Many thermal generating units of an electric power system are supplied with multi-fuel sources such as coal, natural gas and oil. These fuels represent irreplaceable natural resources and conservation is used as a way to increase energy efficiency. Economic dispatch (ED) is one of the significance optimization problems in power system operation for fuel cost savings. This paper proposes a new approach which is hybrid variant of mean-variance mapping optimization (MVMO-SH) for solving this problem. The MVMO-SH is the improvement of original mean-variance mapping optimization algorithm (MVMO). This method adopts a swarm scheme of MVMO and incorporates local search and multi-parent crossover strategies to enhance its global search ability and improve solution quality for optimization problems. The proposed MVMO-SH is tested on 10-unit and large-scale systems with multiple fuels and valvepoint effects. The obtained results are compared to those from other optimization methods available in the literature. The comparisons show that the proposed method provides higher quality solutions than the others. Therefore, the MVMO-SH is a promising method for solving the complex ED problems in electric power system. © ICST Institute for Computer Sciences, Social Informatics and Telecommunications Engineering 2016.