In the era of rising interest rates, policies that enhance power system flexibility can reduce energy costs
The U.S. power industry is facing cost challenges from the growth of data centers and industrial loads. In 2025, electricity generation is expected to grow by 2.3%, with residential electricity prices rising over 5% year-over-year. Experts believe that through demand-side flexibility policies, aggregation of distributed energy resources, and new tariff designs for large loads, excessive investment for rare peak loads can be avoided, thereby controlling electricity price increases.

Facing cost pressures from hyperscale data centers and other new large loads, power system stakeholders told Utility Dive that utilities can effectively limit the resulting increase in customer electricity bills if they make good use of emerging flexible resources to manage demand peaks.
U.S. Energy Information Administration (EIA) September data shows that driven by growing electricity demand from data centers and industrial customers, total U.S. power sector generation will increase by 2.3% in 2025. EIA's Data Browser also shows that from July 2024 to July 2025, average U.S. residential electricity prices rose by more than 5%.
Stakeholders noted that expanding the power system to meet new large load demand while also controlling electricity price impacts seems contradictory, but it is not impossible to achieve.
Severin Borenstein, a professor at the Haas School of Business, University of California, Berkeley, said that if large loads prompt utilities to make expensive grid infrastructure upgrades to meet higher demand peaks, this will push up electricity prices. But he added that if policies incentivize "shaving" peak demand, large loads can avoid the costs that drive up electricity prices.
Analysts believe state policymakers and stakeholders can design and implement these incentives and other policies to enhance the flexibility, affordability, and reliability of the power system.
"Rising electricity prices are drawing the attention of politicians because costs, especially energy costs, are becoming a focus for consumers," said Ted Thomas, former chairman of the Arkansas Public Service Commission and founder of Energize Strategies. "This makes resisting electricity price increases an important issue on which politicians are willing to spend political capital."
Although the scale of load growth remains uncertain, investments made to meet that growth already threaten affordability. However, data shows that appropriate regulation can limit its impact and potentially stabilize customer costs.
Electricity price increase trend
A September report from the Edison Electric Institute (EEI) shows that investor-owned utilities, which serve 57% of U.S. electricity demand, will invest approximately $1.1 trillion in infrastructure from 2025 to 2029, up from $765 billion in the previous five years. The report also found that 91 GW of new capacity is currently under construction, with another 488 GW in the planning or proposed stage.
A September policy brief from the Union of Concerned Scientists (UCS) notes that data center electricity demand has created "a new class of costs" for system expansion and modernization, threatening "the bills of all customers." The brief adds that state-level regulatory practices "allow this worst-case outcome" to occur.
But stakeholders say new policies that encourage large loads such as data centers to flexibly adjust their electricity use and avoid demand peaks could improve these outcomes.

Large load flexibility
Borenstein of the Haas Institute said large loads drive up electricity prices because system operators and utilities must make capital expenditures (CapEx) to meet demand peaks. These peaks occur no more than 50 hours per year, but the infrastructure costs invested to address peaks enter and push up customer electricity rates.
Borenstein believes the most important way to prevent electricity price increases is to enable large loads and system operators to use demand flexibility to manage these rare new peaks. He added that grid connections for data centers and new large loads "can be conditioned on flexibly adjusting electricity use during peak periods."
Borenstein said using flexibility to lower the demand peaks of new large loads can spread the increased electricity sales of utilities across periods when infrastructure and supply are ample. Even if some distribution system investment is needed, "electricity prices will not rise significantly."
A report from Duke University's Nicholas Institute points out that by adopting "flexible demand strategies," the U.S. power system can meet more than 98 GW of new demand from data centers and other large loads. The report adds that "only a 0.5% annual reduction" is needed during peak demand periods.
An August 31 analysis by the Electric Power Research Institute (EPRI) also argues that greater flexibility "can transform data centers from passive customers into grid assets," thereby improving reliability, reducing costs, and accelerating grid connections. A pilot led by software design company Emerald AI in Arizona in May, using dedicated NVIDIA chips, validated the potential of data center flexibility.
Mike Hogan, senior advisor at the Regulatory Assistance Project (RAP), said relying solely on generation to address rare peak demand surges could lead to investing in "rarely used capacity" to meet resource adequacy standards. He added that leveraging low-cost demand-side options "is more valuable than ever."
Former commissioner Thomas said regulators can also implement large load tariffs with contractual commitments, ensuring that new demand is paid for by the party causing it. Thomas said policymakers want both to support economic development and to protect consumers from electricity price increases. He added that some "will stake their next election on the decisions before them."

Flexibility of distributed generation
Analysts, utilities, and other stakeholders say the use of flexible distributed energy resources (DERs) aggregated through virtual power plants (VPPs) is growing significantly.
A 2024 study by Lawrence Berkeley National Laboratory found that since 44% of utility capital expenditures go to distribution systems, shifting distribution system peaks could have the greatest impact on customer costs. September research from Wood Mackenzie shows that aggregated DERs and VPPs capable of shifting distribution demand peaks are growing.
Wood Mackenzie reports: "The VPP market grew more than 33% from 2024 to 2025," with aggregators achieving "commercialization of 433 utility and project programs, up 35% from last year."
In Illinois, Commonwealth Edison (ComEd) has proposed a "bring-your-own-device load reduction program" to "leverage customers' storage resources," said Scott Vogt, vice president of strategy, energy policy, and revenue programs at the company. The program's capital expenditures would be "minimal" and would begin with a competitive solicitation from DER aggregators.
Vogt said aggregators would receive signals to reduce electricity use during peak periods, with the goal of enrolling at least 10 GW of the 47 GW of DERs in ComEd's territory. DER aggregators and owners would be compensated, but for the utility and its non-DER customers, "the value lies in lowering the demand curve."
Former commissioner Thomas emphasized that shifting a significant amount of load away from peak periods can simultaneously reduce costs for new distribution system infrastructure and system maintenance. Regulators across the country are increasingly being asked to recognize the value of demand flexibility at all levels of the system.

Regulatory measures to expand demand flexibility
Elizabeth Cook, vice president of the Association of Edison Illuminating Companies (AEIC), said current regulatory processes and requirements leave utilities "in a bind" and unable to adopt forward-looking "out-of-the-box thinking." She added that to address today's load growth, utilities need to enhance their analytical capabilities and "truly reimagine the way they view their systems."
Cook continued: "Hyperscale data centers also need to adjust their approach to utilities, even if they can acquire the utilities they work with."
According to the DELTa database of the North Carolina Clean Energy Technology Center and the Smart Electric Power Alliance (SEPA), over the past three years, 30 states have designed 49 newly approved or proposed large load tariffs for 44 investor-owned and public utilities and cooperatives.
Ann Collier, senior manager of emerging technologies at SEPA, said, "Balancing numerous goals such as energy affordability, economic development, reliability, and clean energy is a significant effort." Provisions in new tariffs may include curtailing load during extreme demand peaks, building on-site generation or storage, or making long-term contractual commitments to cover infrastructure costs.
Josh Keeling, chief commercial officer of UtilityAPI, said expanding demand flexibility can reduce costs and increase participation, but regulatory processes create too much friction for residential customers. Keeling said few utilities can conveniently provide the customer-level data needed to integrate VPPs. But he added that initiatives such as EPRI's Open Power AI are developing methods to share data while protecting data security.
Keeling said utilities such as Eversource, Pacific Gas & Electric (PG&E), and Con Edison, along with their regulators, have streamlined data access. Maryland, New Jersey, Illinois, and New Hampshire are following the successful data access experiences of New York, California, and Texas.
Scott Harden, senior vice president and chief technology officer at Schneider Electric, said that for demand flexibility to be most effective, program design must include meaningful compensation and be clearly understood by participating customers. He said program assets must be dispatchable.
Harden added that new load growth will give regulators reason to address utilities' traditional bias in technology spending on distribution system management software.
Brendan Reed, vice president at Sparkfund, said regulators can formalize the use of aggregated DERs through distributed capacity procurement (DCP) to reduce demand peaks and their associated costs. He added that DCP would be included as one of the peak capacity resources in the utility planning process for cost recovery.
Will Baker, director of market innovation at Renew Home, said another metric is data from randomized controlled trials. He added that randomized controlled trial data can approximate the value of avoided transmission and distribution costs, ancillary services, and energy.
Hannah Bascom, chief growth officer at Uplight, said state policymakers can mandate that utilities reduce a certain percentage of overall peak load through load flexibility programs. Bascom added that even without mandates, utilities facing load growth are beginning to view flexibility as the most prudent investment to protect affordability and reliability.
Flexibility as a principle of prudence
Multiple stakeholders said that failing to fully consider flexibility in utility planning and rate cases violates the fundamental principles of prudence and reasonableness in electricity regulation.
Mark LeBel, head of research and strategy at the Regulatory Assistance Project (RAP), said that as demand grows, standardized data publicly available in utility planning and rate cases becomes more important. He added that regulatory decisions approving utility rate case investment proposals without sufficient supporting information would be imprudent.
Abigail Anthony, commissioner on the Rhode Island Public Utilities Commission, said that for a proposed rate case investment to be reasonable and prudent, a utility must present a business case explaining its necessity, value, and responsibility to customers. Anthony said a utility might argue that regulatory requirements hinder critical modernization, but it must demonstrate that modernization investments provide net value to ratepayers.
Anthony and others said proving the value of new investments is particularly important when significant uncertainty remains about the scale of load growth.
But regardless of future load growth, "in this world of computing power, utilities need hyperscale data centers to create AI analytics, and hyperscale data centers need utilities to deliver electricity," said Cook of the Association of Edison Illuminating Companies. "Regulation that fosters collaboration could become a win-win, enhancing overall reliability and affordability."