Forecasting electricity demand of Bursa and Turkey using different methods

In this study, medium and long term electricity demand of Turkey and Bursa and maximum peak loads of Turkey are estimated. The time interval for the long-term and medium-term is determined as follows: the long-term forecasts consist of the period between 2019 - 2028, and the medium-term forecasts include the period between January 2019 - December 2019. Multiple Linear Regression Analysis and Artificial Neural Networks methods were used in this study. Electricity consumption, maximum peak loads, climate data (heating and cooling degree days), population and economic growth indicators were also used as input data. Results obtained for each model were compared with other models. The obtained results were also compared with projections made by TEİAŞ (Turkish Electricity Transmission System Operator) and Ministry of Energy and Natural Resources. Estimations of long-term electricity demand to Turkey and Bursa and annual peak load to Turkey are in line with estimations of TEIAS and MENR. In addition, the ANN model offers a wider range of estimation options for the user, but the processing time is considerably longer than the multiple linear regression analysis.