08/27/2026
What does that energy-cost chart really tell us — and what is driving your electric bill?
Good Morning, Neighbors.
First, let's give you some context. The Delaware Project sees power costs discussed hotly, and often inaccurately, among what we'll call the "politically interested" throughout Delaware on a daily basis.
We first saw the chart in question on Sen. Stephanie Hansen's page, took a few minutes to look it up and read the source, and by the time we got back, the comments were already wild—and, quite often, wrong. But to make sure you get another party's take on energy-related issues, take a look at her page and the Delaware State Senate Republican Caucus when you get a chance.
So, we're going to try to keep all sides, including our own, as honest as possible. The chart below, the one we saw on Sen. Hansen's page, is Lazard's annual comparison of what new power plants cost. Lazard is a prestigious global financial firm that specializes in high-level business advice and investing capital, and its chart tells the truth, but only when read closely.
And that's what we did; we actually read it before saying a word about it. Doesn't that sound crazy?
Lazard's "levelized cost of energy" is the lifetime cost of building, financing, fueling, and running a new plant, spread across the electricity it produces. The July 2026 edition, unsubsidized, per megawatt-hour: $40–$98 for utility-scale solar, $37–$99 for onshore wind, $61–$156 for solar with batteries, $51–$129 for combined-cycle gas, $144–$276 for gas peakers. New nuclear, at $175–$255, reflects Georgia's Vogtle reactors, the only recent U.S. build. New coal? With none being built, Lazard shows only an inflation-adjusted older estimate.
Read the fine print. The ranges overlap — a favorable gas project can beat a difficult solar one. The low solar numbers are utility projects, not rooftops. "Unsubsidized" excludes tax credits, and pollution costs, too. And every technology got pricier this year — interest rates, tariffs, supply chains.
Why are wind and solar so often cheapest? No fuel bill, ever. Their cost is mostly up-front, factory-built and quick to install; manufacturing scale cut utility-scale solar costs 82% from 2010 to 2020, per the federal renewable-energy lab. A gas plant buys fuel every hour it runs, at prices that swing; new nuclear carries heavy construction and financing risk.
The chart cannot tell you this: electricity is not one product. A megawatt-hour in a July heat wave, close to home, is worth more than one far away in a mild midday surplus. The U.S. Energy Information Administration warns such direct comparisons "are misleading" on their own. Its averages for plants entering service in 2031 cluster closer — about $57 wind, $58 solar, $77 combined-cycle gas, $94 solar-plus-battery, $88 advanced nuclear — its wind and solar figures now credit-free, since the 2025 tax law ends those credits for projects finished after 2027. Economists have long made this point; careful reviews find integration costs small at low renewable shares, rising unevenly at high ones. Lazard agrees: its own report adds regional "firming" costs — and even firmed wind (about $98) and solar (about $123) on our grid, PJM, still land inside the range of new combined-cycle gas.
And the chart prices new plants; most of our power comes from old ones. Lazard pegs the marginal cost of simply running existing, paid-off plants at roughly $26–$36 per MWh for nuclear, $32–$51 for combined-cycle gas, $34–$69 for coal — often cheaper than building anything. Whether one stays open is decided plant by plant — which brings us to Millsboro.
NRG bought Indian River from Delmarva Power in 2001. Three of its four coal units closed from 2010 to 2013 under state pollution agreements. In June 2021, NRG told PJM it would retire the last 411-MW unit after two straight years of losses as capacity-market revenue collapsed — Lazard's math in an owner's books. But it could not simply vanish: PJM found retirement would violate transmission-reliability standards on the peninsula, so customers paid roughly $50 million a year under a federal "reliability-must-run" agreement while Delmarva built the fix. The final upgrade finished early; the unit closed in February 2025, 22 months ahead of schedule, saving customers an estimated $93 million. This spring, with data-center demand surging, Sussex legislators floated restarting the site, likely on gas. "Replace 411 megawatts of coal with 411 of solar" was never honest accounting; replace the energy, the capacity and the local grid services, and the math works.
Now the costs kept off every bill. A congressionally requested National Academies study put coal power's non-climate damages — deaths, illness, harmed crops and buildings — near $32 per MWh in 2005, versus $1.60 for gas. Old numbers — controls and closures have cut the harm steeply. A 2023 study in Science tied about 460,000 deaths of older Americans to coal-plant particulates from 1999 to 2020, the yearly toll falling from over 43,000 to about 1,600 by 2020. Berkeley Lab estimates 2022 wind and solar delivered climate and health benefits near $143 and $100 per MWh ($36 and $17 for health alone) — society's savings, not a bill credit, and assumption-sensitive. Wind, solar and nuclear have real impacts too: mining, land, wildlife, waste. The durable conclusion is an ordering, not one true price: coal's unpriced burden largest, gas smaller but real, non-combustion far gentler on the people downwind.
Meanwhile, our bills. Residential electricity averaged 13.15 cents per kWh nationally in 2020 and 17.30 cents in 2025 — up about 32% — and this June ran 5% above last. Delaware reached about 19.4 cents this spring, up roughly 7% in a year. The state Public Advocate finds Delmarva distribution rates doubled since 2006, transmission charges up 93.5% since 2020, summer supply rates up 114% since 2021. The renewable-compliance charge? Down 11% this June. Delmarva earns no profit on the power itself — supply costs pass through — while its parent, Exelon, earned $2.77 billion in 2025, up 12.5%, crediting higher distribution and transmission rates among the drivers. That is not proof of gouging — regulated wires earn an authorized return — but it is exactly why Delaware's new prudence law, in force since January, must be used.
What would serving the public look like?
• Plan on whole-system value — energy, capacity, firming, wires, fuel and construction risk, health — never on one number.
• Buy wind, solar and storage where they win competitively; keep affordable existing nuclear and other firm plants; add gas only for demonstrated need, not speculation customers will pay to abandon.
• Build wires, efficiency and demand response faster; make the largest new loads, data centers included, pay the costs they cause.
• Audit utility spending against projections; shield the households least able to absorb one more increase.
Neither slogan is true — not "renewables always win," not "keep every old plant." The public interest is the lowest-risk mix of energy and reliability, with the off-bill costs counted and a plain answer to who pays and who profits. On that fuller accounting, clean energy usually earns its place. Whether its benefits reach your bill depends on planning, oversight — and neighbors paying attention.
IF you'd like to read the Lazard report yourself, we'll put the link in the comments, and supply any of the other sources we relied on upon request.
See Clearly. Do the Work. Build Together.