SUMMARYThe lightning stroke to power system structures especially overhead lines makes severe damages and results in less reliable power supply. The invention of surge arresters was a revolution in these systems for protecting the precise equipments from lightning stroke overvoltages. Nowadays, with ever decreasing prices, using arrester not only for protecting certain instruments but also for decreasing total risk of flashover in overall network, is investigated by academic and industrial pioneers in this area. In this paper, our goal is to introduce a heuristic method for determining optimum positions for placing transmission lines surge arresters (TLSAs) with acceptable approximation, to get lowest possible value of shielding failure risk of flashover in a selected set of overhead lines. Simulation optimization based on neural net (i.e. Meta Model) and genetic algorithm (optimization algorithm) is invoked to suggest best positions for placing TLSAs. A case study on Kerman 230 kV network shows good achievement of simulation optimization for finding optimum positions of TLSAs. Comparison is also made with the results of transient simulation to reveal the effectiveness of the method.