Factor analysis of energy efficiency in the fuel and energy complex based on the selection of dominant indicators

Relevance. The fuel and energy complex represents a fundamental sector of the national economy. Under conditions of rising energy prices and external risks, improving energy efficiency and implementing energy-saving measures becomes a scientific and practical necessity At enterprises of the fuel and energy complex (including electric power, oil and gas, coal, and thermal energy industries), planning based on multifactor forecasting of fuel and energy resource consumption significantly improves management efficiency. In this regard, identifying the key factors affecting the energy intensity of the fuel and energy complex enterprises is of great importance for effective managerial decision-making. Aim. To determine the factors affecting the energy intensity of 23 large industrial enterprises for 2000–2023, as well as to develop a method and algorithm for assessing and establishing the statistical significance of their impact. Methods. Pearson’s correlation analysis, the -test, and -value estimation algorithms based on the Fisher–Yates distribution table. To automate the calculation of -values for large datasets, a the authors have developed the sixth-order polynomial regression model, establishing a relationship between the -statistic and the number of degrees of freedom ( ). Results. The factor (energy losses) in most cases showed a low level of statistical significance and was excluded from the analysis; the factors (specific electricity consumption per unit of output) and (electricity utilization efficiency coefficient) were identified as the most significant. The results of the -test demonstrated a high statistical significance of the relationship between factor d and energy intensity for many enterprises (for example, for the 18th enterprise: ). The proposed polynomial model provided highly accurate approximation of Fisher–Yates table -values ( ) and enabled automated calculations.