Planning, permits and interconnection carry $266 billion of Brattle's $614 billion DER estimate

US DER capacity could more than quadruple if key policy barriers to deployment are removed, according to a Brattle Group report cited by T&D World. That would add more than 400 GW by 2050, roughly a third of the country's current generation fleet, and the report estimates the shift could avoid $614 billion in US power system costs through 2050. The report was prepared for The Pew Charitable Trusts and is titled The National Case for a More Distributed Power Grid. It counts distributed solar, battery storage, managed EV charging and flexible heating and cooling across residential, commercial and industrial customers.

The big number gets the headlines. The smaller one is more useful. On April 28, Pew released a DER policy playbook, Distributed Energy Can Unleash the Resilient, Affordable Grid of the Future, with six recommendations. Brattle's analysis finds that addressing three of them could deliver $266 billion in avoided power system costs through 2050. The three are incorporating distributed resources into distribution system planning, reducing permitting costs and reforming interconnection processes. That suggests a large share of the savings could come from process changes rather than new technology.

By Brattle's own numbers, three process reforms carry $266 billion of the $614 billion, a bit more than two-fifths, so a large share of the value sits in utility planning rooms and local permit offices. For a distribution planner deciding whether a customer battery fleet belongs in the next load forecast, that matters more than any national headline figure. It matters too for the energy lead at a company with a big new load waiting in the interconnection queue.

What the model counts, and what it displaces

Brattle's report page says the study combines long-term power system modeling with an assessment of where local distribution upgrades could be deferred. It covers 25 US regions and looks at both the bulk power system and the distribution system. That second piece carries a lot of the case. Customer-sited storage and flexible load compete most directly with feeder and substation spending, so a model that only looked at the bulk system would likely understate their value.

Ryan Hledik, a Brattle principal and report coauthor, said many homes and businesses already own technologies able to help the grid handle rising demand, and that uptake of those technologies will keep growing. In his view, utilities should count them as resources in their planning instead of leaving what they can do untapped. He wrote the report with Kate Peters, Sophie Edelman, Prisha Tiwari, Carly McAdam and Akhilesh Ramakrishnan.

Maureen Quinlan, who leads Pew's work to decarbonize the grid, pointed to historic demand driven by data centers, new manufacturing and electrification. She said DERs could deliver a significant amount of new energy supply if the barriers to adopting them are removed, and called the report evidence that pragmatic policy changes enabling the technology can save billions and cut emissions.

A separate scenario tests a harder target. Pew's release says that if the generation fleet must be carbon-free by 2050, expanded DERs would offset about 175 GW of utility-scale capacity. In that case, annual generation costs in 2050 would come down by $45 billion. That equals nearly 20% of the added cost of moving the power supply to zero carbon. The repeated 175 is a coincidence worth flagging: one is gigawatts of capacity in the carbon-free case, the other is a count of gas plants in the status quo comparison.

The storage result needs a careful citation. Expanded DERs would move utility-scale storage capacity toward distributed storage and other flexibility close to customers, but Pew's release and the PR Newswire version put that shift at about 105 GW, while Brattle's own report page says about 116 GW. Anyone quoting the figure in a filing or a board deck would do well to take it from the full report rather than a press summary.

A cheaper bet on uncertain demand

For a utility weighing a wave of data center interconnection requests, the overbuild scenario may be the most practical finding in the report. Suppose anticipated load growth falls short. Relying more heavily on distributed resources would then trim what is spent on overbuilt energy infrastructure from 2030 through 2040 by $33 billion, which works out to a cut of roughly 40%. Brattle frames this as flexibility for utilities planning around uncertain demand from data centers, manufacturing and electrification.

The PJM numbers show how big that uncertainty is. Energy-Storage.news, in reporting by April Bonner, cited an October 2025 US Energy Storage Coalition study finding that PJM's summer peak would rise 16 GW by 2028 and a further 30 GW by 2032. That projection assumes all the demand in PJM's 2025 load forecast materializes. The "if" is the whole planning problem. A utility that builds conventional capacity for load that never shows up spreads the cost of the extra plant and wire across every ratepayer.

The pressure is already political. The outlet reported that the projected load increase and interconnection queue backlogs have drawn calls for reform from local leaders and the federal government, and that PJM is considering multiple reform paths.

Forecasting method is part of the problem. The Energy Systems Integration Group's Long-Term Load and DER Forecasting report, covered by T&D World in August 2025, found that forecasts have historically scaled annual energy use or peak demand, an approach that often misses localized adoption of new technologies. It recommends hourly, high-resolution forecasts and scenario-based methods that account for drivers such as policy changes and market signals.

Julieta Giraldez of Electric Power Engineers, who chaired the task force behind it, stressed aligning forecasts across entities to reconcile system-level data with local grid needs. That gap matters for the Brattle case. Battery and flexible-load value lands feeder by feeder, which suggests a planner relying on a system-wide peak forecast could miss much of the deferral value the report's distribution assessment is built to find.

PJM is already paying for distributed capacity

PJM is the country's largest Regional Transmission Organisation, covering parts of 13 US states and the District of Columbia, and is the location of Virginia's "Data Centre Alley," the world's biggest cluster of data centres, according to Energy-Storage.news. A deal there offers a small real-world counterpart to the national model. On June 2, Google and Voltus announced a three-year agreement under which Voltus will aggregate up to 100 MW of distributed resources each year from local businesses and homes into a Google-funded virtual power plant, the outlet reported.

Voltus will then pay customers who participate, "turning Google's capacity demand into real economic benefits for PJM customers," in the companies' words. They describe it as a bring-your-own-capacity deal, a structure the US government has encouraged through steps such as the White House's "Ratepayer Protection Pledge," which Energy-Storage.news called largely symbolic and which Google and other major tech companies signed.

That payment flow mirrors one of the report's mechanisms. Brattle and Pew say DERs let customers, rather than large-scale power developers, earn compensation for services they provide to the grid, with net savings large enough to benefit all customers through lower rates. In the Voltus structure, a data center's capacity need becomes a check to a local homeowner or business.

Set against 400 GW, 100 MW a year is a rounding error. What it shows is a large buyer paying for distributed capacity as part of how it gets power, which is the kind of transaction Brattle's numbers assume will become routine. For data center developers in congested regions, the Google-Voltus structure is a working precedent that procurement teams can point to.

It also sits beside much bigger, more conventional deals by the same buyer. In March, Google announced plans with DTE Energy for a Michigan data center including 1,600 MW of solar paired with 450 MW of storage, which Google will fund and DTE will operate over a 20-year contract. In February, Xcel Energy said it would install 30 GWh of Form Energy's iron-air batteries at a Google-owned data center in Pine Island, Minnesota. Distributed capacity, on this evidence, is one line in a large buyer's portfolio rather than a replacement for utility-scale supply.

What the reforms look like at project level

On April 16, Lightshift Energy, which both develops and operates projects, announced a portfolio of five battery storage projects in Virginia, all built at distribution scale. Transmission-scale projects of 200 MW to 500 MW require years of PJM interconnection studies and tens of millions of dollars in upgrade costs, the outlet reported. Lightshift is avoiding that route and installing smaller batteries, typically 20 MW or less, at distribution substations that already exist, aiming to reduce the impact on regular energy consumers.

Lightshift's design choice shows the interconnection problem from the developer's side: sizing projects to existing substations is a way to stay out of the slowest part of the queue. Reform of the kind Brattle priced would matter most for developers who cannot shrink their projects to fit.

Pat Wood III, co-chair at Pew Charitable Trusts, discussed the nonprofit's DER Policy Playbook with Energy-Storage.news Premium. The playbook outlined ways to incentivize DERs in the US. To make entry easier, Pew recommends automated permitting for residential installations and simpler interconnection processes for commercial providers. The goal of these administrative changes is to lower costs and shorten the delays that slow deployment. On implementation, the report emphasized the importance of aggregating DERs into virtual power plants.

Those two administrative fixes line up with two of the three reforms behind the $266 billion, and the aggregation point describes what Voltus is doing for Google. Read together, the PJM deals look like early, small-scale versions of the playbook's process changes.

The household number, read carefully

By 2050, Brattle estimates, the average household could pay $122 a year less for electricity in current dollars than it would under the status quo. That saving would cover about half of the projected bill increase above the inflation rate the study assumes. The number is real, but it takes a long time to arrive. It describes where things stand in 2050, so a regulator can't count on it in the next rate case.

The strongest objection to the headline figure is that it depends on policy. Most of the levers belong to state regulators and local permitting offices, and only the $266 billion is tied to named reforms. The rest of the headline figure could depend on other barriers coming down as well. For planners, the safest reading treats the $266 billion as the part with a clear to-do list attached.

The first real-world check

Brattle has published the full report and a technical appendix, so utility planning teams can test its assumptions against their own territories. The closer signal is in PJM. The Google-Voltus contract runs three years with a 100 MW annual target, and whether Voltus fills that each year will show more about how fast distributed capacity can be recruited than any 2050 projection can.