New Approach: Integrated Risk-Stochastic Dynamic Model for Dam and Reservoir Optimization

Abstract

Multiple studies have developed management models to identify optimal operating policies for reservoirs in the last four decades. In an uncertain environment, in which climatic factors such as stream flow are stochastic, the economic returns from reservoir releases that are based on policy are uncertain. Furthermore, the consequences of reservoir release are not fully realized until it occurs. Rather than explicitly recognizing the full spectrum of consequences that are possible within an uncertain environment, the existing optimization models have focused on addressing these uncertainties by identifying the release policies that optimize the summative metric of the risks that are associated with release decisions. This technique has limitations for representing risks that are associated with release policy decisions. In fact, the approach of these techniques may conflict with the actual attitudes of the decision-makers regarding the risk aspects of release policies. The risk aspects of these decisions affect the design and operation of multi-purpose reservoirs. A method is needed to completely represent and evaluate potential consequences that are associated with release decisions. In this study, these techniques were reviewed from the stochastic model and risk analysis perspectives. Therefore, previously developed optimization models for operating dams and reservoirs were reviewed based on their advantages and disadvantages. Specifically, optimal release decisions that use the stochastic variable impacts and the levels of risk that are associated with decisions were evaluated regarding model performance. In addition, a new approach was introduced to develop an optimization model that is capable of replicating the manner in which reservoir release decision risks are perceived and interpreted. This model is based on the Neural Network (NN) theory and enables a more complete representation of the risk function that occurs from particular reservoir release decisions.