When electricity prices were regulated, hydropower optimization often considered only the inflow uncertainty. In a deregulated electricity market, price uncertainty must be also considered in addition to inflow uncertainty. This makes the operation problem more challenging due to inclusion of the objective of minimizing risk. It also makes the objective function nonlinear while the estimation of risk is problematic. For dealing with uncertainty, a set of finite scenarios of inflow or price sequences may form the evolution of information over the stages that are used in optimization algorithms. Such implicit optimization methods can be seen as an extension of deterministic optimization. A disadvantage is the number of scenarios may grow exponentially in multi-stage optimization problems, even with only a few branches at each stage. An explicit method, denoted as the Fletcher-Ponnambalam model (FP), has been recently developed tor the first and second moments of the storage state distributions in terms of moments of the inflow distributions. This method provides statistical information on the nature of random behaviour of the system state variables without any discretization and hence suitable for multireservoir problems. Also not considering scenarios makes it computationally inexpensive, as there is little growth to the size of the original problem. In this paper, we introduce the price uncertainty into the FP model; our results indicate that the method could achieve optimum policy considering also the reduction of risk, using the second moment information. Our study is for medium term operations of a single reservoir. The results are compared with corresponding results from simulation and where possible, with the well-known Benders' Decomposition method (BD).
