Optimal Operation of Danjiangkou Reservoir Using Improved Hedging Model and Considering the Effects of Historical Decisions

Abstract

For reservoirs with multiple water demands, losses in water use utility because of shortages in the water supply from historical operation decisions can be considered information to improve the current decisions. This study proposes an improved hedging model that uses the parameter optimization method to incorporate the effects of historical decisions. A new cumulative water shortage impact index is proposed to evaluate the severity of historical water shortages and their impact on release decisions, and the Palmer drought severity index (PDSI) is incorporated to improve the hedging model. Based on the new index, a weighting factor of previously derived hedging rules is optimized, which represents the relative importance of water release and water storage, and more realistic hedging rules and operation decisions for reservoir managers are obtained. The new model is employed at the Danjiangkou Reservoir, which serves as one of the major storage points on the middle route of the south-to-north water transfer project (SNWTP). The results show the following: (1) when the reservoir inflow is insufficient, the initial weighting factor (initial w) and its corresponding scaling factor can be obtained to guide decisions by considering the forecasted inflow, starting regulation level (SRL), and expected water level at the end of year (WLE); (2) by following the optimal solution, the water delivery process becomes more uniform, and a more reasonable water supply order can be determined according to the unit water supply benefit for each water user; and (3) in an average dry year and in the driest year in the historical record, with initial w values of 0.7 and 0.15, the SRL is greater than 158 m, and the delivery rates of the SNWTP water in the initial stages should be controlled at approximately 0.85 and 0.4, respectively.