This thesis proposes methods to enhance performance of the automatic fault analysis system for transmission network. These consist of three main aspects, which are how to select a suitable fault location algorithm, the newly proposed algorithms, and how to determine fault clearing time, more correctly. The two newly proposed fault location algorithms include a one-terminal fault location algorithm based on the principle of short circuit calculation, in which it has the advantages of yielding accurate fault location even for a case of high fault impedance, employing the single formula regardless of fault types. In addition, it requires less measurments and parameters associated only withpositive sequence components. The other algorithm is an optimization based algorithm, utilizing data from multi-terminals. This method will relax the constraint of having complete infomation from both ends of the faulty line. In this research, performances of the proposed algorithms have been verified using data obtained from fault simulation on the IEEE 14 bus test system, as well as field measurements from the Electricity Generation Authority of Thailand's transmission system. An accurate one-terminal fault location algorithm based on the principle of short circuit calculation gives much better accuracy when compared to that obtained from the standard one terminal simple reactance method. Nonetheless, the accuracy depends on correctness of the bus impedance matrix which should well represent the system condition during fault period. For the case of optimization based algorithm, it will give an accurate result when compared to that obtained from the one terminal simple reactance method; yet, less accurate than that of the two terminal fault location techniqe in a repective case. However, this proposed algorithm can be applied to a real system which normally has no complete data. In the future, it can also be used to calibrate errors and time unsynchronization of various monitoring devices in the transmission network.
