Abstarct: Construction of tunnels has been increasing in recent years due to development of infrastructures such as highways, underground, railways and passive defense facilities. Traditional Drilling and blasting methods, as well as modern techniques like using TBM (Tunnel Boring Machine) are common methods to excavate tunnels. After excavation process, support systems need to be applied to make the tunnel stable. There are several approaches to support tunnels. Using steel frxames coupled with shotcrete is usually used as primary support system in loose ground conditions, which is the case in this research. Several experimental and numerical methods have been proposed for analysis and design of such composite structures. The equivalent section can be named as the most applicable approach to analyze these composite primary support systems. Quick installations of steel frxames and rough boundaries of the tunnel due to excavation circumstances lead to different shotcrete thickness and different spacing between steel sets. As a result the thickness of shotcrete and spacing between steel sets can considered as Random variables. In this research, an appropriate distribution functions for spacing of steel frxames and shotcrete thicknesses in tunnels are obtained according to statistical data collected from a recent tunnel project in north of Iran. To achieve this, different distribution functions are utilized to fit the data and each function is evaluated by three well known methods called Chi-Squared, Kolmogorov-Smirnov and Anderson-Darling. According to the results obtained from all three methods, the Generalized Extreme Value (GEV) distribution function is recommended for thickness of shotcrete and the Wakeby distribution function is suggested for spacing of the steel sets. The above distribution functions are then utilized to simulate the real thickness and spacing of steel sets using monte carlo technique. The monte carlo technique was used as tool to evaluate the possibility of failure in various circumstances and the results were presented. The results indicates that effect of imperfections and deviations of shotcrete thickness should be considered in determining safety factor for design of composite sections of steel sets and shotcrete. In order to have optimum design of a horse shoe tunnel, it is recommended that two different safety factors are used in design of the tunnel wall and tunnel arch. Eventually, safety factors of design are determined to reach several certain amounts of failure probabilities.