I. Introduction
Electric power systems have become vulnerable to voltage collapse, due to lack of reactive power support, which has significant impact on the reliability of the entire power system. In this context, sensitivity analysis is a promising tool to asses the effect of voltage collapse on power systems reliability. Sensitivity analysis of reliability indices of composite systems with respect to system parameters such as failure and repair rates has been extensively addressed in the literature. Nonetheless, sensitivity of the reliability indices with respect to voltage and reactive power limits has not been given much attention thus far. In performing reliability studies of a given power system, a power flow solution is repeatedly carried out. In the context of power flows, the DC power flow model and linear programming with the objective of minimum load curtailment is probably the most common model used in the traditional sensitivity analysis based methods. Nevertheless, calculating the sensitivity of the reliability indices with respect to system operating constraints such as voltage and reactive power constraints requires determining Lagrange multipliers for such vital constraints. When integrating the DC power flow model into the linear programming optimization problem, the only Lagrange multipliers that are accessible are those associated with power demand, generation, and transmission lines constraints. However, in order to calculate the sensitivity of the reliability indices with respect to voltage bounds and reactive power constraints, these constraints should be modeled in the optimization problem, which is not applicable using the DC power flow model.