Fundamental market model design in turkish power market

Abstract

Privatization in Turkish electricity industry has started during early 2000s and evolved into a state where private investors are highly motivated to invest in all the layers of supply chain from generation to distribution. As a natural outcome of this development, Turkish energy markets have structured in time, enabling many buyers and sellers to penetrate. The penetration level varies for different energy markets; such that, while the competition in natural gas market is still imperfect and a state firm is dominant, competition level in power market is relatively much higher even though the liberalization is not yet complete. Moreover, as a result of this liberalization and increasing variety of resources used in power generation, the price for kW of electricity in forward, future and spot markets, is determined based on the laws of supply and demand. The elements of supply and demand in power markets are of high variety and the underlying dynamics are ever changing. Nevertheless, the most outstanding principle of microeconomics still apply. Given the supply is sufficient enough, there is always a price tag for higher volumes. In power markets the volumes are the installed generation capacities of various technologies while the price tag on the volume is determined by whether the power plants are running on a cost basis, namely, fuel driven power plants or renewable power plants. On the other hand on the demand side, substitution of other commodities to power is a key element and demand is driven by consumption points and their reactions to drivers like temperature, capacity utilization and GDP. In this thesis a comprehensive modelling of supply and demand elements in power markets are handled with uppermost care. The main purpose of the thesis is to underline the motivations of different power market players and how they act in a common market place to maximize their individual objectives. After sufficient descriptions and historical evidences are presented, a stack model is built to reflect the supply side of power market and demand is handled based on various drivers. Finally the demand and supply is combined in a linear programming model for cost minimization to determine price and quantity factors for each delivery period. As model results clearly show, fundamental models can be utilized for different analysis requirements like the effect of electrical vehicle adoptions on prices or the effect of extensive photovoltaic penetration on hourly price curves and profits on conventional power plants and so on. The outcomes also show an outstanding model performance where prices of the last 10 years have been simulated with 5.8% mean absolute percentage error. Moreover, the generation levels of different dispatchable generation technologies are re-produced with high accuracy. One of the most outstanding highlights is that reservoir hydro power plants have been acting or are dispatched as substitute to gas power plants, which later on will be discussed in the thesis under shadow pricing behaviour of hydro power plants.