Empirical Analysis and Future Projection of the Ratio of Primary Energy Production to Consumption (RPC) in Turkey and Upgrading Policies

Abstract

This study was aimed at obtaining numerical equations to estimate of the ratio of primary energy production to consumption (RPC) and the ratio of primary energy import to consumption (RIC) of Turkey by using different energy and economic indicators and an artificial neural network (ANN) approach. It intends to contribute to the strategies necessary to preserve the supply/demand balance of Turkey. For this purpose, three different models were used to train the ANN approach. In Model 1, main energy indicators such as installed capacity and gross generation were used. In Model 2, gross domestic product and population were used. In Model 3, gross national product and population were used in the input layer of the ANN. The RPC and the RIC were in the output layer for all models. In order to train the neural network, economic and energy data from 1994 to 2005 were used for the projections for all models. The aim of using different models was to estimate the RPC and RIC values with a high confidence for future projections. The R² values of RPC for training data were found as 0.999947, 0.999446, and 0.999634 for Model 1, Model 2, and Model 3, respectively. Similarly, the values for RIC were obtained as 0.999956, 0.999963 and 0.999959 for Model 1, Model 2, and Model 3, respectively. R² values for testing data are obtained as nearly 1 for RPC and RIC. The results of the analysis indicate that the RPC of 64.95% will reduce to 12% within 10 years while the RIC will increase from 77.78% to 99.5%. Consequently, the utilization of renewable energy sources and nuclear energy is strictly recommended to provide the stability of the RIC value for Turkey.

Keywords:

• artificial neural network
• economic indicators
• energy consumption