Energy innovation for a power-hungry world

In the high desert of southwestern Utah, a crane lowered a 60-foot-long metal tool into a well, heading for fractured rock a mile deep where scientists are refining techniques for Earth’s heat to generate electricity almost anywhere. 

Sarah Sausan, a PhD student on a team led by Stanford energy scientist Roland Horne, watched from the well site as the device descended through layers of rock to depths where temperatures reach over 400 degrees Fahrenheit. 

Sensors housed within the tool measure levels of chloride seeping from fractured rock and enable Horne’s team to map how fluids move through the well. This kind of map helps geothermal energy engineers decide where to drill and how to manage circulation so water picks up the right amount of heat to drive a turbine and generate electricity. 

“You want a wider distribution of flow over a larger volume of rock to get the most out of your system,” said Horne, a professor of energy science and engineering at the Stanford Doerr School of Sustainability.

Horne’s work at the U.S. Department of Energy’s geothermal field lab in Utah, conducted in collaboration with Sandia National Laboratory in the summer of 2024, is part of a push by researchers to make reliable, clean, and affordable energy a reality for all. Among other efforts, Stanford scholars are expanding geothermal power beyond natural hot spots, converting agricultural waste into fuel while trapping carbon, collaborating with farming communities to deploy solar panels among crops, and finding ways for the existing power grid to absorb more renewable energy and meet more demand than utilities once thought possible. 

“We need to innovate to eliminate the green premium, making affordable clean energy available to everyone,” said Stanford materials scientist William Chueh, director of the Precourt Institute for Energy. 

Rising demand and prices

The need is urgent. Oil and gasoline prices have soared in 2026 amid supply disruptions tied to the war in the Middle East. Average U.S. residential electricity prices have climbed more than 4% since last year and more than 30% in the last five, straining household budgets. At the same time, extreme weather events and aging grid infrastructure threaten to make energy less reliable for tens of millions of U.S. households.

Solar and wind have become the cheapest options in many cases to add new generating capacity. Although 750 million people worldwide still lack access to electricity, many developing countries are on track to achieve nearly universal access by 2030.  

Related: Study finds high energy use provides little benefit for health and well-being in richer nations

Fossil fuels still supply most of the world’s primary energy. Emissions from burning them continue to drive climate change and intensify extreme weather events such as heat waves, wildfires, hurricanes, severe winter storms, and floods. These events can spike demand for heating and cooling, damage energy infrastructure, and lead to power outages. According to the International Energy Association, energy production and use account for three quarters of greenhouse gas emissions globally.

In the United States, electricity demand has been rising since 2020 after nearly two decades of slow growth. “This wasn’t much of a surprise because it was really a national priority to bring manufacturing on shore. We were beginning to electrify the vehicle fleet. But at the same time, something else emerged: demand for electricity to power data centers,” said Sally Benson, the Precourt Family Professor of energy science and engineering in the Doerr School of Sustainability. 

“Adding huge loads very quickly is always challenging,” Benson said. “As we add these large new generators to meet data center demand, it provides an opportunity to upgrade the wires going into those facilities and add energy storage. Not only can these benefit the data center operators, they can also benefit communities more broadly.”

Expanding geothermal energy

Data from Horne’s project in Utah, paired with simulations in his lab on campus, have already informed advances in enhanced geothermal systems, or EGS, which adapt fracking and horizontal drilling techniques developed for natural gas and oil fields. These systems generate electricity around the clock without emitting greenhouse gases. 

Conventional geothermal power plants rely on rare, natural combinations of heat, water, and permeable rock that are common near the surface in volcanic areas, but relatively scarce across the globe. EGS aims to engineer these conditions deeper beneath the surface and in many more places.

Ongoing work on EGS could help accommodate surging demand from data centers, said Horne, the Thomas Davies Barrow Professor. “AI farms need 24/7 power and that makes them very well matched for geothermal, which wants to run constantly as well,” he said.

Related: The future of geothermal for reliable clean energy and highlights from the 51st annual Stanford Geothermal Workshop

Harvesting electricity and crops

At an educational farm on campus and in California’s Central Valley, ecologist Aidee Guzman and collaborators are working to harvest electricity alongside crops as part of a project supported by the Stanford Sustainability Accelerator, which is based at the Doerr School of Sustainability.

Starting with experimental plots at the O’Donohue Family Stanford Educational Farm, the team is constructing rows of shading structures to test how shade affects tomatoes and other crops grown extensively in the Central Valley. 

Installing photovoltaic solar panels in agricultural lands, which make up nearly 40% of the planet’s land surface, can offer myriad benefits. Locally, these “agrivoltaic” installations can provide shade for plants and animals, reducing heat stress. They can also slow water evaporation and give farmers access to locally sourced, clean energy and a new revenue stream. More broadly, agrivoltaics offer a possible route to reducing competition between renewable energy development and food production.

“There are still unknowns and we’d like to know more, but agrivoltaics is super-promising. Recent research suggests that if you covered 1% of the world’s cropland with agrivoltaics, it would solve global energy needs,” said Guzman, an assistant professor of biology in the School of Humanities and Sciences.

Related: Fast facts about solar energy from the Precourt Institute’s Understand Energy Learning Hub

In the coming years, Guzman’s team plans to take what they’ve learned to agricultural fields in Allensworth, a historic town of fewer than 700 residents outside of Bakersfield, where community-led organizations are working to transform the local economy and food landscape, partly through new approaches to farming.

Through the Accelerator project and other work in her lab, Guzman’s team is working to understand how diminished sunlight affects crop productivity and carbon storage in soil. “If you’re reducing photosynthesis via the shade of the panels, you’re potentially reducing carbon flow down below ground,” Guzman said. “We want to know more about the effects of shade on different crops and on carbon and soil dynamics, which are crucial to keeping plants healthy.”

“We have an opportunity with Allensworth to utilize this technology as a way to reach their community goals of energy and food sovereignty, while also contributing to the science that will inform what’s going to happen when these agrivoltaic systems emerge on a larger scale,” said Gisel De La Cerda, a PhD student in Guzman’s lab and member of the Accelerator project team. 

By working closely with residents and farmers to understand their needs and concerns, Guzman and De La Cerda hope to provide a model for earning local support for clean energy development. “Community engagement is a central tenet to how we all move forward,” Guzman said.

Related: Stanford’s farm goes fully electric

Turning waste into fuel

Agriculture already faces pressure to produce food for a growing global population as rainfall patterns shift and droughts grow more extreme.

Every year, farmers discard more than 5 billion tons of biomass left over from harvests of staple crops such as wheat, rice, corn, and sugarcane. These discarded stalks, husks, chaff, and other plant parts contain carbon absorbed from the air through photosynthesis. 

Farmers often burn the leftover biomass to dispose of it. The method is quick and relatively cheap, but it releases harmful pollution along with carbon dioxide and other greenhouse gases. Even composting it or leaving residues to decompose in the field releases carbon dioxide and methane, a potent greenhouse gas, into the air.

Former Accelerator postdoctoral fellow Divya Chalise has developed a process to convert agricultural waste into biochar, a charcoal-like material that can be burned like coal for energy or added to soil to promote water retention and nutrient uptake. The material stabilizes carbon compounds, potentially reducing the amount that enters the atmosphere as carbon dioxide.

Chalise’s interest in energy traces back to experiences growing up in Nepal, which does not have fossil fuel reserves. As a teenager, Chalise said he saw many young people from his community leave the country for work opportunities in the oil-rich Persian Gulf. “Fossil fuels are unevenly distributed throughout the world,” he said. “Places that have them are often rich, whereas places that don’t often are poor.” 

Related: Decarbonization improves energy security for most countries, study finds 

The conventional process for producing biochar at industrial scale involves gradually heating organic matter to high temperatures in an environment with little oxygen. With support from the Accelerator and the Stanford High Impact Technology Fund, Chalise created a reactor that can produce biochar at lower than conventional temperatures without specialized vacuum chambers, cutting energy requirements and reactor costs. “The biggest thing with biochar is just being cost competitive, and then all the other good things follow,” he said.

The new process also avoids gasifying the biomass, thereby capturing more carbon compared to conventional methods, Chalise said. With Doerr School of Sustainability dean Arun Majumdar and Accelerator faculty director Yi Cui, who advised Chalise during his two years as a postdoc at Stanford, Chalise has now founded a startup to commercialize the technology. 

One of the next steps is verifying how much carbon stays in the biochar when it’s made and added in soil, and for how long. Researchers including Stanford’s Kate Maher, a professor of Earth system science, are working to standardize measurement and verification of carbon claims, an unsettled regulatory area. A study from Maher’s group published in 2025 found that current standards may underestimate the carbon storage potential of many biochar projects, and proposed a new process to accurately account for factors like local climate and soil composition that can affect biochar’s long-term durability.

Related: Biochar: Bridging a gap in carbon removal strategies

Squeezing more from the existing grid

Civil and environmental engineer Ram Rajagopal focuses on keeping electricity flowing reliably through the grid as the energy mix shifts. His research suggests that existing infrastructure could handle more intermittent renewable sources like solar and wind without an expensive buildout of new backup power plants and transmission lines if utilities use real-time data and updated software. 

“The grid can hold a lot more than we thought,” Rajagopal said.

Rajagopal and colleagues are looking for ways to reduce data centers’ impacts on the grid. One problem is fast-varying loads. Once servers finish a task, their power needs can drop by 80% in microseconds. Batteries and other forms of storage could help smooth out these swings, Rajagopal said, reducing strain on the grid and the need for expensive power from rapid start-up gas generators when demand outstrips supply. 

Related: The hidden transmission capacity of the American West

Another approach to balancing out electricity supply and demand is helping data centers coordinate with local energy supplies through “grid aware” systems, which distribute computing loads more efficiently, Rajagopal said. Like battery storage systems, these software solutions could help utilities stave off use of “peaker” power plants.

Simona Onori, an associate professor of energy science and engineering, is also researching ways to get more value from energy storage systems already in use. As part of a project supported by the Sustainability Accelerator, Onori’s team is developing new technologies to accurately estimate the state of charge of lithium iron phosphate batteries deployed for utility-scale energy storage projects. 

This type of battery is a common choice for grid storage because it is relatively stable, can survive many charge and discharge cycles without degrading, and costs less than alternatives that rely on expensive materials like nickel and cobalt. But the voltage remains nearly flat across most of the battery’s operating range. This leads to unreliable estimates of how much charge the battery has left, which is a critical liability for grid operators who must make day-ahead pricing decisions and real-time dispatch calls based on accurate knowledge of available energy. 

“Accurate knowledge of a battery’s true state of charge allows operators to maximize revenue, increase usable capacity, optimize dispatch strategies, and reduce conservative operating margins,” Onori said.

Chueh, whose lab studies the fundamental chemistry, materials, and physics that underpin advanced energy storage technologies, is involved with another battery effort. Together with dozens of researchers from 15 institutions, he is part of the Aqueous Battery Consortium led by Stanford and SLAC National Accelerator Laboratory, which aims to develop a new kind of battery that makes a major component mostly from water. 

According to project director Yi Cui, the Fortinet Founders Professor in the School of Engineering, the consortium is taking on “the grand challenge of electrochemical energy storage in a world dependent on intermittent solar and wind power. We need affordable, grid-scale energy storage that will work dependably for a long time.”

That emphasis runs through multiple Stanford projects. “By emphasizing affordability, you naturally start to address other issues,” Rajagopal said. “Energy affordability allows more data centers that rely on more solar and wind, because renewables are now cheaper in many cases to install than non-renewables.”