Analysis of the SARIMA Method for Short-Term Electricity Demand Forecasting in Peru's Interconnected Electricity System

Abstract

This work evaluates the performance of the SARIMA model in forecasting short-term (daily) electricity demand in the Peruvian electricity system, highlighting its ability to capture seasonal patterns and intraday variations inherent in load behavior. By incorporating autoregressive, differentiation, and seasonal components, the model adequately reproduces daily fluctuations, generating estimates that are consistent and coherent with the historical dynamics of the series. Likewise, the results demonstrate the robustness, interpretability, and computational efficiency of the proposed approach, making it a competitive alternative to more complex methods. Thus, the SARIMA model is positioned as a competitive tool for operational planning and short-term decision-making processes within electric power systems.