Comparison of supply-demand based models and time series models in turkish electricity market modeling

Price forecasting problem supports professionals within a wide range of scope, from taking investment decisions for power plants to hourly optimization of a running power plant. When making those decisions, forecasting of market-based prices provide the most important input. Two different models have been proposed to solve the electricity price forecasting problem in the Turkish electricity market. In the first model, time series models have been applied to subsequent month pairs, where the first month has been trained and the second month has been forecasted and compared with the actual prices. Supply-demand based models detail supply in power plant breakdown and intersect with monthly demand in merit-order. Under certain assumptions, price is formed where monthly available capacities are lined per the short run marginal cost principal based on their fuel costs and intersected with the demand. Both model results have been compared and it has been concluded that time series models have performed better with respect to absolute error values.