The same 500-megawatt data center or other large load can raise electricity rates for existing customers in one region and have far less effect in another, according to a new report from the Energy Systems Integration Group’s Large Loads Rate Impacts Task Force.
The report finds that the availability of system headroom, market structure, whether the load materializes as forecast, and how tariffs and cost allocation are designed all shape the outcome.
Prepared by The Brattle Group and ESIG, Impacts of Large Loads on Electricity Rates: A Quantitative Assessment builds on the framework established in the task force’s May 2026 primer and applies it through quantitative modeling.
The task force included state utility commissions, utilities, a grid operator, consumer advocates, state and federal agencies, and researchers.
The analysis models a 500 MW large load with a 15-year contract with a prototypical utility in four jurisdictional contexts: a vertically integrated utility in the Southeast, a vertically integrated utility in MISO, a utility in PJM exposed to wholesale capacity markets, and a transmission and distribution utility with competitive retail energy supply in ERCOT.
Each host utility has a 3 GW peak, making the new load equal to 16.7% of existing peak. In each context, the model estimates the rate impact for existing customers by comparing incremental revenue from the new large load customer with incremental costs incurred by the utility to serve them.
“The rate impact of a large load isn’t predetermined,” said Long Lam, Managing Energy Associate at The Brattle Group and lead author of the report. “In our modeling, the same 500-megawatt customer lowered rates for existing customers when the system had headroom and raised rates when it didn’t. Forecast error and tariff design also determined how large those impacts became.”
In the modeled Southeast utility, if there is available generation and transmission headroom to serve the full load, rates for existing customers are estimated to reduce by 0.88 ¢/kWh, or about 6% on average over the 15-year contract term. Without headroom, the utility has to build new infrastructure at a higher cost than system’s average embedded cost, raising existing customers’ rates by 0.44 ¢/kWh.
Forecast error increases the rate impact more than faster-than-expected growth reduces it. If the load arrives four years late, the average impact rises to 0.73 ¢/kWh; if it reaches only half its contracted size, 0.93 ¢/kWh; and if it departs five years early, 0.87 ¢/kWh.
With all three at once, in an unlikely worst-case scenario, the impact reaches 1.32 ¢/kWh -- three times the base case. By comparison, a load that reaches full size in half the expected time lowers the impact by only 0.07 ¢/kWh.
Tariff protections can be put in place to mitigate risks associated with differences between actual and forecasted loads, a news release related to the report said.
Under the combined worst case, a modeled termination payment reduces the rate impact by 27%, while an 80% minimum billing demand reduces it by 52%.
The report also examines approaches that address cost shifts the tariff mechanisms do not fully cover.
In the modeled worst case, assigning generation and transmission costs directly to the new load eliminates the rate increase for existing customers. “And” pricing, under which the new load also pays a share of existing system costs, produces the largest rate reduction in the model.
The report notes that both approaches would require careful implementation, and that “and” pricing would be a significant departure from traditional cost allocation practices.
“The results show that there are regulatory mechanisms that can measurably protect existing customers,” said Trieu Mai, Senior Director of Programs and Research at ESIG and a co-author of the report. “This analysis gives commission staff and utility planners a way to identify which tariff elements would be best suited to address the highest priority risks in their own jurisdictions.”
The same large load produced markedly different outcomes across jurisdictions. Before tariff interventions, the average impact of the 500 MW addition is 0.44 ¢/kWh in the Southeast utility, 0.22 ¢/kWh in the MISO utility, 0.07 ¢/kWh in the PJM utility, and 0.02 ¢/kWh in the ERCOT utility.
The rate impacts from a 500 MW load are lower in MISO, PJM, and ERCOT because some costs are shared across the region. But customers in regional markets may also be exposed to costs from large loads added in other utility territories, the news release noted.
“Regional markets cut both ways,” said Natalie Northrup, Energy Research Associate at The Brattle Group and a co-author of the report. “A single large load has a smaller effect on rates when its costs are spread across a broader regional market, but customers in those markets are also exposed to growth happening elsewhere in the region. When we scaled load growth to the same 16.7% increase in peak demand across each market, the modeled impact rose to 0.29 cents per kilowatt-hour for the MISO utility, 0.65 cents for the ERCOT utility, and between 1.35 and 2.53 cents for the PJM utility.”
The variation across modeled utilities shows that rate impacts depend heavily on local system conditions and market structure, rather than on the presence of a large load alone.
The results are illustrative rather than utility-specific, but they consistently show that system headroom, forecast accuracy, and tariff design can determine whether the same large load lowers or raises rates for existing customers, ESIG said.
The analysis does not estimate broader economic-development benefits or costs, distribution system costs, or, outside the regional scenarios, the cumulative effect of multiple large loads.
The report arrives as ESIG launches its Large Loads Working Group, which holds its first meeting this week at ESIG’s 2026 Fall Technical Workshop in Reston, Virginia.
The working group builds on ESIG’s Large Loads Task Force, which brought together nearly 500 participants across seven project teams and published 11 reports on large load forecasting, interconnection, modeling, reliability, resource adequacy, transmission planning, and markets.
The Large Loads Working Group will provide an ongoing forum for collaboration, knowledge sharing, and the identification of future task forces and other initiatives that help the industry address the challenges and opportunities of integrating large loads.
It sponsors targeted task forces, including the Large Loads Rate Impacts Task Force, and will work closely with ESIG’s other working groups.
