A COMPARISON MODELLING HYDROELECTRIC GENERATION AND CONSUMPTION BY USING TIME SERIES: A CASE STUDY IN TURKEY

Abstract

Hydropower energy generation and energy consumption in Turkey are modelled and compared using time series analysis based on data analysis system (data mining). In this study, 51 years of data on Turkey's hydroelectric production and consumption were used. For this, various time series are used, such as four-day Simple Moving Average (SMA4), seven-day Simple Moving Average (SMA7), Cumulative Moving Average (CMA) and four-day Waiting Moving Average (WMA4). Models are generated by using simple linear regression analysis. At the end of the study, it was seen that the best prediction model for energy consumption was SMA7, and the best prediction model for hydroelectric production was CMA. Besides, it was found out that the energy consumption was 10.79 times higher than the generation. Thus, the increase in hydropower energy generation was found inadequate to meet the consumption. Increased energy consumption is thought to be provided from natural gas cycle power plants, thermal power plants, and similar fossil fuel sources. This has increased the dependence on fossil fuel. However, new policies towards wind and solar energy have shown promise for the future.