COMPOSITE SYSTEM OPERATING STATE RISK EVALUATION

There is considerable interest in the application of probability methods m composite system reliability evaluation. The problem is extremely complex because of the need to include detailed modeling of both generation and transmission facilities and their auxiliary elements. Quantitative adequacy assessment of a composite power system is generally performed for individual load points and the overall system. Many utilities have difficulty in interpreting some of the calculated indices as the existing models are often not perceived to include actual system operating conditions. A security constrained adequacy evaluation technique can be used to alleviate the utility concerns. The performance of a composite system can be examined using a set of predefined operating states in terms of the degree to which both adequacy and security constraints are satisfied. This technique is extended in this thesis to examine the impact on system reliability of common mode and station originated outages. A new approach to combine the dependent and independent outages is developed. Composite system operating state risk (CSOSR) depends on many factors such as the actual physical power system, the system operating conditions and the applied constraints. This risk index can be compared with the loss of load probability (LOLP) which is commonly used in HL I studies. The effect on the CSOSR of a wide range of factors is extensively examined in this thesis. Most utilities are concerned with not only the system risk but also the system well being in terms of the degree of margin or comfort. A new technique associated with composite system performance is developed to combine deterministic considerations and probabilistic indices to describe the well being of an electric power system. 'Generating unit commitment analysis is usually performed at HL I based on a presumed acceptable risk level and the predicted system load. Two new procedures are presented to conduct composite system unit commitment for operational planning and daily operation. A matrix multiplication method is utilized to calculate the required time dependent state component probabilities which satisfies the precision, speed and simplicity requirements. The concepts, techniques and procedures developed in this thesis are illustrated numerically using two reliability test systems. It is believed that the concepts and procedures presented will provide useful tools for power system managers, planners, designers and operators and permit them to perform composite system risk assessments, well being analyses and unit commitment studies