No description available.

No description available.

No description available.

A research program was undertaken to quantify the effect of flow induced erosion on the stability of natural river banks along the Red River in Winnipeg, Manitoba. The study correlated the percent decrease in factor of safety to intensity of river flow and duration. Two methods to quantify flow induced erosion were assessed, the first method based on observed erosion and the second on theoretically calculated erosion. The first method involved aligning annual historical river bank cross-sections and measuring the distance between cross-sections to represent the erosion induced from the flow year between successive cross-sections. Due to the fact there are no sites along the Red River regularly surveyed, the analysis did not provide for a correlation between erosion from a specific flow event to percent decrease in factor of safety. The second part involved the use of theoretical equations to quantify erosion given the river elevation of a specific flow year. The study showed a 1% to 8% decrease in factor of safety from low to high intensity flows and as high as 14.5% for high intensity flows of long duration. Additionally, the evolutionary stability of the riverbank was generated showing the percent decrease each year in factor of safety due to erosion and the years during which failure occurs. The results correlated well to the previous analysis showing a 1% to 5% decrease for low to high intensity flows respectively and as high as 10% for high intensities of long duration.

This thesis presents a novel technique for prediction of the transient stability status of a power system following a large disturbance such as a fault, and application of the tech-nique for subsequent emergency control. The prediction is made based on the synchro-nously measured samples of the magnitudes of fundamental frequency voltage phasors at major generation/load centers. The voltage samples are taken immediately after a fault is cleared and used as inputs to a binary classifier based on support vector machines to iden-tify the transient stability condition. The classifier is trained using examples of the post-fault recovery voltages (inputs) obtained through simulations and the corresponding sta-bility status (output) determined using a power angle-based stability index. Studies with the New England 39-bus test system indicate that the proposed algorithm can correctly recognize when the power system is approaching transient instability. The proposed sys-tem is then applied to Venezuelan power system and Manitoba Hydro power grid to demonstrate the applicability for large practical power systems. Performance of the pro-posed transient stability prediction scheme under the presence of asymmetrical faults, voltage sensitive loads, unlearned network topologies and measurement noise was found to be satisfactory. Once an impending transient instability situation has been detected, appropriate emer-gency control strategies are triggered to minimize the impact of this on the safe operation of the network and reduce the possibility of a blackout. This thesis examines two differ-ent emergency control schemes: a) A fuzzy logic based emergency load and generator shedding scheme and b) A high voltage direct current (HVdc) power order reduction scheme based on synchronized phasors measurements. These strategies were developed for two power systems with contrasting characteristics: one for the Venezuelan power system which is a conventional power system completely based on alternating current (AC) transmission, and the other for the Manitoba Hydro network which heavily depend on long HVdc transmission for power transfer. The proposed wide area control systems demonstrated good performance on the Venezuelan and Manitoba Hydro power grids.

No description available.

No description available.

No description available.

No description available.

No description available.