Due to the worldwide energy crisis, research and development activities in the field of renewable energy have been considerably increased in many countries. In Germany, wind energy is becoming particularly important. Although considerable progress has already been achieved, the available technical design is not yet adequate to develop reliable wind energy converters for conditions corresponding to low wind speeds and urban areas. The Savonius turbine appears to be particularly promising for such conditions, but suffers from a poor efficiency. The present study considers a considerably improved design in order to increase the output power of a Savonius turbine with either two or three blades. In addition, the improved design leads to a better self-starting capability. To achieve these objectives, the position of an obstacle shielding the returning blade of the Savonius turbine and possibly leading to a better flow orientation toward the advancing blade is optimized. This automatic optimization is carried out by coupling an in-house optimization library (OPAL) with an industrial flow simulation code (ANSYS-Fluent). The optimization process takes into account the output power coefficient as target function, considers the position and the angle of the shield as optimization parameters, and relies on Evolutionary Algorithms. A considerable improvement of the performance of Savonius turbines can be obtained in this manner, in particular a relative increase of the power output coefficient by more than 27%. It is furthermore demonstrated that the optimized configuration involving a two-blade rotor is better than the three-blade design. (C) 2010 Elsevier Ltd. All rights reserved.