Towards a geo-agnostic, source-agnostic modeling of climate influences on renewable power plant-level generation
Open Access
- Author:
- Chiluveru, Vijay
- Graduate Program:
- Energy and Mineral Engineering
- Degree:
- Master of Science
- Document Type:
- Master Thesis
- Date of Defense:
- June 11, 2024
- Committee Members:
- Renee Obringer, Thesis Advisor/Co-Advisor
Russell Richard Barton, Committee Member
Jeremy Gernand, Program Head/Chair
Mort D Webster, Committee Member - Keywords:
- geo-agnostic
climate change
renewable energy systems
predictive modeling
resilience - Abstract:
- Energy infrastructure is critical to modern society. However, the ongoing climate crisis is already impacting existing energy infrastructure through extreme weather events which are increasingly frequent and intense. In fact, these climate-induced impacts may create roadblocks for the energy transition, particularly if the climate impacts on low carbon and renewable energy technologies are not well-understood. Here, I propose a data-driven methodology to model these complex interactions defining the renewables-climate-risk nexus over large spatiotemporal scales. In particular, this study leverages an open-source dataset containing hydro, wind and solar energy systems across the United States. Using tree-based ensemble learning techniques, it is shown that we can model the non-linear effects of climate variables on these renewable systems. This study further demonstrates the potential of training Random Forests to produce geo-agnostic, source-agnostic models which are aimed to have consistent and comparable performance with respect to sourcespecific modeling. This study and research work is aimed at envisioning a future, in line with the trends of rising renewables in the mix, with the explicit research need to look at common data pipeling and modeling frameworks in developing data-driven models which can be geo-agnostic as well as source-agnostic to aid with the long-term planning and operations of energy systems in a future actively impacted by climate change.