AI-Powered Grid Modernisation: NVIDIA and Southern California Edison Join Forces
As utilities worldwide grapple with similar challenges, the NVIDIA-SCE partnership serves as a model for leveraging AI to modernise power grids.
Image for illustration purposes.
Southern California Edison (SCE), a major utility serving 15 million people, faces unprecedented challenges as electric vehicles and smart buildings drive an 80% surge in demand across its vast service area. To meet this growing need, SCE is partnering with tech giant NVIDIA to harness the power of artificial intelligence (AI) in modernising its power grid.
The utility’s latest projections show a 35% increase in demand over the next decade, with data centres and other new sources adding another 20 TWh. To address this, SCE is pursuing a multi-pronged approach: introducing clean energy sources like hydrogen and geothermal, expanding energy storage, and reluctantly upgrading power lines.
However, SCE’s collaboration with NVIDIA offers a more innovative solution. By implementing AI tools, the utility aims to optimise power flow and identify equipment faults, potentially deferring costly infrastructure upgrades. The partnership will develop AI-powered software for various applications, including vegetation management, asset inspections, load interconnection, and incident management.
NVIDIA’s AI models will revolutionise SCE’s operations, from microsecond-level grid optimisation to large-scale network planning. This approach is particularly crucial as the U.S. power grid struggles to keep pace with substantial load growth. Grid operators nationwide are forecasting significant increases in demand, necessitating extensive generation and transmission expansions.
The partnership will leverage NVIDIA’s advanced AI platforms, such as Jetson for smart meter integration, Omniverse for digital twin modelling, and NIM microservices for generative AI. These technologies enable real-time management of local transmission and distribution equipment, including cross-arms, fuses, transformers, and substations.
One key advantage of NVIDIA’s edge AI approach is its ability to deploy AI models directly on local devices, rather than relying on cloud computing. This allows for faster, more efficient data processing and decision-making at the grid’s edge.
The collaboration also aims to enhance SCE’s ability to manage distributed energy resources (DERs) like EV chargers and rooftop solar panels. By optimising power flow, SCE can potentially avoid building new transmission and distribution lines, focusing instead on its goal of delivering 100% clean energy by 2045.
Source: EE Power
#AI#asset management#clean energy transition#digital twin#distributed energy resources#edge computing#electric vehicles#energy storage#Grid modernisation#grid resilience#Load forecasting#machine learning#NVIDIA#power flow optimisation#power grid#renewable energy#smart buildings#smart meters#Southern California Edison#Utility Operations
