Ph.D. Thesis

Goran Majstrović, Stochastical Strength of Power Grid in Open Market Environment, Ph.D. Thesis, University of Split, Faculty of Electrical Engineering, Mechanical Engineering and Naval Architecture, Power Systems Department, Split, Croatia, October 2007

Description: Introduction of market conditions in power sector assumes many uncertanities in the area of power system control, maintenance and planning. System operator has no more all functions and information that had in monopoly environment: unit commitment, generation prices, cross border trading, demand forecasting. Also, system operator does not define the size and location of new power plants. Besides that, it has to preserve and guarantee non-discriminatory network access, power system security and reliability and all of that with limited and regulated income.

Consequently, in this thesis mulitvariant method and tool was developed in order to help system operator as well as the other market participants to predict influence of one variable change to the other variables in power system under market uncertanities. Multivaijant method assumes mutual correlations between dependent and independent variables that forms characteristic correlation matrix. Correlation matrix analsis could give detailed power system characteristics. With this approach power system is treated as closed stochastical space with fluctuating electrical and non-electrical variables.

This approach can be used in the process of power system control, maitenance and planning. Definition of interrelation between power plant engagement change and network elements loading could help in detection and quantification of market power, while correlation between generation price change and network element loadings could be indicator for congestion management. Mutual correlation between load changes can affect decisinos in the process of network maintenance and planning.

These analses resulted with definition of one more variable, not found in the literature so far - power network stochastical strength. This value is used for the purpose of the most efficient usage of existing and new power network elements as well as in order to predict new power system states within risky and uncertain market condistions.