IMPACT OF MODELING ACCURACY ON INTEGRATED OPTIMIZATION OF A HYDROTHERMAL PORTFOLIO ON SHORT-TERM MARKETS FOR SCHEDULED ENERGY AND SYSTEM RESERVE

摘要

Due to the liberalization of the European energy market, power generation and trading companies aim for profit maximization instead of cost minimization. Especially for a short-term time horizon this motivates a new modeling of the short-term hydrothermal unit commitment problem as a closed-loop optimization taking into account all relevant technical and market-related restrictions particularly on reserve markets. This work presents a computational efficient method for modeling both hydraulic and thermal power plants in a mathematically exact approach of mixed-integer quadratic programming. In order to achieve practical results each unit is described in detail by its characteristic constraints. This leads to reasonable power plant schedules especially distinguishing between provision of reserve capacity and supply of reserve energy. In this context the closed-loop formulation of the resulting optimization problem overcomes drawbacks of prior approaches by ensuring that the solution is the guaranteed optimum. After validating the method's functionality it is applied to a realistic portfolio by optimizing short-term generation and determining the optimal share of reserve and scheduled energy for the considered system. Comparing the different power plant schedules achieved by including, simplifying or even neglecting single optimization constraints in different simulations the additional value of the new optimization method and the applied modeling accuracy can be quantified.