Seleshi, Yilma (PhD)Shemeles, Paulos2018-06-202023-11-112018-06-202023-11-111998-12http://etd.aau.edu.et/handle/12345678/2070Koka is the only single Reservoir found in the Upper Awash River Basin being utilised both for Hydropower and Irrigation uses. The current maximum capacity of the reservoir is estimated at 1,000 million rrr'. Up to the inlet to the Reservoir, the Upper Awash River Basin drains a catchment area close to 11,300 Km2. A number of important schemes found at the flood prone areas, downstream of the reservoir, are in need of a workable operation rule for Koka which can give them some protection against flooding during wet seasons. In this research an operation rule curve is established. The operation rule curve is established in such a way that all important variables for the hydropower operation and flood control are optimised subject to the constraints. Linear programming in MS-FORTRAN language is used to develop the optimisation model for Koka reservoir operation. The dry season water requirements of downstream irrigation schemes are implicitly guaranteed since the optimised rule curve will ultimately lead to a full storage at the end of every wet season which is desirable by those schemes. Emphasis was also given on the flexibility of the model to allow adaptive operation rules responding to the changing constraints and boundary conditions which are quite noticeable in the Koka reservoir system. The Real-Time operation procedure is formulated to be used in conjunction with the established rule-curve to enable the reservoir operators to make decisions, regarding releases for various purposes, in a considerably shorter period of time (e.g.,daily) using the current hydrometeorological information. The most important component in Real-Time reservoir operation is the Rainfall-Runoff model. These models enables, for a known rainfall, to predict the expected runoff which is an important information for the Real-Time operation of the reservoir and for assessing the possibility of flood hazards well in advance of its catastrophic consequences. It is, however, unfortunate that most models are data intensive that the scarcity of data is the major bottleneck in using them. In this research, a rainfall-runoff model is established employing the SCSenKoka ;ReservoirThesis