John Nichols is a Fellow at the Center for Energy & Environmental Policy at the Caesar Rodney Institute.
In 2025, the Delaware legislature directed a cost-benefit study of a proposed statewide battery storage program, with the goal of improving grid reliability and lowering peak costs. Battery storage systems store electricity when it is plentiful and release it later, when demand is high. The study reported $1.53 billion in projected benefits. But it did not calculate what the program would actually cost to build and operate. This piece does, using the study’s own numbers and established federal methodology. Using those inputs, the total cost comes into focus at roughly $3.8 billion over 10 years. Every figure is independently verifiable.
The battery storage study examined here, conducted by Gabel Associates, was commissioned by Energize Delaware under Senate Joint Resolution 3, using funds derived from charges paid by Delaware ratepayers through their electric bills under the Regional Greenhouse Gas Initiative, the state’s carbon reduction program, which some lawmakers have sought to withdraw Delaware from. Under law, 65% of Delaware’s annual RGGI auction proceeds flow to Energize Delaware — $39 million in 2025 and $239.2 million since 2009.
The study Gabel Associates delivered in April 2026 reported $645 million in projected benefits for behind-the-meter batteries (installed at homes and businesses) and $888 million for utility-scale systems. Combined, that totals $1.53 billion in projected benefits. Commissioned by Energize Delaware — the operating name of the Delaware Sustainable Energy Utility — the study was presented to the Delaware Energy Stakeholders Group, the state working body overseeing the program, and is available from that body on request; Energize Delaware has not posted it publicly, despite it being commissioned with public funds.
The study, however, never answers the most important question: Do the projected benefits exceed the costs Delaware ratepayers would ultimately bear? In fact, the study’s disclaimer states plainly, on Page 2, that it does not calculate costs against benefits. Readers should also be aware that Gabel Associates has a documented institutional position favoring battery storage expansion in PJM markets, having prepared a report for the American Council on Renewable Energy arguing that PJM should loosen grid connection rules to expedite battery storage development through an accelerated interconnection pathway known as the “surplus interconnection service.”
The legislature asked for both sides of the ledger — costs, as well as benefits — from a firm with a documented prior position favoring battery storage expansion. What it received instead was a projection of benefits without a corresponding analysis of costs, how those costs would be paid or who ultimately would bear them.
Three bills ratepayers will pay
If Delaware households fund a battery storage program, they will not pay one bill; they will pay three.
SJR 3 originally contained a provision allowing utilities regulated by the Delaware Public Service Commission — including Delmarva Power — to recover all reasonable costs of the battery pilot projects from ratepayers through billing. That provision was deleted by amendment before passage, leaving the cost allocation question open.
In practice, that question rarely stays open. Every comparable state program has ultimately recovered these costs through charges on customers’ electric bills. California’s Self-Generation Incentive Program committed $450 million in ratepayer funds to behind-the-meter storage. Maryland’s storage program draws on its Strategic Energy Investment Fund. New Jersey’s Garden State Energy Storage Program targets 1,000 megawatts through competitive solicitations, with required ratepayer savings analyses. Virginia has set a 3,100-MW storage goal by 2035, backed by utility procurement mandates.
Every state that has set deployment targets at this scale has recovered the costs through ratepayer-backed mechanisms. Delaware’s own history — the Renewable Energy Portfolio Standards Act, the qualified fuel cell provider project and community solar consolidated billing — confirms that it follows the same model.
The cost allocation question has been deferred, not resolved. In practice, Delaware households would likely pay for battery deployment through three separate layers of cost.
First, the batteries themselves. The study’s own numbers project a battery build-out costing approximately $3.3 billion to $3.6 billion over 10 years. Its cost table prices utility-scale batteries at $1,902 per kilowatt and commercial systems at $1,668 per kilowatt. Its deployment schedule projects 1,000 MW of utility-scale storage and 450 MW of customer-sited storage by 2036. Applying those figures to that schedule using the study’s own 2.3% annual inflation assumption produces that $3.3 billion to $3.6 billion estimate. These are the study’s own numbers applied to its own plan.
That cost projection also rests on a federal subsidy structure that no longer exists. The study was delivered nine months after President Donald Trump signed the “Big Beautiful Bill” on July 4, 2025, eliminating the 30% federal residential battery tax credit, effective Dec. 31, 2025 (Section 70506 of Public Law 119-21; see also IRS Fact Sheet 2025-05). The same legislation also imposed new domestic content requirements and foreign-entity-of-concern restrictions on larger battery projects seeking federal tax credits for construction beginning after 2025 — requirements that the Gabel study, delivered in April 2026 despite modeling deployment through 2036, does not address. Its 450-MW behind-the-meter deployment model is benchmarked against the three fastest-adopting states among 11 benchmarks — Massachusetts, Nevada and Rhode Island — states that built customer-sited storage when the federal credit was available.
The study projects 450 MW of behind-the-meter storage statewide without specifying what a typical residential installation would actually look like for a homeowner. The same industry source Gabel itself relies upon — the National Renewable Energy Laboratory’s annual Technology Baseline — defines a representative residential battery as a 5-KW, 12.5-kilowatt-hour system. At the study’s residential cost estimate of $3,086 per KW, that system would cost more than $15,000 with no remaining federal tax credit offset.
Delaware cannot replicate that benchmark trajectory without replacing the eliminated federal subsidy through state funding. That substitution cost does not appear in the study’s analysis.
Cost is not the only problem a Delaware homeowner faces. Reliability is another. A typical 5-KW home battery provides only a few hours of backup power under full household demand. That same limitation leads PJM to project that the reliable capacity value of batteries will fall to about 24% of their stated capacity by the mid-2030s, as more batteries compete during the same peak demand window.
In Delaware, that limitation compounds in the months when backup power matters most. Dover averages 3.6 peak-sun hours per day in December and 3.8 in January, compared to nearly six in July, according to NREL’s National Solar Radiation Database. A battery drained overnight on a consecutive string of cloudy winter days cannot reliably recharge.
After dark, in every season, the home is back on the grid.
Second, the electricity that charges them. A battery does not generate electricity; it stores it. That electricity is not free. Even when prices are low, the cost has already been paid by ratepayers through other charges on their bills, including renewable energy program charges and capacity payments. Charging the batteries adds another layer of cost, not a new source of savings. The low wholesale price at which batteries propose to charge is not cheap electricity; it is electricity whose true cost has already been collected from ratepayers twice: first through the renewable compliance charge that helped fund the resources suppressing the wholesale prices and again through rising capacity charges, as reliable power plants became less economical and left the market.
Lastly, the gas turbines and backup power plants that must stay. PJM’s own capacity accreditation data, cited in the study, shows that a four-hour battery currently provides a reliable capacity equal to about 55% of its rated size but projects that the figure will fall to just 24% by the mid-2030s, as more batteries compete for the same peak window. At that rate, 1,450 MW of proposed batteries would provide reliability equal to about 348 MW of reliable on-demand generation.
The gas turbines and other reliable backup power sources required to cover the remaining capacity obligation must stay under contract and available at ratepayer expense. Keeping existing dispatchable generation available currently costs Delaware ratepayers $122-$156 per KW per year in PJM capacity market charges, according to auction clearing prices published by Monitoring Analytics, the Federal Energy Regulatory Commission-designated independent market monitor for PJM, in its “2025 State of the Market Report for PJM.” Should the reliability gap grow large enough to require new plants — a risk the declining effective load-carrying capability trajectory makes increasingly plausible — that cost rises sharply. The Brattle Group, the independent consulting firm retained by PJM itself to set the cost benchmarks underlying the regional capacity market, places the all-in annual cost of a new natural gas combustion turbine at approximately $304 per KW per year for the 2028-29 delivery year — 43%-46% higher than its estimate just two-and-a-half years earlier and before the major tariff increases announced in April 2025 are accounted for.
What the study counts as benefits
Of the $1.53 billion in combined benefits the study reports, more than half — $769 million — consists of jobs, wages and tax revenue generated by spending approximately $3.3 billion on construction and maintenance programs, estimated through an economic modeling tool called “IMPLAN,” which measures how spending circulates through a local economy.
The U.S. Office of Management and Budget’s Circular A-4 — the federal guidance the study itself cites to establish its discount rate — states that economic activity generated by spending cannot be counted as a net program benefit because the same dollars spent on any alternative would generate similar jobs, income and tax revenue. IMPLAN is a legitimate tool for measuring how dollars circulate through a local economy, but it is the wrong tool for a ratepayer cost-benefit analysis.
The study uses federal rules to set its discount rate, but those same rules say it cannot count spending as a benefit.
Spend $3.3 billion on gas turbines, transmission upgrades or road repairs, and IMPLAN would return a similar jobs figure.
The study’s own methodology narrows this claim further. It excludes roughly 84% of construction costs from the IMPLAN calculation — battery modules, inverters and transformers are manufactured outside Delaware and do not circulate through the local economy. When the General Assembly is next presented a jobs figure from this study, that figure rests on just 16 cents of every construction dollar circulating inside Delaware’s economy. The other 84 cents leave the state.
After removing the IMPLAN jobs figures, the study’s remaining claimed grid benefits total approximately $764 million. But, even within that figure, roughly $300 million consists of modeled values assigned to lower carbon emissions and hypothetical avoided outages, rather than direct savings to ratepayers.
Each of those remaining benefits also shrinks under examination.
Any savings from buying electricity during low-price periods and discharging during high-price periods primarily benefit the battery owner, not automatically the broader body of ratepayers funding the system. The peak demand reduction is temporary because most batteries can discharge for only about four hours before underlying demand is fully exposed again — which helps explain why PJM projects that the reliability value of batteries declines as more of them are built. That distinction matters when evaluating whether large-scale ratepayer subsidization of battery ownership produces benefits proportional to its costs.
Those reliability limitations also affect several of the study’s projected systemwide benefits. The study’s claimed $165 million in transmission deferral savings does not account for the grid interconnection infrastructure the batteries themselves require. Its emissions benefit assumptions are also less certain than presented. The analysis assumes that batteries charge using cleaner electricity and discharge when dirtier generation would otherwise be used. But the study’s own Section 6.C acknowledges that some modeled years produced net emissions increases instead.
What the numbers actually show
Projected program costs of at least $3.8 billion exceed the study’s remaining claimed grid benefits of roughly $764 million by approximately 5-to-1 before backup capacity costs are even added. Based on data for Delaware from the U.S. Energy Information Administration’s Form EIA-861, the average Delmarva Power residential customer faces an estimated addition of approximately $89 per month. The average commercial account faces an estimated increase of approximately $269 per month. The average industrial account faces approximately $1,870 per month. The 38 industrial accounts in Delmarva Power’s Delaware-bundled retail base captured in Form 861 are predominantly smaller industrial customers; larger Delaware industrial customers typically procure supply competitively, and their per-account impact under a delivery service cost allocation would differ.
These figures are approximations calculated by annualizing the $3.8 billion 10-year program cost and allocating it across Delmarva Power’s customer base, weighted by each class’s share of total 2024 annual revenue using Form 861 data for Delaware. The Gabel study modeled costs statewide across all three Delaware electric providers — Delmarva Power, the Delaware Electric Cooperative and the Delaware Municipal Electric Corp. — in proportion to each utility’s share of total statewide electricity sales. This analysis instead allocates the full projected statewide program cost to Delmarva Power customers alone, reflecting that Delmarva Power is Delaware’s only investor-owned utility subject to Public Service Commission regulation and, therefore, the most likely vehicle for ratepayer cost recovery. Had costs been distributed proportionally across all Delaware utilities, Delmarva Power customers would bear approximately 65% of the statewide total, producing lower per-customer estimates than those shown here. The actual per-customer impact will depend on the cost recovery mechanism the General Assembly authorizes and the Delaware PSC ultimately implements, neither of which has been determined.
The study acknowledged that ratepayer-funded programs require a cost analysis it declined to provide, calling such analysis “useful.” It is more than useful. It is the only analysis that answers the question ratepayers need answered before Delaware commits to the program: What will this cost?
The study outlines what Delaware might gain from battery storage. It does not show what it will cost. Using the study’s own assumptions, that cost approaches $3.8 billion over 10 years, against roughly $764 million in measurable grid benefits.
That gap is not explained in the analysis provided to lawmakers.
Instead, it has grown since the study was produced. The federal residential battery tax credit no longer exists. The deployment model the study relies on was benchmarked against states that built storage when it did. Replacing that eliminated federal subsidy with state funding is the predictable next step — and an additional ratepayer cost the study declined to calculate.
The General Assembly has already authorized a study. It will next be asked to authorize the program itself — before the cost of building and operating that program has been calculated and before the mechanism for recovering those costs from ratepayers has been determined. Whether those costs will ultimately appear through household electric bills, special charges or some other mechanism was deliberately left open. Neither the General Assembly nor the Delaware PSC has determined how Delaware households and businesses would ultimately pay for it.
Promises first, costs later
Delaware has seen this pattern before. In July 2011, the General Assembly amended the Renewable Energy Portfolio Standards Act to create the regulatory framework for a 30-MW Bloom Energy fuel cell project — structured as a jobs bill built on promises of 375 construction jobs, up to 900 manufacturing positions and up to 600 supplier jobs. The Delaware Economic Development Office and the Delaware Department of Natural Resources and Environmental Control designated Bloom Energy an “economic development opportunity” before any cost analysis was performed.
Independent consultants retained by the Delaware Public Service Commission calculated the project’s above-market costs at approximately $113 million in net present value — roughly $1.34 per month for the average residential customer. The commission approved the tariff anyway.
Under the authorizing legislation, once approved, the ratepayer charge became irrevocable and could not be reversed by future state utility regulators. The economic development case was built on IMPLAN modeling. Cumulative ratepayer charges since 2012 now exceed $400 million.
The Gabel study leads with 776 jobs per year, derived from IMPLAN modeling. When the General Assembly is next asked to authorize the battery storage program, the Delaware PSC will ultimately be left determining how to recover costs from ratepayers after the broader commitment has already been made. In other words, the cost recovery question will be answered after ratepayers are already committed.
The Renewable Energy Portfolio Standards Act established the template. The qualified fuel cell provider project repeated it. Benefits and jobs are claimed. Costs are ignored. Ratepayers are committed.
The bill is $3.8 billion.
Reader reactions, pro or con, are welcomed at civiltalk@iniusa.org.
