The Department of Energy (DOE) has highlighted the potential benefits and risks associated with using artificial intelligence (AI) to manage the U.S. electric grid in a recent report. According to the DOE, AI has the ability to reduce emissions, lower costs, and improve grid resiliency, but it also poses significant risks if deployed without proper consideration.
In their report titled “AI for Energy: Opportunities for a Modern Grid and Clean Energy Economy,” the DOE outlined priority use cases for AI in managing the U.S. electrical system, which include grid planning, permitting and siting, operations and reliability, and resilience. Additionally, the DOE identified four broad categories of grid risk in their report titled “Potential Benefits and Risks of Artificial Intelligence for Critical Energy Infrastructure,” which include adversarial attacks against AI systems, unintentional failures of AI models, the use of AI for cyber or physical grid attacks, and supply chain compromises.
According to Puesh Kumar, the director of DOE’s Office of Cybersecurity, Energy Security, and Emergency Response, artificial intelligence holds both incredible promise and potential challenges for the U.S. energy sector. The assessments published by the DOE were directed by an executive order signed by President Biden, which aimed to identify core principles regarding the development of AI and assess its threats to critical infrastructure.
One particular risk highlighted in the report is the potential for “poisoning attacks,” where an attacker could alter the data used to train an AI model in order to manipulate its behavior. This could lead to misunderstandings in normal energy system operations or prevent the detection of equipment maintenance needs.
Despite the risks associated with AI, the DOE believes that the technology has the potential to address climate change and improve grid resiliency. Secretary of Energy Jennifer Granholm emphasized that AI can help solve some of the toughest challenges faced by the energy sector, and the agency is accelerating its AI work to stay globally competitive and manage the increasing energy demand.
As the U.S. aims to achieve a carbon-free electric system by 2035, the DOE sees AI as a crucial tool in managing renewable and distributed generation. AI can help better maintain generation assets, forecast non-dispatchable generation, and improve operational safety. By predicting outages and optimizing the dispatch of repair crews, AI can help utilities restore power more efficiently and enhance grid resilience.
To support AI initiatives in the energy sector, the DOE announced investments in AI-powered tools for clean energy infrastructure, the establishment of a working group on powering AI and data center infrastructure, and upcoming discussions with key stakeholders on the impacts of load growth.
Experts predict that the global power demand created by AI, excluding China, could reach 13.5 GW to 20 GW by 2028, with a significant portion of that demand expected to be in the United States. Despite the challenges posed by the growing power demands of AI, the DOE believes that by leveraging AI, they can address critical needs such as enhancing cybersecurity, developing new battery materials, and advancing grid-scale storage solutions.
Overall, the DOE sees AI as a transformative technology that, if managed properly, can help usher in a cleaner, more efficient, and resilient energy future for the United States.
In recent news from Utility Dive, the US Department of Energy is exploring the use of artificial intelligence (AI) to manage risks in the electric grid. This technology has the potential to improve grid reliability and efficiency, helping to prevent outages and other disruptions. This development could be a major step in modernizing and securing our power infrastructure.