232 IEEE JOURNAL OF OCEANIC ENGINEERING, VOL. 30, NO. 1, JANUARY 2005 North East Pacific Time-Integrated Undersea Networked Experiments (NEPTUNE): Cable Switching and Protection Mohamed A. El-Sharkawi, Fellow, IEEE, Aditya Upadhye, Shuai Lu, Harold Kirkham, Senior Member, IEEE, Bruce M. Howe, Tim McGinnis, and Phil Lancaster Abstract—The objective of the North East Pacific Time-In- tegrated Undersea Networked Experiments (NEPTUNE) is to establish a permanent, subsea observatory surrounding the Juan de Fuca tectonic plate. To achieve this objective, a special power distribution system is designed to provide continuous power to science equipment, vehicles, and laboratories located as deep as 5 km below the water surface. The NEPTUNE power system is significantly different from terrestrial power systems in many aspects and it requires different switching, protection, and control strategies. In this paper, we address the design of system switching and fault isolation equipment. Index Terms—Circuit breakers, fault diagnosis, power elec- tronics, transient analysis, underwater cables. I. INTRODUCTION T HE vast oceans of the world have not yet been explored completely [1]. Present technology allows the deployment of battery-operated instruments, low-power instruments pow- ered from shore, autonomous underwater vehicles (AUVs), or remotely operated vehicles (ROVs) that can explore the deep ocean floor and carry out scientific experiments. All of these methods are power-constrained. In shallow waters, the ROV is powered by a generating unit on the ship through a flexible ma- rine cable. However, the cable weight and length makes the op- eration of the ROV in deep waters an extremely difficult task. An alternative method is to outfit the ROV with batteries. However, the weight, volume, and the limited energy storage capacity of the batteries restrict the mission of the ROVs to short periods. Carrying out scientific experiments on the ocean floor for ex- tended time is a challenging task that demands a permanent source of energy to create a permanent observatory. This is the main task of the power system design of NEPTUNE. The topology of the planned North East Pacific Time-Inte- grated Undersea Networked Experiments (NEPTUNE) power Manuscript received September 22, 2003; accepted August 26, 2004. This work was sponsored by the National Science Foundation by Grant OCE 0116750 “Development of a Power System for Cabled Ocean Observatories.” Associate Editor: J. Lynch. M. A. El-Sharkawi, A. Upadhye, and S. Lu are with the Department of Elec- trical Engineering, University of Washington, Seattle, WA 98195 USA (e-mail: elsharkawi@ee.washington.edu). H. Kirkham is with the Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA 91109 USA. B. M. Howe and T. McGinnis are with the Applied Physics Laboratory, Uni- versity of Washington, Seattle, WA 98195 USA. P. Lancaster is with Alcatel Submarine Network, Greenwich, SE10 0AG London, U.K. Digital Object Identifier 10.1109/JOE.2004.839938 system is shown in Fig. 1. It is a fiber-optic/power cable net- work around and across the Juan de Fuca tectonic plate off the west coast of North America [2]. The NEPTUNE power system consists of a backbone circuit that covers the entire service area, and branching circuits to reach specific sites. The backbone net- work is comprised of 3000 km of cable connecting about 30–40 evenly distributed branching units (BU). The red circles in Fig. 1 show some of the system BUs. These BUs can be viewed as the switching yards in terrestrial systems. The branching circuits are not shown in the figure, but they are cables connecting the BUs to the load nodes on the seafloor. The length of the branching cables can be as long as 100 km. Each node provides standard power to the scientific equipment, and internet communication interfaces between the scientific equipment and shore. The com- munication network has a capacity of about 10 Gb/s, and the power network delivers 200 kW with a design life of 25 yr [3]. The main design features of NEPTUNE power system are as follows. 1) The total capacity of the network is 200 kW provided by two shore stations: one will be located on Vancouver Is- land, Canada; and the other on the Oregon coast, USA. Each of these stations is capable of providing 100 kW. A redundant power supply will be located at each of the shore stations. 2) The cable will have a voltage rating of 10 kV. The resis- tance of telecom cable is around 1 , so that for high currents and distances of a few hundred km, the cable volt-drop can approach the voltage rating of the cable. Hence, a current rating of 10 A has been set for the power network. 3) The submarine cable is to be selected from among the existing conventional submarine telecommunications ca- bles. This is done to reduce the cost of the backbone system. The conventional submarine cables are highly re- liable and are being used in subsea telecom systems all over the world. Fig. 2 shows one of the proposed cables. The cable is to serve a dual purpose; its hollow core car- ries fiber optics for communications, and its copper sheath is used to transmit the electric power. 4) The cable network is dc to reduce the effect of cable ca- pacitance [4]. 5) The ocean provides the return path for the current, so a single conductor cable can be used. 6) The anode of the system must be at the shore station to limit the corrosion of the equipment in deep water. Hence, 0364-9059/$20.00 © 2005 IEEE