A METHODOLOGY FOR RISK ANALYSIS IN RAIL CAPACITY PLANNING UNDER OPERATIONAL UNCERTAINTIES

This work proposes a statistical methodology for risk analysis in the execution of planned volumes in rail transport, considering operational uncertainties. The approach replaces the traditional deterministic model with a stochastic formulation based on random variables, adjusted from historical data. Capacity is modeled as a composition of probability distributions via discrete convolution, allowing to derive executable volume curves by flow, fleet, and system. Risk metrics, such as the P80 percentile, are used to support operational decisions. The methodology was applied at VLI Multimodal S.A., integrated with monthly and annual planning. The solution allows you to simulate scenarios, adjust assumptions, and extract execution probabilities, demonstrating gains in robustness in logistics planning and greater visibility of the risk of underexecution.