Machine Learning-Enhanced Optimization for the Real-Time, Secure, and Efficient Operation of the Energy-Water-Hydrogen Nexus

Given the escalating urgency of climate change, our research addresses the imperative need to curb carbon emissions, particularly from the electricity sector, which accounts for a quarter of total emissions in the U.S. We propose integrated systems, such as the energy-water nexus (EWN) and an innovative concept known as the energy-water-hydrogen (EWH) nexus, which integrates renewable energy sources (RESs) with green hydrogen production through water electrolysis. These concepts align seamlessly with the nation’s commitment to global climate agreements, including the Paris Agreement. Furthermore, green hydrogen—a clean and efficient energy source produced via water electrolysis powered by renewable energy—emerges as a pivotal solution to addressing climate change challenges. These innovative engineering concepts show promise in bolstering greater renewable energy integration, ensuring the secure operation of water and power systems, and managing energy while reducing operational costs. However, the complexity of power, water, and hydrogen systems intensifies when collaborating, particularly with the high penetration of intermittent RESs. Advanced AI-based optimization methods are crucial to addressing these intricacies and uncertainties. Our research focuses on applying advanced optimization techniques, such as convexification and machine learning, to efficiently solve these complex problems across various concepts, from virtual energy storage and virtual power plant applications to secure operations and real-time optimal operations of the EWH nexus.