Optimal network congestion management using wind farms

With the increased use of wind energy for the power generation several TSO have increasing difficulties for congestion forecasting due to the unpredictable nature of the energy source. An actual method used to deal with days-ahead congestion planning is based on an order of disconnection of the generation of the type "last generation installed, first generation limited". This paper proposes to enhance the congestion management using a real time supervisor. This supervisor is developed to perform automatic and dynamic re-dispatching using both wind and conventional generators. In order to reduce the production constraints to the minimum, the real time congestion management is based on an indicator of the efficiency of a re-dispatching on the power flowing in the overloaded line. This approach leads to reduced re-dispatching costs and increased network reliability. Actual and proposed methods are compared in the paper using Matlab/Simulink simulations of a realistic test grid. It is shown that the real-time supervisor allows maximization of renewable production during congestions while ensuring network reliability.