Environmental performance evaluation of electric enterprises during a power crisis: Evidence from DEA methods and AI prediction algorithms

Abstract

With the continuous occurrence of power crisis events worldwide, meeting society's demand for electricity has kept coal-fired power generation high, which leads to a large amount of carbon dioxide emissions and environmental pollution problems. Therefore, improving the environmental performance of thermal power plants under the background of a power crisis has become particularly important. This research provides a new path for production optimization and environmental performance improvement of electric power enterprises from the perspective of combining prediction algorithms and DEA methods. We apply the modified non-radial directional distance function (MNDDF) method to a set of panel data containing 16 different cogeneration units and 27 different periods in eastern China and construct the grey prediction model to predict the future environmental performance of these units. Empirical analysis shows that our model has high prediction accuracy and offers a reasonable basis for future production decision-making adjustments for electric power enterprises. Based on the results herein, we present valuable suggestions for thermal power plants to improve their environmental performance during a power market crisis, including optimizing resource utilization, improving management level and operating conditions, strengthening technological innovation, and enhancing production factor allocation.