In electrical power systems, the impact of interruptions due to failures can be reduced through expansion planning studies. While high investments result in very expensive and more reliable decisions, reduced investments can lead to unreliable systems. Therefore, it is evident that economic and reliability constraints are conflicting, which makes decision-making difficult in planning and operation stage. The reliability theory, based on probabilities and stochastic processes, allows modeling the random behavior of equipment to estimate performance indices such as Loss of Load Cost. However, parameters as equipment failure rate and repair time are subject to random variations due to limited or nonexistent operating histories, aging and statistical errors. This paper proposes a technique for considering uncertainties on stochastic equipment data in power systems expansion planning. Based on the Monte Carlo Simulation, the proposed technique uses Interval Arithmetic as a method for calculating uncertainty through the theory of imprecise probabilities (P-Box). The application in a test system and a real transmission system allows observing the behavior of the reliability cost as well as the final cost of alternatives for expansion of these systems with the consideration of uncertainties along the expansion horizon.