The demand for electricity to power AI data centers is skyrocketing, placing immense pressure on traditional energy sources.  

“If we continue pursuing clean energy for AI and data centers, we will need to triple the energy supply for data centers by 2030,” says Woodruff Professor Anna Erickson, a nuclear engineering expert from Georgia Tech. Nuclear power, with its high energy density and continuous operation, is well-suited to provide the steady base load of electricity required. 

According to Erickson, the recent headlines of the restarting of Pennsylvania’s Three Mile Island Unit 1 reactor (TMI-1) could play a crucial role in meeting these demands sustainably. 

This decision, supported by a 20-year agreement with Microsoft, aims to provide carbon-free energy to meet the escalating power demands of AI data centers. The company’s goal to be carbon negative by 2030 aligns with the broader push for sustainable energy solutions.  

According to the United States Energy Information Administration, as of Aug. 1, 2023, the United States has 93 operating commercial nuclear reactors across 54 nuclear power plants in 28 states. The most recent reactor to begin commercial operation is Unit 4 at the Alvin W. Vogtle Electric Generating Plant in Georgia, which started on April 29, 2024. 

The commercial start of Unit 4 completes the 11-year expansion project at Plant Vogtle.

A Historic Site With a New Mission 

Three Mile Island, infamous for the 1979 partial meltdown of its Unit 2 reactor, has remained a symbol of nuclear caution. However, the reopening of TMI-1, which operated safely for decades before its 2019 shutdown due to financial constraints, represents a potential renaissance for nuclear power. The plant’s revival is seen as a strategic move to address the increasing strain on conventional electricity grids, exacerbated by the energy-intensive needs of AI technologies. 

Expert Insights on Safety and Innovation 

Erickson stresses the importance of rigorous safety measures and technological upgrades in the reopening process.  

“Reopening TMI-1 will require addressing several critical safety concerns, primarily focused on aging infrastructure and modern regulatory standards,” she explains. Comprehensive inspections and upgrades to emergency cooling, radiation monitoring, and digital control systems will be essential to ensure structural integrity and operational reliability. 

Erickson notes, “We can expect to see developments in advanced radiation detection, novel sensors, and AI-driven security systems.” These technologies not only enhance safety but also improve the efficiency and reliability of nuclear power plants. She also highlights the potential for innovative advancements in reactor technology.  

Economic and Environmental Implications 

The reopening of TMI-1 is expected to bring notable economic advantages. According to Erickson, upgrading existing infrastructure is likely to be more cost-effective than new construction and can be completed more quickly.  

“The implications of restarting are significant,” she explains. “It supports clean energy goals and provides a reliable power source for the growing needs of data centers.” 

Environmental considerations are also paramount. The plant’s carbon-free energy production aligns with efforts to combat climate change and reduce greenhouse gas emissions.  

“Nuclear energy is a clean and reliable power source that can help us achieve our climate goals while meeting the growing energy demands of AI,” Erickson emphasizes. 

Public Perception and Regulatory Oversight 

Despite the potential benefits, public perception of nuclear energy remains cautious, primarily due to historical incidents like the Three Mile Island accident. Erickson acknowledges these concerns and indicates the importance of transparent regulatory oversight and effective communication.  

She says the U.S. Nuclear Regulatory Commission (NRC) “does a lot to ensure safety and security, but as experts, we need to do a better job of explaining technological advances and the benefits of nuclear energy.”  

The reopening of TMI-1 is subject to approval from the NRC and other regulatory bodies, ensuring that all safety and environmental standards are met. 

Georgia Tech’s Executive Vice President for Research (EVPR) search committee has selected three finalists. Each candidate will visit campus and present a seminar sharing their broad vision for the Institute's research enterprise. 

The seminars are open to all faculty, students, and staff across the campus community. Interested individuals can attend in person or register to participate via Zoom (pre-registration is required).    

All seminars will take place at 11 a.m. on the following dates:  

• Candidate 1: Monday, January 13, Scholars Event Theater, Price Gilbert 1280 (register for webinar)  

• Candidate 2: RESCHEDULED to Wednesday, January 29, Scholars Event Theater, Price Gilbert 1280 (register for webinar)

• Candidate 3: Monday, January 27, Scholars Event Theater, Price Gilbert 1280 (register for webinar)  

Each candidate’s bio and curriculum vitae, along with further details, will be accessible through the EVPR search site 48 hours prior to each visit. Georgia Tech credentials are required to access all materials. Information is being made available in this manner to protect the confidentiality of the finalists. Following each candidate’s visit, the campus community is invited to share their comments via a survey that will be posted on the candidate’s webpage.   

The search committee is chaired by Susan Lozier, dean of the College of Sciences. Search committee members include a mix of faculty and staff representing colleges and units across campus. Georgia Tech has retained the services of the executive search firm WittKieffer for the search.  

When it comes to finding solutions to climate change, there’s no shortage of technologies vying for attention, from renewable energy to electric vehicles to nuclear energy. One such contender, solar geoengineering, is favored by proponents who say it could quickly cool the planet and give the world time to fully implement efforts to limit emissions and remove carbon from the atmosphere.  

But that promise comes with risks, which include potentially poorer air quality or depleted atmospheric ozone – both of which can cause serious health issues of their own. 

A new Georgia Tech School of Public Policy-led study published in the Proceedings of the National Academy of Sciences (PNAS) suggests that while those risks deserve further consideration, solar geoengineering could save as many as 400,000 lives a year through a reduction in temperature-related deaths attributable to climate change.  

“An important question is how the reduction in climate risks from solar geoengineering compares to the additional risks its use entails,” said lead author Anthony Harding of the School of Public Policy. “This study offers a first step in quantifying the risks and benefits of solar geoengineering and shows that, for the risks we considered, the potential to save lives outweighs the direct risks,"  

Harding co-authored the PNAS article with Gabriel Vecchi and Wenchang Yang of Princeton University and David Keith from the University of Chicago. 

The researchers studied a climate change mitigation strategy called stratospheric aerosol injection (SAI), a type of solar geoengineering that involves spraying tiny reflective particles into the upper atmosphere. Those particles would then redirect some sunlight back to space and help cool Earth. 

The authors used computer models and historical data on how temperature affects death rates to see how much solar geoengineering might affect death rates, assuming a 2.5-degree Celsius increase in average temperature from pre-industrial levels and similar approaches to climate change as seen in the world currently. 

They found that cooling global temperatures by 1 degree Celsius with solar geoengineering would save 400,000 lives each year, outweighing deaths caused by solar geoengineering’s direct health risks from air pollution and ozone depletion by a factor of 13. This means that the number of lives saved due to solar geoengineering-caused cooling would be 13 times the number of lives potentially lost from solar geoengineering’s known risks. 

Many of those deaths would be avoided in hotter, poorer regions, the study notes. Cooler, wealthier regions could actually face increased cold-related deaths. 

Solar geoengineering has generated millions of dollars in funding and a recommendation by the National Academies of Science that the federal government should provide millions more toward research and the development of a risk-risk analysis similar to what Harding’s team produced. But the tech has also drawn concern, including from the Union of Concerned Scientists. That group says there’s too much environmental, ethical, and geopolitical risk to proceed without much more research. 

The authors caution that their study is an important starting point in better understanding solar geoengineering's promise and peril but is far from a comprehensive evaluation of the technology’s risks and benefits. 

They say their models are based on idealized assumptions about aerosol distribution, population and income growth, and other factors. They also can’t capture all of the real-world complexities that solar geoengineering would entail. They note their study also does not address all of the potential risks of solar geoengineering, such as possible impacts on ecosystems, global politics, or the possibility governments will rely on the technology to delay politically difficult emissions cuts. 

Still, the researchers say, the study suggests that for many regions, solar geoengineering could well be more effective at saving lives than emissions reductions alone and is worth keeping in the mix as the world searches for the optimal ways to cool our warming planet. 

“There’s no perfect resolution to the climate crisis,” said Harding. “Solar geoengineering entails risks, but it could also alleviate real suffering, so we need to better understand how the risks compare to the benefits to inform any potential future decisions around the technology.” 

The paper, “Impact of Solar Geoengineering on Temperature-Attributable Mortality,” was published in the Proceedings of the National Academy of Sciences on Dec. 17. It is available at https://doi.org/10.1073/pnas.2401801121.