Factors influencing China's decreasing renewable energy consumption

What is it about?

Renewable energy consumption (REC) holds the key to sustainable development. Therefore, many studies have considered the role of REC. However, the factors influencing the REC share in total energy usage (SREC) are not well investigated. Especially, the factors of China’s fast-shrinking SREC are understudied. This research void on the world’s largest renewable energy producer and consumer’s, i.e., China’s decreasing SREC, is alarming and requires thorough investigation. Our study intends to fill this gap by analyzing the factors of China’s decreasing SREC. The study uses both the conventional (descriptive and directional correlational analyses) and some unconventional (Automatic linear modelling (ALM) and Artificial neural network (ANN) Multilayer perceptron (MLP)) approaches to investigate the factors of China’s decreasing SREC. The initial hypothesis testing and most reliable model validation were achieved via directional correlational (Pearson and Spearman) and ALM. The results of ANN MLP (two hidden layers) showed that “Combustible renewables and waste” is the most critical factor. It was followed by “Urbanization, Gross savings, and Alternative and nuclear energy,” respectively. It is suggested that the Chinese government and private investors prioritize their investments based on factors' importance ranking.

Featured Image

Photo by danilo.alvesd on Unsplash

Photo by danilo.alvesd on Unsplash

Why is it important?

First, based on relevant literature and economic theory/insight, the study hypothesizes the relationship between various key factors and the SREC. Second, China was chosen as a case study because its SREC is rapidly declining, which adds to the relevance and practicality of our research. In addition, when compared to other related studies, our study offers some methodological innovations. (1) ALM was used to determine whether the directional relationship-based factor selection presents the most important predictors (independent variables) of China's declining SREC. In other words, we used ALM to find the most stable (reliable) model for the ANN analysis. Over traditional linear regression approaches, ALM has several computational and technical advantages. (2) ANN MLP analysis is used in this study instead of regression-based tools. One of the primary advantages of ANNs over traditional statistical techniques (such as regression) is their adaptability and lack of distributional assumptions.

Perspectives