APPLICATION OF RELIABILITY WORTH IN POWER SYSTEM PLANNING
MetadataShow full item record
In the past two decades there have been increasing pressures and incentives for utilities to optimize their investments and operations and to more adequately justify plant expansions. A major aspect of the justification and optimization of system expansion is cost-benefit assessment of power system reliability. Power system reliablity assessment has been a topic of much research and is in common usage by most utilities. The assessment of power system reliabllity worth is an area of more recent concern but one where a fair degree of attention is focussed. The melding of system reliability indices and worth, the application of reliablity worth, is an area which has received minimal attention and is addressed as the main emphasis of the thesis. The thesis presents an overview of reliability worth data and of power system reliablity assessment. Considerations and problems associated with the application of reliability worth are discussed for generation, composite and distribution systems. The majority of studies were found to have been performed in the area of generation reliability optimization. The major and more current ones are discussed and compared. Major conclusions are that generation reliability worth assessments result in relative indicators not abselute ones and that interruption cost data do not adequetely include the indirect effects of generation type outages. Composite system reliability worth assessment is discussed with the conclusion that application of worth data is presently difficult because of the immature state of composite relibablity assessment techniques. The "COMREL" composite reliability program is utilized in an example costing application. The thesis places a strong emphasis on the application of reliabilty worth in distribution system studies. Distribution indices are fairly absolute measures of user reliability. Interruption cost data are most applicable in distribution system studies. A major contribution of the thesis is the development of a probalistic simulation program which is used to obtain the probability distributions associated with the reliability indices and interruption costs for simple radial distribution systems. Distributions for the Load Point Failure Rate, Outage Duration, and Annual Interruption Time and for the SAlDl, SAIFI, and CAIDI indices are presented. It is found that the use of the average outage duration to calculate interruption costs can, in a significant number of cases, result in large errors as compared with using the entire duration distribution. The use of the $/KWHR interruption cost coefficient form 1-5 compared with the duration specific $/KW form. An analytical technique for constructing load point outage duration distributions is presented and shown to be computationally efficient as compared to the use of simulations.