Particle Contamination of High Voltage Dc

A scaled dc insulator contamination experiment in which the electric field, air flow, and particle dynamics around an insulator string are modeled has been implemented. Outstanding contamination features, including sharp clean rings and leeward wakes on the insulator surfaces, and diffuse contamination bands on the pin caps, are explained using simple trajectory models, which are both qualitatively and quantitatively verified. A simplified contamination model is demonstrated, in which the deposition rate on surfaces that are contaminated is given by the average product of particle mobility, particle mass density, and normal electric field. Large scale deflection of incident charged particles is shown to be unlikely. Segregation of oppositely charged particles to upper and lower regions of the insulator string is observed, but explained via a trajectory perturbation analysis. A dynamic particle charging model suggests that corona ions may be capable of charging initially neutral particles to values comparable to naturally occurring mobilities. Design concepts for improved dc insulators, aimed at minimization of the normal electric field on insulator surfaces, are offered. The time average "particle mass densitycollection time product" is indicated as a good parameter for evaluating contamination severity at both existing and proposed dc transmission line sites. THESIS SUPERVISOR: James R. Melcher TITLE: Professor of Electrical Engineering Acknowledgements: I would like to extend my thanks to the many individuals whose help and influence led to the completion of this thesis. Prof. James R. Melcher, my thesis advisor, provided me with overall direction, and gave good insight and unique approaches to the problems encountered. He likewise introduced me to the theoretical fundamentals upon which the models presented here are based. Dr. David Jolly, staff researcher at MIT Electric Power Systems Engineering Laboratory (EPSEL), managed the contract under which this work was performed. He also helped me with various experimental problems, and provided overall perspective of the significance of this work to the engineering world. His infamous dry wit has been a source of great moral support. Dr. Jolly, and research assistants Fred Donahue and Ed McHale were responsible for the initial stages of construction ofthe experimental test chamber. Several undergraduates deserve recognition for their contributions to the experiment. Craig Poole performed admirably during the unpleasant and messy orange dust phases of the experiment, when the management techniques (often poor) and measurement proceedures (often tedious) for the test contaminant were developed. My apologies to the members of EPSEL for fluorescent fallout experienced during this peroid of time. Rick Baer constructed the generating voltmeter used for the voltage distribution measurements reported in Appendix C. Alan Barnett constructed the mobility measuring device, unofficially named the Barnettometer, as part of his Bachelor's thesis. The electrolytic tank was constructed rather creatively by R. Steve Colby, who, along with Alan Presser, performed the tedious macroscopic field measurements of Chapter 4. Prof. Markus Zahn of the University of Florida, who was at MIT for the academic year 1976-77, contributed many helpful discussions about some of the early theoretical aspects of the