ECS-9721447 Pahwa Restoration of power after an extended outage in a distribution system results in load much higher than normal load due to cold load pickup. In addition to initial transients, the load during cold load pickup has an enduring component that can be 2 to 5 times normal load and may last for several hours. The enduring component is mainly due to loss of diversity in the cycling type of loads, such as air conditioners, refrigerator etc. Since utilities would like to utilize the existing assets to the maximum extent, cold load pickup is an important issue in design and operation of distribution systems. The PI received funding for a project (Award Number 9311257) in 1994 from the National Science Foundation to investigate the impacts of cold load pickup on distribution systems from both operation and from design points of view. Techniques for determination of the optimal sequence of sections for step-by-step restoration of distribution systems were investigated. Also, system design to determine the optimal number of sectionalizers and the optimal size of substation transformer needed for the system was investigated. Three journal papers and three conference papers were published based on the research. The research also formed partial dissertation of one Ph.D. student and full Ph.D. dissertation of another student. This proposal is being submitted to request funding for further investigation of the impacts of cold load pickup on distribution systems. Specifically, the research will address 1) modification of the transformer loading model, 2) further investigation of the optimal restoration sequence, 3) inclusion of better models for cost of transformers and cost of interruption in the optimization problem, 4) inclusion of voltage drop and feeder capacity constraints in the optimization problem, and 5) issues related to estimation of the cold load pickup model parameters. The proposed research lays a foundation for integration of distribution automation into design of distribution systems. So far distribution automation has mostly been discussed from the point of view of system operation. However, as a result of distribution automation, new methods of designing distribution systems will be needed. The new designs should utilize automation effectively and at the same time keep the system cost low. As the penetration of distribution automation increases, the new design procedures will become an integral part of distribution system design. These issues will gather more attention in the future due to the changes that are taking place presently in the utility business. These changes will result in the restructured utility environment under which the distribution companies will be required to manage their assets more efficiently and effectively. The proposed project will offer a novel opportunity for a graduate student to pursue the Ph.D. degree and receive training in the field of distribution systems as well as in various mathematical and computational techniques. The results of this research also will be extremely useful for the PI for integration into research-based courses.
