-Normal model appreciably describes monthly water stream flows entering to the reservoir. Thus, this model was applied to compute the stream flow rates with 95, 90, 80, 70, and 50 percent probabilities in each month. There was a good consistency between the drought periods estimated from precipitation data and stream flow rates and 51 months drought period was used in modeling procedure. The results of the maximum allowable release and storage optimization showed that the linear programming model is superior to dynamic modeling approach.