CAREER: Computation-efficient Resolution for Low-Carbon Grids with Renewables and Energy Storage

Rochester Institute Of Tech, Rochester NY

Investigators

Abstract

To realize the vision of sustainable power systems, Energy Storage Resources (ESRs) play critical roles in reliable and resilient grid operation. This NSF CAREER project aims to develop novel modeling and optimization approaches for grid operation with sustainable energy and ESRs. The project will bring transformative change to enable power grid operators to leverage ESRs more efficiently for a sustainable and efficient energy future. This will be achieved by ESR modeling, novel formulation tightening techniques, and innovative optimization methods. The intellectual merits include novel ESR market participation models considering dynamic State-of-Charge (SOC) limits and renewable uncertainties; an innovative machine learning-based formulation tightening approach to improve computational efficiency; and an Ordinal Optimization-based optimization approach for exponential complexity reduction to efficiently manage a large number of ESRs in grid operations. The broader impacts of the project include the development of a training module for Independent System Operators (ISO), Regional Transmission Operators (RTO), and software developers, and a training module for graduate and undergraduate students, focusing on students at an early stage in STEM disciplines; and broader outreach activities to K-12 students. The project addresses several technical challenges in low-carbon grid operation with renewable energy and ESRs including ESR market participant models, inconsistency between day-ahead scheduling and real-time dispatch, and computational difficulty caused by unique features of ESRs such as bidirectional discharge and charge operations and time-coupling SOC. The technical components of the project include the establishment of various ESR participant models from self-scheduling to being fully managed by ISOs/RTOs considering dynamic SOC limits; development of a novel convex hull-oriented deep learning-based formulation tightening approach for computational benefits; and an Ordinal Optimization-based optimization approach for exponential complexity reduction to efficiently solve grid operation problems with a large number of ESRs. The resulting models and methods with plug-and-play capabilities can be integrated into ISOs/TROs’ existing platforms developed by vendors for efficient utilization of ESRs, leading to economic and environmental benefits. The results of the project will also facilitate education and outreach activities related to ESRs for a sustainable and efficient energy future. 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.