Fused Portfolio, Site, and Device Sizing Optimization for Harnessing North Carolina’s Coastal Renewable Energy Resources

The objective of this project is to create an integrated framework that fuses (i) site selection, (ii) number of each type of device at each site, and (iii) device sizing sub-problems into a single tool, simultaneously accounting for the resource models, technical performance models, and cost models. The proposed work will lie at the intersection of two stated interests of the NCROEP, namely (i) resource assessment (via synthetic resource modeling) and (ii) device efficiency improvement (via optimization). In developing the aforementioned integrated tool, our team will build upon independent resource modeling, portfolio optimization, and device performance analysis to create a nested integrated optimization framework. We will use efficient optimization tools to select a portfolio for each candidate site, where each iteration of the portfolio optimization contains a device sizing optimization therein. For the current fiscal year, the team will focus on portfolios consisting of (i) offshore horizontal axis wind turbines, (ii) point absorber wave energy converters, (iii) tethered, stationary turbines, and (iv) underwater kites; however, the tool will enable new technologies to be integrated moving forward. Noting the importance of dynamic behavior in techno-economic performance, previous time-series synthetic data generation based on adversarial neural networks will be enhanced by considering time chronology and correlation among different sites and technologies. High-level mathematical optimization modeling and analysis will be done in Python using the Pyomo library from Sandia National Labs and the neural networks development will be done in Python using the TensorFlow library from Google Brain Team.

The mai project goal can be partitioned into the following tasks: • Develop enhanced synthetic modeling tools capable of recovering the probability distribution of the wind, wave, and ocean current data, accounting for extreme energy generation conditions that can occur over device lifetimes (e.g. 20-30 years). • Develop and run the nested framework of the Figure below (middle), to include consideration for (i) wind turbines, (ii) wave energy converters, (iii) stationary tethered OCTs, and (iv) kites. Based on these tasks, the team will compute optimal site locations and build efficient frontiers (risk x return curve) for different combinations of wind, wave, and ocean current resources; this effort corresponds to the right portion of the Figure below. These efficient frontiers, along with the corresponding synthetic resource modeling and optimization tools, will represent the ultimate project deliverables.

Vermillion, C. (PI), de Queiroz, A.R. (co-PI), Fused Portfolio, Site, and Device Sizing Optimization for Harness North Carolina’s Coastal Renewable Energy Resources, Funded by the UNC Coastal Studies Institute, Renewable Ocean Energy for North Carolina Program, 2022-2023

Analysis Framework - Fused Portfolio, Site, and Device Sizing Optimization for Harnessing North Carolina’s Coastal Renewable Energy Resources