When a historic cold snap froze Texas in February 2021, the lights went out for millions as power plants failed and electricity prices surged. Hundreds died. Billions of dollars in estimated damages piled up as the state’s power system revealed critical vulnerabilities.
The disaster wasn’t just about cold weather. It exposed what can happen when market incentives, regulations and infrastructure fall out of sync. Across the U.S., power systems are facing similar strains, from climate extremes and aging infrastructure to rising demand from artificial intelligence and electrification.
“Historically, it’s always been a challenge to get the economics to sufficiently match the physics of how the grid works, but I certainly think it’s the case that those challenges are larger now,” said Jacob Mays, an electricity markets experts and assistant professor in the School of Civil and Environmental Engineering at Cornell. “It’s a challenge for the operators. It’s a challenge for the regulators. It’s a challenge to get new rules in place fast enough to adjust to the changing reality.”
Mays, who is also an assistant professor of systems engineering and a faculty fellow at the Cornell Atkinson Center for Sustainability, focuses his research on how to design electricity markets and fund infrastructure upgrades in ways that make the grid more reliable, affordable and sustainable.
“All three of those priorities are making headlines right now,” Mays said. “The grid is becoming increasingly central to more parts of our economy, more parts of our lives, and that’s expected to continue as we progress with climate action and new technologies.”
Who foots the bill?
Behind every ambitious plan to upgrade transmission lines or integrate new energy sources into the grid lies a deceptively simple question: Who should pay?
Mays’s research centers on a principle known as “beneficiary pays,” the idea that those who stand to gain from a grid upgrade should bear the costs. The concept is simple. The politics are anything but.
In a study published in the Energy Law Journal, Mays and his co-author show how a carefully structured beneficiary-pays model helps regulators fairly allocate costs among the diverse, and often competing, parties involved in large-scale grid projects: private developers, regional utilities and state governments, among others. The approach offers a way out of the political deadlock that can styme investment in shared infrastructure.
An illustrative example of this model highlighted by Mays comes from a proposed multi-state transmission line offering different benefits across state lines. If the line delivers economic gains to Ohio and emissions reductions in New Jersey, the costs are split accordingly – Ohio pays for its economic benefits, while New Jersey covers the share tied to environmental gains. The model ensures each state pays only for what it gets.
“It was slightly easier to coordinate across these entitles in the past,” Mays said, “because we mostly were dealing with a small number of really large energy plants that had to be added to the system. Now there are more diverse technologies – electric vehicles, smart thermostats, data centers – on the demand side, and we have lots of different solar and wind farms popping up all over the place.”
That technical complexity of adding renewable energy to the grid is now colliding with political friction. As states pursue divergent energy goals, it’s become harder to reach agreement on who should pay for what. Without an accepted cost-sharing framework, utilities are inclined to prioritize modest local projects instead of major regional transmission lines that would deliver broader economic and environmental benefits. The result is a fragmented grid that can’t scale to meet today’s challenges.
Those challenges came to a head in 2024, when the Federal Energy Regulatory Commission issued Order 1920, mandating new planning and cost-allocation rules for regional transmission projects. The order aims to jumpstart multi-state coordination and prevent gridlock, but has faced legal challenges from a coalition of states concerned about rising electricity costs.
“I don’t think the fights are going to stop anytime soon,” Mays said, “but beneficiary pays is likely the only approach to cost allocation that doesn’t result in some customers free-riding off their neighbor’s transmission investments.”
Planning the grid in a volatile world
Climate change is already upending assumptions about how, when and where power is needed. Heat waves, wildfires and winter storms are testing the limits of aging infrastructure, and those events are arriving with growing intensity and less predictability.
“We have big uncertainty in the forecasted amount of wind and solar and demand,” Mays said. “We have unexpected failures from transmission lines, unexpected failures from large plants. So we have to be smart about how we operate under uncertainty.”
Mays sees data-driven optimization playing a critical role – using algorithms and analytics to make smarter decisions in the face of volatile conditions – but even that has its limits.
“There are cases where no amount of optimization really helps,” he said. “That’s when things are breaking and it’s not about tweaking algorithms and trying to get better performance, but about hardening infrastructure so we can operate through the worst scenarios.”
In a recent study in the journal Energy Economics, Mays and his co-author developed a planning model that tackles this uncertainty. Their approach accounts not just for different possible futures – changes in demand, fuel prices, policies – but for ambiguity, when the odds of those futures can’t even be agreed upon. By modeling many possible futures, Mays shows that the differences in beneficiary results can be large, and he argues for flexible cost-sharing methods that adjust after the fact to better match who truly benefited.
In one case study he examined, a region was expected to benefit from a new transmission line based on lower electricity prices from increased imports. However, in certain future scenarios, the region actually experienced losses because the transmission line increased competition for local generators, reducing their revenues. This mismatch between expected and actual outcomes underscores the need for cost-sharing to be flexible enough to adjust after the fact.
Such modeling is especially important as the grid becomes more dynamic, with shifting generation patterns, distributed energy resources and load growth arriving unpredictably.
The data center dilemma: AI’s growing footprint
Although the Energy Economics study doesn’t model artificial intelligence or data centers directly, the ambiguity posed by their growth is the kind of challenge Mays’s research approach was designed to handle.
Tech companies like Google, Amazon and Microsoft are racing to build new data centers to support AI and cloud computing. These facilities can demand hundreds of megawatts of power, comparable to a small city, prompting utilities to invest heavily in new substations, transmission lines and other upgrades. In Virginia alone, proposed data centers have multiplied rapidly, with estimates that meeting their demand will require $40 billion in new grid infrastructure over the next five years, according to the state’s energy operator. The operator is now proposing to raise power bills by 15% over the next two years.
“AI has thrown a wrench into the grid conversation because of the concern that rate payers are going to be asked to foot the bill for transmission upgrades that are going to benefit data centers,” Mays said. “What are the rules for that, and how do we make sure that we’re coming up with something that’s fair?”
In some states, regulators are exploring policies requiring developers to cover more of the grid connection costs. Some data centers have responded by insisting they’re willing to pay their fair share, but not for upgrades that also benefit others.
“It’s a really challenging problem,” Mays said. “Sometimes an upgrade is clearly just to serve a single company, in which case it makes sense for them to cover most or all of the cost. But other times, that upgrade might also benefit the broader system or future customers, and then it wouldn’t be fair to ask one company to pay for all of it.” Mays hopes modeling such as his beneficiary-pays principles will help regulators settle on a framework for cost-sharing that’s both economically sound and mutually agreeable, especially for large-scale projects involving renewables and AI.
Market design for a resilient grid
The importance of market design might seem abstract, until it shows up in your mailbox.
Electricity markets are the financial and operational backbone of the power system. They determine how electricity is bought, sold and priced in real time, and they send long-term investment signals to developers building new grid infrastructure. But these markets are struggling to keep up with the pace and complexity of today’s grid transformation.
Mays is currently researching how energy storage facilities, for instance, provide critical flexibility and reliability benefits to the grid, but aren’t fairly compensated under current market rules. Large batteries can store electricity when supply is high and release it when demand spikes, helping balance the grid and smooth out fluctuations from renewable energy.
Mays said Texas, which operates an energy-only market, has seen more market-driven battery deployment because its pricing system allows batteries to earn revenue during periods of grid stress. In contrast, states like California have relied more on mandates, forcing utilities to contract with storage providers regardless of whether the market offers adequate incentives.
As more states look to scale up battery storage, Mays warns that traditional capacity markets – designed around slow-ramping, fuel-based plants – will need to evolve.
“It will require some adaptations in the way we design those market mechanisms and the way we allow participation in them,” said Mays, who proposes mechanisms in a forthcoming research paper. “We need to get the right level of regulation on the market to get all the benefits of competition and ingenuity, but also some guardrails to ensure the system actually performs reliably.”
As batteries, solar and AI-driven demand reshape the grid, Mays’s work is a reminder that resilience isn’t just about keeping the lights on, it’s about designing systems that are fair, flexible and built to last. The grid’s transformation is already underway. The question now is whether markets, rules and regulations can keep up, or whether they’ll need a hard reset of their own.