PIRE: Multi-Domain, Multi-Scale, Policy-Aware Digital Twin for Offshore Wind Energy Infrastructure

For the US to achieve the offshore wind goals of 30 Gigawatts by 2030, approximately 2000 offshore wind turbines (OWTs) need to be installed in the coming years. Only 7 are currently operating in US waters. For comparison, close to 6000 foundations are operating in European waters.

The size, expense, and importance of OWTs to mitigating climate change necessitates to consider them as civil infrastructures. Such infrastructures must be built to last for over 50 or even 100-years. However, a conception of OWTs as infrastructure has not caught up with their rapid growth.

Most OWTs are typically designed for a 25-to-35-year service life. In a few decades, the industry will face a consequential decision-making challenge as to whether to decommission, rebuild, or retrofit these assets. This challenge is not limited to the US.

It has global implications considering the global expansion of offshore wind energy. Here, the team develops a joint modeling framework for decision making about OWT safety, operation and maintenance, life extension and design. International collaboration is essential because the developers, designers, and operators for offshore wind energy farms are almost entirely from Europe.

Leveraging the European experience from their international collaborators, the researchers use both quantitative and qualitative data collected directly from OWTs, the workers who service them, and the decision makers who enable their development. Their goal is to find solutions to improve the resilience of offshore wind turbines, notably while confronted to more frequent extreme weather events.

By improving OWT service life, this project paves the way to efficiently develop the US clean-energy infrastructures. The project also provides support and training to 1 postdoctoral associate, and graduate and undergraduate students notably from underrepresented groups in science and engineering.

More specifically, the team develops an extensible and customizable joint modeling framework for policy and safety aware digital twins. They use a physics-data-policy-safety co-modeling paradigm, integrated with the help of Agent Based Models. A digital twin is a computational model of an actual OWT (or systems of OWTs) that is maintained and updated based on measured data.

The proposed framework is formulated and studied within the context of the Block Island Wind Farm in Rhode Island state waters, the Coastal Virginia Offshore Wind Pilot Project, in US federal waters, and the Levenmouth Demonstration Turbine in the United Kingdom. It enables short, medium, and long-term modeling, learning, and assessment of the offshore wind farms.

It uses measured data and integrate with policy, labor, and safety aspects. The multi-domain, multi-scale nature of the renewables and related structural elements is modeled by a physics-based Bayesian Assimilation Framework. It is complemented by data-driven machine learning and transfer learning.

The project includes national and international stakeholder engagement programs to facilitate a diverse and inclusive co-production of knowledge. The project has integral educational components including K-12 outreach, professional trainings, and development of graduate students within a transdisciplinary research environment to support the future workforce needed in the offshore wind industry.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.