Switch-Mode Power Transformer in a Wave-Powered, Reverse Osmosis Desalination Plant

Abstract

In the reverse osmosis (RO) desalination process, a salt water solution is pressurized to overcome the osmotic pressure across a semi-permeable membrane. A few groups have proposed that a wave energy converter (WEC) having a seawater based, hydraulic power take-off can could be used to pressurize the feedwater for an RO system. However, coupling the wave energy harvesting process and the RO desalination process imposes unique design constraints on the fluid power system, such as pressure limits of conventional RO system components. In this study, a fluid power circuit with a switch-mode power transformer is used to transfer power while keeping the pressure of the power take-off and RO processes relatively decoupled. The switch-mode power transformer studied herein adds fewer costly components and less significant loss mechanisms to the system than a conventional hydraulic transformer performing the same function. The switch-mode power transformer uses the inertia of a hydraulic motor driven electric generator and switching of the hydraulic motor inlet between high and low-pressure sources to decrease the pressure at which power is being transmitted to the RO process. This process is analogous to DC-DC switching power transformers in the electrical domain. This study seeks to demonstrate this unique switch-mode system as a potential solution for coupling the wave-energy harvesting process with the reverse osmosis process. The system is modeled and studied in the context that the transformer and RO system are onshore, 500 meters from the WEC. Power captured from the WEC is transmitted through a long pipeline to shore. A distributed parameter model is used to model the pipeline dynamics, simultaneously revealing the significance of these dynamics and the robustness with which the switch-mode transformer decouples the pressure dynamics at the RO feed from the pipeline dynamics. The switch-mode power transformer is estimated to be 76% efficient while the system, as a whole, is estimated to be 45% efficient.