Utilities Rethink Resiliency with AI Planning Tools
- Grid Services & Demand Response News Policy & Markets
Marin Holt
- 0
- 6 minutes read
- Utilities are increasingly using AI to modernize electrical grids and enhance resilience.
- Challenges include aging infrastructure, climate threats, and new energy demands.
- Rhizome, an AI-powered software platform, uses high-resolution intelligence to create risk models for infrastructure assets.
- Machine learning is employed by platforms like Rhizome.
Utilities are increasingly turning to artificial intelligence (AI) to navigate the complex challenges of modernizing their electrical grids and enhancing resilience. Aging infrastructure, escalating climate threats, and a surge in new energy demands are compelling grid operators to make strategic investments, but identifying the most impactful ones has become a significant hurdle.
Mishal Thadani, co-founder and CEO of Rhizome, an AI-powered software platform, highlights this shift. Rhizome utilizes high-resolution intelligence to create detailed risk models for infrastructure assets. By employing machine learning, the platform analyzes geographic, building, and system data to pinpoint potential asset failures and quantify risks, such as wildfire threats, enabling utilities like National Grid to make more informed investment decisions.
Traditionally, utilities have operated in a reactive mode. However, this approach is no longer sufficient. The U.S. Department of Energy’s Electric Emergency Incident and Disturbance Reports show an 86% increase in weather-related outages between 2003 and 2023. This surge, coupled with increasing regulatory demands for quantifiable risk reduction in wildfire mitigation and reliability plans, has pushed utilities to adopt more proactive strategies.
The framework of Risk-Spend Efficiency (RSE) has emerged as a critical tool. RSE allows utilities to compare the financial investment in grid improvements against the actual reduction in risk achieved, providing a transparent and auditable method for prioritizing a portfolio of mitigation strategies. This approach sits at the nexus of grid resilience and utility risk management.
Despite the growing need for sophisticated planning, many utilities still rely on outdated methods, such as static spreadsheets and incomplete data. The sheer volume of data generated by detailed wire- and pole-level analyses can overwhelm planners, making it difficult to derive actionable insights. In some regions, these reports are crucial for regulatory funding allocations and can take months to prepare, undergo multiple reviews, and may even require revisions. This data overload, combined with stringent regulatory requirements, can transform resilience planning into a time-consuming compliance exercise rather than a strategic decision-making process.
AI-driven platforms like Rhizome are addressing these data roadblocks by integrating near-real-time data with advanced analytics and intuitive visualizations. This capability transforms the RSE framework from a mere compliance task into a powerful decision engine, enabling utilities to make timely, defensible infrastructure decisions and build more reliable systems for the future.
Why It Matters
This development signifies a move towards more data-driven grid management, impacting the broader energy industry by enabling more precise allocation of capital. As weather-related outages have climbed 86% from 2003 to 2023, utilities are compelled to adopt advanced analytical methods to optimize their multi-billion dollar investments in grid hardening and wildfire mitigation.
Critical Perspective
The article notes an 86% increase in weather-related outages between 2003 and 2023, a statistic that underscores the urgency for grid modernization. However, relying on AI platforms like Rhizome to quantify risks, as National Grid is doing, echoes the promises made by companies like Palantir with their predictive maintenance software for infrastructure. History shows that similar technological promises, such as those surrounding smart grid initiatives in the early 2000s, often faced significant implementation hurdles and did not fully deliver on their projected efficiency gains. Given the substantial investments being made, are these AI tools truly optimizing for long-term grid stability, or are they merely accelerating a cycle of expensive, technology-driven upgrades that could become obsolete just as quickly?
