A total of 593 dam failures, which were used by Zhang (2007) to analyze the cause of dams failure, were due to leakage due to the failure of 30% of the dams. Insufficient control of leakage or inadequate preparation of dam and embankments have been reported due to many problems of leakage and failure of the dam. The flow of fluid in the rock mass is influenced by several factors, the most important of which is the discontinuity pattern in the rock mass. Therefore, the correct fracture modeling can be helpful in predicting the permeability of the rock mass. In order to produce a fracture network that is as close as possible to the actual model, a Discrete Fracture Network that simulates water flow system is used. 3DEC software has the capability of generating a random fracture network and solving the flow in the interfacing system. This software has the ability to model the joints in a discrete manner, based on statistical distribution functions, and uses a distinct elemental method to solve it. To study the case, Chamieshir dam data was used to construct the model. First, the fracture network was constructed in accordance with the region and then, with the geological conditions, it was applied to the three-dimensional model of the support, and in order to adapt the permeability of the model, the results of the Lugen test were used. The results show that the calculated leakage rate is strongly affected by the aperture of the joints. Experimental methods are the most common method for measuring the permeability of dams and supports. These methods are fast, easy and based on measurable parameters. A lot of empirical formulas have been proposed for the calculation of permeability The use of any of these methods will have different results. Therefore, studies on these methods are important in order to select the most appropriate method and to calculate the uncertainty of the results. In this study, for estimating the amount of experimental methods, the permeability is calculated using different methods (Hoek, Altovsky, Dupuit , ...) and compared with the results of the field. And the best relationship has been chosen based on the RMSE indicator.