Multiport Energy Router based Integration of Renewable Energy Sources for Offshore Grid
Abstract
Integrating Renewable Energy Sources (RES) is a potential solution to improve the reliability and cost of the electrical offshore architecture for harnessing deep-water hydrocarbon resources. However, conventional subsea electrical architectures based on HVAC transmission and distribution require bulky line frequency transformers and have reactive power losses in the system that increase with the longer step-out distance. In turn, subsea HVDC cables are more efficient, lighter, and cost-effective per unit of transmitted power for long distances than HVAC cables. Furthermore, integrating RES via HVDC or MVDC systems requires voltage regulation without reactive power compensation. However, a DC microgrid with several power sources needs a proper architectural design. To address this issue, the dissertation proposes an HVDC offshore architecture that utilizes wind turbines, battery storage systems, and an onshore grid to power subsea oil and gas loads. The architecture is based on the proposed Multiport Energy Router (MER), a modular DC-DC converter that interconnects the sources above. MER is realized by a series-parallel combination of multiport DC-DC Triple Active Bridge (TAB) converters. The converters are critical to enhance the proposed system's power density, reliability, and efficiency. The TAB modules in MER have an inherent capability of reaching zero voltage source (ZVS) switching due to the port inductor or leakage inductor transformer. This feature assists in eliminating the turn-on losses along with the EMI issues. However, the soft-switching can be compromised depending on port voltages and operating power region variation. To address this issue, a control algorithm for efficiency optimization of TAB by selecting optimal control variables is proposed. The algorithm improves the TAB converter efficiency by extending the ZVS range and minimizing the circulating current in the transformer windings. The TAB converter in MER is a multi-input multi-output (MIMO) system. The conventional PI controller for TAB utilizes a static decoupling matrix, which introduces potential operational challenges for the converters. The dissertation proposes a Model Predictive Control (MPC) for TAB, which is preferred for MIMO systems. The proposed MPC controller can handle fast dynamics, incorporate various constraints, and have straightforward digital implementation.