Assessment of the impact of demographic, economic, and climatic factors on long-term forecasting of electricity consumption

Abstract

Relevance: long-term electricity demand forecasting is a crucial task for efficient energy resource planning, de
velopment of energy supply strategies, and ensuring the stability of the energy system. In the context of global
changes such as changes in ambient temperature, population growth, and economic development dynamics, the
accuracy of electricity demand forecasts becomes particularly important for making informed decisions in energy
strategy formulation. Demographic, economic, and climatic factors shape the overall electricity consumption
picture, but the impact of each of these factors can vary significantly depending on specific conditions. The
weight coefficients of factors, calculated within regression models, allow for a quantitative assessment of the
impact of each factor on long-term changes in electricity consumption. Thus, the assessment of the weight coeffi
cients of factors is an important part of the analysis, as it allows improving the accuracy of forecasts and identify
ing the most significant elements that require attention when planning electricity demand in the long-term fore
casting. Therefore, the study of weight coefficients of demographic, economic, and climatic factors and their
impact on electricity demand forecasting is an important step toward more accurate and grounded forecasts in the
energy sector, which has practical significance for the development of sustainable and adaptable energy supply
models in changing conditions.
Aim: to analyze the main factors influencing the long-term forecasting of electricity consumption.
Methods: there were used statistical methods of analysis and forecasting.
Results: the analysis revealed that current electricity consumption has the greatest impact on long-term consump
tion (59.54%), followed by demographic (21.91%) and economic factors (12.28%). The climatic factor has the
least influence (6.27%). The model demonstrated high accuracy (RMSE = 0.1611) and can be applied for long
term forecasting in the power system.