Analysis of the Brazilian Energy Policies for Natural Gas Using Artificial Neural Networks

Finding appropriate forecasting methods for the effective management of energy resources is extremely important for improving energy consumption efficiency and decreasing its impact on the environment. Natural gas is a primary source of electricity in Brazil as well as the world because it is a promising alternative fuel for reducing polluting gas emissions. In Brazil, the price of natural gas is higher than that in other countries. A new policy for the gas market, also known as the new gas market, has recently been approved and is being implemented to solve this problem. Thus far, the impact of a changing gas energy policy on national demand is unclear. This work uses multi-layer perceptron type of artificial neural networks (ANNs) to analyze Brazilian energy changes. Future demand for new and old gas markets are forecast from 2021 to 2030. Brazil’s total demand for natural gas is described using seven demands (industrial, automotive, residential, commercial, generation, cogeneration, and others). These demands are predicted individually using four input variables (time (in months), US dollar price (converted to Brazilian real), natural gas price, and GDP) for ANNs. The results show that adopting a new gas market would lead to an increased demand. Conversely, if there were no changes in the energy policy, the demand for natural gas would decrease over time. Because the input variables can be quantified for any country, the proposed approach can also be used to forecast the demands of other countries as well, which would be useful in analyzing their energy policies.