Toward cascading failure mitigation in high voltage power system capacitors

Rights statement
Awarding institution
  • University of Strathclyde
Date of award
  • 2022
Thesis identifier
  • T16190
Person Identifier (Local)
  • 201558037
Qualification Level
Qualification Name
Department, School or Faculty
  • As electrical power networks adapt to new challenges, advances in high voltage direct current interconnection offer one means to reinforce alternating current networks with flexibility and control, accordingly improving diversity to become a present-day, viable alternative to network flexibility and energy storage measures. High voltage capacitors support these links and offer simple means of voltage support, harmonic filtering, and are inherent to established and emerging converter designs. Where research literature predominantly explores use of modern dielectrics in efforts toward improved capacitor technologies, but reveals little about: existing capacitor designs; associated failure modes or statistics; or avenues in monitoring or maintenance, simulation modelling equips engineers with an approach to pre-emptively anticipate probable incipient fault locations toward improving designs for systems yet to be commissioned. This Dissertation presents a high-voltage capacitor simulation model, before exploring two questions about these hermetically sealed, highly modular assets: where are incipient faults most likely to arise; and how can internal faults be externally located? Nonlinear voltage distributions are found within each and among connected units, induced through parasitic effects with housings supported at rack potential. Consequent implications are considered on: stresses within unit dielectrics, susceptibility to cascading failure, and an ability to locate internal faults. Corroboration of fault detection and location is additionally found possible using unit housing temperatures. A model is presented, developed to be scalable, configurable, and extensible, and made available for posterity. Opportunities in asset design, modelling, manufacture, and monitoring are proffered toward improvements not only in operational longevity, but in understanding and early awareness of incipient faults as they develop.
Advisor / supervisor
  • Stewart, Brian
Resource Type