In this work, a simulation methodology is proposed to obtain future scenarios of demand in buses of an electric system. This application is useful in the power systems industry for operation planning and system expansion analysis. Quantile auto-regressive models are proposed to model the behavior of the buses. The models rely on the previous creation of scenarios of energy signals that guide the long-term behavior (growth, shrinkage or stagnation) of the load levels in the buses. They count as well with the historical data of the buses to characterize the short-term movements (intra-day and intra-week movements). The proposed methodology was implemented in the Julia language and was applied to real data of the Brazilian electric sector.