Infrastructure, operations, and circuits design of an undersea power system

Natural processes in the oceans take place in an episodic or intermittent fashion: massive storms, submarine volcanoes, earthquakes, marine mammal feeding and hunting patterns, for example [1]. Traditionally, the oceans have been studied with ships, satellites, and instrumented moorings, all of which have been limited by time, space, power and communication capabilities. The NEPTUNE program will deploy a regional cabled ocean observatory on the seafloor of the northeast Pacific Ocean, enabling the continuous study of the ocean processes over this large region. This dissertation investigates the approaches to designing the NEPTUNE power system.; Located on the seafloor, the NEPTUNE power system poses a number of design challenges: it requires high reliability and compact sizes, cannot use commercial off-the-shelf components for power conversion and protection, lacks measurements to identify topology changes and to locate faults, and has no communications available to assist in the system startup. Solutions to these challenges are proposed in this dissertation.; First, a novel backbone circuit configuration aimed at increased reliability is described. Based on this configuration, the system operation modes are presented. An automated and coordinated protection scheme, which does not require dedicated communication capability between protection units, is presented. Then algorithms to detect an opened switch and to detect and locate a fault in an interconnected power network are proposed. Appropriate models and approaches for analyzing the various types of stability problems in a large scale do power system are proposed or summarized. The operation design and implementation circuits for the branching unit system and the science node startup system are presented and their functionality is verified through lab tests.; With increasing research interest on the Earth's oceans, similar observatory systems will be needed and constructed. The solutions proposed in this dissertation address the typical constraints and difficulties in building power systems for this type of observatories. They may find more applications as scientists conceive methods to explore the ocean environment.