| Abstract [eng] |
Analyzing the prospective needs for electricity is crucial for national organizations to make informed decisions on new laws, projects, or investment planning, to optimize the distribution of electricity generation, to ensure energy security and reliability, to carry out targeted planning of power plants, and grids, to meet the energy objectives of the EU and other international organizations, to promote efficient electricity consumption and to integrate renewable energy sources. Today, the Lithuanian household sector is one of the main electricity-consuming sectors, with electricity needs growing on a trend. Thus, forecasting electricity demand in this sector can have a significant impact on risk management, security of supply and the achievement of strategic objectives for the electricity sector. In particular, this Master’s thesis analyses scenarios of economic growth, population change, energy efficiency promotion, and changes in the EU wholesale electricity price that affect the prospective electricity needs of Lithuanian households. Secondly, statistical models - regression and econometric - and a simulation MAED model are developed to perform the forecast. Thirdly, based on the assumptions of the forecast and the developed forecasting models, a forecast of the electricity demand of Lithuanian households for the period 2023-2050 is made. Finally, the assessment of the significance of the results and the comparison of the models is carried out based on the MAE, MAPE, and RSME statistical indicators. The results of the study identified that electricity consumption of Lithuanian households will grow in the period 2023-2050. The average annual growth rate of electricity demand projected by the quadratic regression model is 4.6%, by the econometric model it is around 1.8% and by the MAED model it is 3.62%. The MAPE of all models was found to be below 10%, so the forecasts can be considered accurate. The quadratic regression model with the lowest MAE and RMSE compared to the other forecasting models was identified as the most appropriate model to analyze the future electricity needs of Lithuanian households if the country's economic growth and population trends remain similar to the past. On the other hand, the econometric and MAED models, whose MAE and RMSE were higher than those of the quadratic regression model, are recommended to be used if the analysis of the prospective electricity needs of Lithuanian households is to be carried out by taking into account the various economic, social, and technological factors that determine the electricity needs of Lithuanian households, as well as the different scenarios of their development. |
