Natural Gas Development in the Brazilian Industry: A Short-Term Projection

摘要

The aim of this study is to estimate the demand for natural gas of the Brazilian industrial sector. Industry is responsible for more than half of the national consumption of this type of energy in Brazil. Therefore, being able to precisely predict the natural gas consumption of the industrial sector is crucial for policy makers. Planning and managing natural gas supply operations are essentials. However, only a limited number of studies specifically address the gas consumption of the industrial sector, both at the national and global level. The existing literature has mostly addressed the composite demand for natural gas and households’ consumption. This paper aims at filling this gap in the literature. We applied the Kalman filter, a Bayesian structural state-space model, to a comprehensive dataset of the energy consumption in Brazil and its industrial sector obtained from the Brazilian Association of Piped Gas Distributors. The Kalman Filter is a simple econometric dynamic model, it acts as an efficient recursive filter, which allows the adaptation of its parameters to each period, thus allowing a better accuracy in demand projections. We based our estimations on an extended version of the model. The proposed framework is innovative in the frame of natural gas consumption projections. We evaluated the robustness of the proposed framework comparing it with two routinely adopted methods. The results of this work proved that the Kalman filter delivers a more accurate projection of the industrial natural gas consumption in the short term compared to the proposed benchmarks. The methodology suggested in this work allows the analysis of time-varying parameters and may be readily employed to obtain demand projections for several other products and energy sectors.