Comprehensive probabilistic assessment of voltage sags in distribution network

Overall aim of this thesis is to develop a comprehensive probabilistic method for the assessment of the number and characteristics of voltage sags at power system buses and inside the industry facility, by taking into account all relevant system parameters and properties. Initially, definitions, nature and detailed characteristics of voltage sags together with their causes and consequences are discussed. In the stochastic approach, voltage sags are traditionally predicted by the method of fault positions and the concept of the area of vulnerability though these two methods are associated with some disadvantages. Those disadvantages and the area in need of improvement are highlighted in the thesis. Then, the development of a comprehensive method for voltage sag prediction is discussed. The method combines the ideas of the method of fault positions and the concept of the area of vulnerability and it takes into account the probabilistic fault rate at each fault location, and the probability of the failure of the primary protection system. Also, the method considers the probability of the occurrence of a fault on a particular phase(s) at a given fault location as any asymmetrical fault can occur at any phase(s) of a three-phase system. After that, the influence of different fault distribution patterns with different distribution parameters on the prediction of voltage sags is discussed. Faults are modelled along the line according to several most characteristic distributions, such as uniform, nonnal and two types of exponential. In reality, it would be unnecessary to model all possible fault distribution patterns along each line in the power system and also it might take enormous amount of time. Therefore, a methodology for identifying the most crucial lines where different fault distribution patterns need to be modelled is introduced in the thesis. The full details of this methodology are also given in the thesis. Traditionally, faults in power systems are assumed to be cleared by the primary protection system. It is however almost impossible to obtain such a 100% reliability in reality. Therefore, the effects of the failure of the primary protection on the duration of voltage sags are studied and a straightforward method for determining the probability of the failure of the primary protection is developed. Though the number and characteristics of voltage sags at system buses can be assessed by the comprehensive probabilistic method, those at equipment terminals will be different due to the transformer employed at the entrance to the industry. A method to assess the performance of voltage sags at the low voltage level is therefore established by considering the effects of different types of transformer connections on the voltage sag propagation. When induction motors are connected, the characteristics of voltage sags, particularly the duration, might changed. Therefore, the analytical method is developed to assess the actual voltage sag performance at the equipment's terminals by considering the interaction between voltage sags and induction motors. The results obtained by the methodology developed in this thesis strictly obey the theory of induction motors and entirely agree with those of the previous studies. It is demonstrated on the realistic size distribution network in order to show the difference in voltage sags performance with and without induction motors. Finally, the thesis describes the comprehensive modular software purposely developed for dealing with all above-mentioned studies. The software is essentially database application and as such could equally use databases obtained during simulations or databases formed following prolonged monitoring in realistic power networks. The software is capable of handling detailed analysis of voltage sags in the network as well as in-depth analysis of voltage sag performance at individual buses and inside the industry facility.