Maintaining the stability and reliability of power grid system is one of the main tasks of system operators. For its success it is important to know which of the elements are the most important ones and would cause the most damage or have the most impact on conditions in the system The purpose of this master's thesis is to create a method, which would determine these elements on the basis of multiple factors. The method will also be tested on a simplified version of the Slovenian power grid system, the results and the procedure will be presented in the thesis.
Most of the previous approaches were focused on methods prioritizing only one variable as its main criterium, while the method shown here focuses on many different ones. Main used methods are: Loadflow analysis, contingency analysis and Saaty method. This thesis will aslo briefly present the following previous methods: performance indices, sensitivity analysis, voltage stability indices, artificial intelligence, probability indices and entropy-based methods.
The assignment is carried out using python and a program called Power Factory, which conducts the loadflow simulations. The algorithm will turn one element in the grid off, conducts a simulation, gathers all the necessary data and turns the element back on. This procedure is then repeated for every line, generator, transformer and compensation device in the model. For every case of contingency the coefficients are calculated fort he following quantities: p.u. voltage, voltage angle, losses, power flows, overloading and ENS. Wherever possible (losses and ENS being the exceptions) the basis for calculating these coefficients is the MAPE (Mean absolute percentage error) formula. Every coefficient is then normalized using the maximum index within that category so that the values are comparable in the final results. Before all of them are summed up in the final importance equation, each quantity is assigned a weight, depending on the importance of each quantity. This is done using the Saaty method, where the inputs in the matrix are the relationships between the quantities. They are divided into three groups. The first group includes losses and voltage angle, the second includes p.u. voltage, ENS, and power flows, and the last group consists of overloads, which can potentially cause the greatest damage to the system. Then the index of each quantity is multiplied by its weight and added to the final equation of importance which sum all weighted indices, providing the importance of each element.
The final product of this thesis is the graph, showing all the elements in the grid, ranked by their importance. The graph shows, which elements in the grid are the most important, meaning that they should be focused on, in terms of maintenance and upgrades, in the future. The results presented here, are not from any of the real grids, but its purpose was to develop a method which can be used on them in the future.
|
