UC Irvine

Reservoirs store a large proportion of the Earth's limited fresh water resources. This water sustains life and makes regional developments possible. The construction of reservoirs does not reveal the overall system success since it is assessed only by operating and measuring whether their proposed benefits are achieved for that single reservoir. Managers of the reservoir systems often face the challenges of fulfilling various conflicting operational objectives while complying with the contractual agreements, institutional limitations, water rights privileges, environmental concerns, and hydrological and demand uncertainties. Obtaining optimum reservoir operation strategies by honoring these restrictions necessitates computer modeling. In this thesis, we introduce the Elaborate Reservoir Operation Management model, EROM, which simulates the storage and power generation of a single reservoir or a system of reservoirs that have one or multiple purposes in response to the record of inflow values and site characteristics under a predetermined operation strategy. As the name suggests, the model differs from the other long term reservoir operation analysis models with its sophisticated tools that provide more thorough representation of the real systems to be modeled. We perform three case studies to test the model with the data of a real-world project which balances water supply, power generation and flood protection. Within these case studies, rule curve based operation optimization analysis of the system are successfully performed by integrating EROM with the evolutionary multiobjective optimization algorithm, AMALGAM, and considering two objective functions as the maximization of average annual generated energy and the minimization of shortage index. The results of the case studies are presented and discussed.