Abstarct: In this thesis an equilibrium-baxsed model is proposed for analyzing physical capacity withholding strategy by dominant producer in day-ahead market in the presence of wind producers. The market under study is a highly concentrated day-ahead market including one dominant producer with 50% capacity of dispatchable generating units and a number of small producers that participate with maximum capacity. Additioally a number of wind products participate in this market. Between the empirical evidence valid spatio-temporal correlation among geographicaly close wind farms. Analyzing the impact of this spatial-temporal correlation on the wind production pattern and market outcomes is important. An ARIMA-Fuzzy method is used to forecast and generate scenarios for modeling uncertainty in wind production. The offering problem of dominant producer is modeled as a bi-level optimization problem. In the upper-level problem, the dominant producer derives strategy of physical capacity withholding optimally restricted by constraints of it, own generating unites. In the lower-level problem, market operator clears the day-ahead market on an hourly basis. The bi-level optimization problem is recast to a mathematical programming with equilibrium constraints (MPEC) problem, and then, by linearization techniques converted into mixed integer linear programing(MILP). Finally, the wind penetration as well as correlation between wind farms on market power exercised by dominant producer is analyzed.