Determine the Optimal Diameter of Carbon Steel Pipes Around Undersea Power Cable

• Main Author: Lau, Alvin Kouk Shenn
• Format: Thesis
• Language: English
• Published: Universiti Malaysia Sarawak 2021
• Subjects: TK Electrical engineering. Electronics Nuclear engineering
• Online Access: http://ir.unimas.my/id/eprint/36603/1/ALVIN.pdf; http://ir.unimas.my/id/eprint/36603/

This work is on the experimental investigation into feasibility of underwater power transmission without using the traditional cross-linked polyethylene (XLPE) technology. It utilizes the vast advantages of overhead transmission lines over XLPE based cables by suspending bare high voltage cables within submarine pipes at the star point of insulators which are spaced to minimize sag. The technology of submarine pipes has been well established by the oil and gas industry, so this research furthers that development by utilizing them for electric transmission. This research is focused on determine the optimal radius of carbon steel pipe where there is no eddy current loss when electric cables suspended inside the pipe. Eddy current plus joule heating levels were made by measuring the heat level on different radius of the pipes for various changes in parameters such as current, voltages in alternating current (AC) and direct current (DC). An open-sourced machine learning algorithm was created to allow the machine to analyse more result in the future by different user and different parameter value, which produce more accurate pipe radius calculation. The result significantly shows that with 50 A electricity carrying cable suspended inside a 1.11 cm pipe radius, the temperature of the pipe increased by 27.6 °C while no temperature changes when the optimal pipe were chosen for different value of electric current. The conclusion of this research shows that using DC and high voltage are the optimal mode of power transmission because power loss through heat energy is minimal compared to AC. Keywords: Undersea Power Cables, cross-linked polyethylene, Eddy Current, Machine Learning Algorithm