A large scale unit commitmeent algorithm for systems wit radially connected areas

Since many countries worldwide are going toward the liberalisation of electric energy markets, the need of forecast tools for unit commitment is growing. These tools allow us determining a reliable forecast for the hourly production and price and use of each generator. These are crucial pieces of information for all the players in the market, which enable to compute lower bounds for expected prices on real markets. Although unit commitment is a classical problem, and many professional or research software products are available for its solution, it is not easy to find a tool that is able to manage systems with a large number of generators, taking into account the dynamic constraints between different hours on a long time horizon, and the transmission constraint between several areas. The proposed methodology for optimal unit commitment has been developed with the purpose of giving a reliable solution for systems with hundreds of generators, radially connected areas. This solution performs commitment and dispatch hour by hour, searching for overall optimal solutions, taking into account the above-mentioned constraints, on a time horizon on 1 year. The algorithm has been tested on a real system, namely the Italian one and final results are produced within a reasonable computation time.