SUMMARY

With the ever advancing state of computer systems, it is imperative to maintain the most up-to-date and reliable safety evaluation data for nuclear power systems. Commonplace now is the practice of updating old accident simulation results with more advanced models and codes using today's faster computer systems. Though it may be quite an undertaking, the benefits of using a more advanced model and code can be significant especially if the result of the new analysis provides increased safety margin for any plant component or system. A series of parametric and sensitivity studies for the Loss of Normal Feedwater Anticipated Transient without Scram (LONF ATWS) for Southern Company's Vogtle Electric Generating Plant (VEGP) Units 1 & 2 located near Waynesboro, GA was performed using the best-estimate thermal-hydraulics transient analysis code RETRAN-02w. This thesis includes comparison to the results of a generic plant study published by Westinghouse Electric Corporation in 1974 using an earlier code, LOFTRAN, as well as Vogtle-specific analysis. The comparative analysis exposes and seeks to explain differences between the two codes whereas the Vogtle analysis utilizes data from the Vogtle FSAR to generate plant-specific data. The purpose of this study is to validate and update the previous analysis and gather more information about the plant actions taken in response to a LONF ATWS. As a result, now there is a new and updated evaluation of the LONF ATWS for both a generic 4-loop Westinghouse plant and VEGP using a more advanced code. Beyond the reference case analysis, a series of sensitivity and parametric studies have been performed to show how well each type of plant is designed for handling an ATWS situation. These studies cover a wide range of operating conditions to demonstrate the dependability of the model. It was found that both the generic 4-loop Westinghouse PWR system and VEGP can successfully mitigate a LONF ATWS through the core's operating cycle.