Date of Award

Summer 8-2025

Level of Access Assigned by Author

Open-Access Thesis

Degree Name

Doctor of Philosophy (PhD)

Department

Civil Engineering

First Committee Advisor

Kimberly Huguenard

Second Committee Member

Richard Kimball

Third Committee Member

Andrew Goupee

Additional Committee Members

Jason Dahl

Daniel Zalkind

Abstract

Recent advancements in floating offshore wind technologies are accelerating the transition toward commercial maturity by reducing costs and enhancing reliability. These innovations extend beyond individual components, addressing the performance and integration of component systems and ultimately the design of entire floating wind farms. This dissertation focuses on system-level innovations that improve the performance, resilience, and cost-effectiveness of floating wind farms, emphasizing predictive modeling and optimized infrastructure design. The dissertation is structured in two main parts, each named after an aquatic creature: DOLPHINN and TRTLE. Part I, DOLPHINN (Dynamics Observation of Long/short-crested Phase-resolved wave Harmonics Integrated with Neural Network), introduces a predictive framework that integrates accessible sensor data with physics-informed, data-driven modeling to forecast wave conditions and system states. This framework supports applications ranging from descriptive to control strategy integration, while supporting anomaly detection, signal filtering, and predictive forecasting. By reconstructing and forecasting spatio-temporal wave fields at locations of interest using upstream measurements, DOLPHINN enables more accurate real-time predictions of floating platform responses. The framework also identifies abnormal system behavior by benchmarking against healthy baseline conditions. Its practical implementation spans digital twin applications such as sensor data pre-processing, continuous model training from field measurements or physics-based simulations, and forecasting of system dynamics. Case studies demonstrate the framework’s ability to predict wave fields, platform motion across six degrees of freedom, and other critical system responses under a variety of conditions, including fault scenarios. Effective deployment of the framework depends on: (1) clearly defining the system’s intended level of autonomy, (2) understanding the underlying cause-effect relationships between input measurements and target system responses, and (3) assessing the practical value and intended use of the predicted outputs. Part II, TRTLE (Turbine Reconfiguration Technique for Layout Efficiency), presents a conceptual design framework for optimizing floating wind farm layouts and mooring configurations at the system-of-systems level. Unlike traditional design approaches that focus on single turbines, TRTLE integrates array-level parameters early in the design process to improve energy yield and reduce costs. For example, wake-dodging strategies that reposition upstream turbines can enhance the farm’s annual energy production (AEP). One key innovation explored is the use of shared mooring systems, which reduce the overall number of components and simplify installation and maintenance. Although shared moorings introduce challenges related to system coupling and parameterization, this study identifies best practices to address these complexities. Preliminary findings suggest that mooring systems can be designed to enable passive or active platform relocation, helping to mitigate wake losses and further increase AEP. Economically, this results in minimal impact on capital expenditure (CapEX), while offering clear advantages in reducing installation complexity and lowering operational expenditure (OpEX).

Download

Files over 10MB may be slow to open. For best results, right-click and select "save as..."

Share