Author: Daniel Hou, Graphics: Sylvia Tirado
The BRB Bottomline: As hurricanes strip power grids of electricity, AI’s ability to pinpoint erratic weather patterns has become key.
Hurricanes Helene and Milton recently ravaged power grids in the southeastern US, leaving millions without electricity and causing billions in damages. As climate change exacerbates extreme weather, AI’s ability to pinpoint erratic patterns has taken center stage.
Guarding the Grid: Preemptive Patching
A few months ago, Hurricanes Helene and Milton swept the Southeast, raking in economic losses at an estimated $225–$250 billion. Negligence in planning, maintenance, and foresight all played a role in the disaster. Entire portions of Florida’s power grid were torn apart as Helene plunged 3.4 million homes and businesses into darkness.
In the aftermath of these storms, the vulnerability of power grids was more evident than ever. Ferocious winds and relentless rainfall deprived millions of electricity, incurring damages projected to cost tens of billions of dollars. With climate change accelerating, natural disasters like hurricanes have become more frequent and severe. Compounding the immeasurable humanitarian loss, the economic impacts of these hurricanes have reverberated beyond the energy sector and disrupted power-intensive industries like manufacturing, tech, and retail.
Though damages caused by hurricanes are often unpredictable, many of these recent outages could have been prevented. For energy companies, the challenge lies in predicting where weak points in power grid systems will appear before disaster strikes. Sprawling, aging infrastructure, and limited resources complicate the process. But what would once be deemed an impossible task is now feasible through predictive artificial intelligence (AI), a set of tools with the capacity to transform approaches to grid resilience. Data companies like Rhizome have led the charge, harnessing AI to help utilities infrastructure adapt to dynamic environmental challenges.
AI’s Applications and Advantages
Rhizome is part of a growing movement of businesses leveraging AI to help utilities adapt to the risks of a warming world. By employing machine learning models, Rhizome and similar companies can process vast datasets—weather conditions, structural reports, and even past grid failures—to rapidly identify areas that are more susceptible to failure in extreme weather conditions. Predictive AI models can additionally offer practical implementations of “hardening the grid,” including reinforcing power lines in hurricane-prone areas or upgrading transformers that have a history of overheating in extreme heat waves. This selective targeted approach can reduce the frequency of blackouts and also lower the cost of post-disaster repairs by funneling resources where they are needed most.
The economic advantages of AI-based solutions to power grid failures in post-disaster environments are manifold. By preventing outages, companies can circumvent the billions in losses incurred by disruptions in electricity supply. By employing AI-enabled grid management protocols in cities where natural disasters have become commonplace, companies ensure that their businesses—particularly those in sectors reliant on continuous power like data centers and hospitals—can operate smoothly despite extreme weather conditions.
Environmental Trade-off: Worth It?
While AI capabilities have tremendous potential upside in areas affected by extreme weather, the technology also brings alternative environmental challenges. Training large AI models is immensely energy-intensive, with a single model capable of emitting around 300,000 kilograms of CO2. Moreover, data centers—where AI models are traditionally housed—consume between 1% and 1.5% of global electricity, a figure set to rise as AI usage expands. To meet rising demand, data centers’ energy consumption is set to surge from 3-4% to an estimated 11-12%, reaching approximately three times their present levels by 2030.
Can AI truly combat climate change while dwarfed in the shadow of its immense carbon footprint?
Balancing Cost with Benefit
While AI-incited emissions are an unavoidable issue, progress in renewable energy and data center efficiency is helping to offset them. More data centers are now powered by clean energy, and innovations in sectors like energy efficiency are already showing promise.
Beyond power grids, AI has contributed to climate solutions in surprising ways. In the aviation industry, for one, AI models are reducing the environmental impact of planes by minimizing aircraft contrails, a significant source of greenhouse gas emissions. In manufacturing, AI designs inspired by biological patterns and molecular structures are making industrial processes more efficient, optimizing sustainability across industries.
AI is also being deployed to predict and mitigate climate risks in vulnerable regions. In Africa, the UN’s International Climate Initiative uses AI to forecast extreme weather events, offering early warnings to communities in Chad and Sudan. Stanford researchers have even leveraged AI to connect rising temperatures with global warming trends, offering critical insights into how future heat waves may unfold.
Perhaps most crucially for our current conversation, AI is monitoring polar ice melt, which directly influences sea levels. Understanding these dynamics helps scientists predict coastal erosion, flooding, and storms—all of which place immense pressure on power grids and infrastructure globally.
AI’s Critical Role in Climate Resilience
While AI may incur environmental costs, its capacity to protect infrastructure and predict disasters scales far beyond these drawbacks. Today, companies like Rhizome are already modeling climate threats and actively deploying predictive tools to save lives and resources. In the future, AI may advance to assess complex climate risks holistically, integrating real-time data from vast networks of satellites to offer increasingly precise real-time insights.
As natural disasters like Hurricane Helene increase in both frequency and lethality, AI looks to take a grander role beyond simply forecasting the weather. It may very well offer us an unprecedented understanding of these disasters’ origins—a glimpse into the eye of our planet’s storm.
Take-Home Points
- Recent hurricanes Helene and Milton dredge up concerns about natural disasters and their potential to wreak costly havoc.
- Utility damages form a substantial financial burden for the federal government, but are often avoidable.
- AI emerges as a tool to pinpoint infrastructure vulnerabilities and provide targeted solutions preventing widespread outages.
- Although AI’s environmental costs are significant, advancements in renewable energy and efficiency are reducing its carbon footprint.
- AI’s role extends beyond energy grids, offering critical solutions in emissions reduction, climate prediction, and industrial sustainability.