The long-term forecast of energy demand and uncertainty evaluation with limited data for energy-imported cities in China: a case study in Hunan

Abstract

The long-term forecast of energy demand has assumed significant importance in the energy planning of energy-imported cities. However, the required data are usually limited or unavailable. In this paper, we developed a combined model based on Long-range Energy Alternate Planning and econometric models to forecast the energy demand of Hunan province from 2012 to 2030. Based on historical data, population was predicted by Leslie matrix, the industrial structure, traffic turnover and productions of high energy-consuming industries were forecasted by Autoregressive Integrated Moving Average and Vector Autoregressive models. Three scenarios were designed with different energy-saving polices and population growths. The uncertainties of forecasting results were analysed by Monte Carlo method. Results showed the total energy demand increased but the growth rate gradually slowed down under each scenario. In 2030, compared with business as usual scenario, the energy demand in second child scenario only increased 29 thousand tons coal equivalent (tce), showing a slight population growth effect; while the energy demand in energy saving scenario significantly decreased 450 thousand tce, showing a significant potential of energy conservation and efficiency improvement. Although the uncertainty of forecasted results increased over time, the uncertainty was still less than 15 %, showing a good accuracy of this combined model.