Day 1 - 2 SUYDERHOUD-SINHA RAY-ANDRUS-10-06-25 IFNEC AI Summit.pdf

1、Accelerating Nuclear Power Plant Design,Licensing,and Deployment with AILRS/MIS-25-06246Agenda1.Opportunities and Challenges for Accelerating Nuclear Deployment2.Application Areas for AI in Design,Licensing,and Deployment3.Early R&D at INL4.Looking to the FutureThe Opportunity to Accelerate Long-Ter

2、m Projections for Nuclear Power Increase:U.S.Department of Energy(DOE):100 400 GWe by 2050 International Atomic Energy Agency(IAEA):337 992 GWe by 2050 Need to factor in existing fleet retirement But there is immediate demand from data centers:U.S.DOE:176 TWh 325-580 TWh by 2030(6.7%-12%of total U.S

3、.Energy)International Energy Agency(IEA):500 TWh 945 TWh by 2030The Challenge Nuclear power has a historical mean cost overrun of 120%worldwide Vogtle 3/4 in Georgia Initial$14B,final$30B(+$17B)Initial 2016,final 2024(+8 yrs)Based on current estimated capital costs of nuclear,this would cost anywher

4、e from$2.5 4.2 Trillion in the United States,not accounting for inflation Commercial plant engineering and licensing is a multi-year effort that accounts for 33%of total project costs and up to 50%of new energy deployment time to market How can we reverse this trend and accelerate nuclear plant depl

5、oyment?Opportunities for AI in Nuclear Power Plant Design A systems engineering design process for plant design(Requirements Architecture Detailed Design)Analytical workflow(FEA,multiphysics,neutronics,hazards/safety Analysis)automation with data transformations Plant,system,and component level opti

6、mization across risk&reliability,cost,supply chain,etc.Verification&Validation(V&V)Opportunities for AI in Licensing Acceleration New document generation from existing design engineering data Existing document conversion from one regulatory framework to another Regulatory framework comparison using