Build, Repeat. (A Paces Podcast) The AI Power Puzzle: Grid Flexibility vs. Off Grid Microgrids
Apr 15, 2025
Join Kyle Baranko, Head of Product at Paces, and Tyler Norris, a Ph.D. student at Duke University, as they tackle the surge in energy demands from AI data centers. They dive into solutions like off-grid solar-plus-storage microgrids and the need for flexible grid connections. The duo discusses the critical balance between reliability and sustainability, examining how current market frameworks affect scaling innovations. With insights into the barriers facing developers, this conversation is vital for anyone curious about the future of energy and infrastructure.
AI Snips
Chapters
Transcript
Episode notes
Off-grid Microgrids Rival Gas Costs
- Off-grid solar-plus-storage microgrids can supply AI data centers cost-effectively and quickly.
- These systems can rival traditional natural gas setups in price and enable faster deployment than grid connections.
Grid Underutilization Enables Load Growth
- The U.S. power system is underutilized, with average load at 53% of peak capacity.
- Modest curtailment of AI data centers' peak demand could add up to 98 GW load without new generation.
Curtailment Can Preserve Uptime
- Data centers can manage curtailment without reducing server uptime by shifting flexible loads onsite.
- Not all data center infrastructure runs at full capacity; cooling loads can be temporarily reduced to save energy.