Waste collection is a highly visible municipal service that involves large expenditures and difficult operational problems. In addition, it is expensive to operate in terms of investment costs (i.e. vehicles fleet), operational costs (i.e. fuel and maintenances) so that generating small improvements in this area can lead to huge savings in municipal expenditures. Among the issues raised in the process of decisions making by managers and associated policy makers, one can point to determining the optimal weekly policies of waste collection. In this paper, the periodic capacitated arc routing problem (PCARP)in mobile disposal sites is described and the authors seek to determine the optimal routes of required edges (streets or alleys) per week, number and location of mobile disposal sites, and the number of required vehicles. We present two simulated annealing algorithms, which are different in cooling schedule and number of iterations of each temperature. To evaluate the performance of these algorithms on small-sized problems, the solver “CPLEX"in application “GAMS” is used. The experimental results show that the presented algorithms have appropriate performance and a reasonable time range.