Energy / Electric utilities · Deep dive

Dominion Energy, Inc.

A Richmond-based investor-owned utility ($21.8B FY2025 revenue) whose Virginia service territory now sits on top of the largest data-center concentration on earth — an unearned tailwind the Virginia SCC keeps trying to price back to hyperscalers instead of shareholders, at the same time a $11.65B, six-months-late offshore wind build absorbs another tariff-driven overrun and a 2020 pure-play restructuring is still delivering a smaller company than the one investors bought.

at risk

Dominion sits on the single best load-growth tailwind in US utilities — Virginia data-center demand — but the SCC has now spent two rate cases explicitly designing tariffs and deposits to shift that upside to hyperscalers rather than shareholders, the CVOW build has slipped six months and blown through $11.65B against a $7.8B original budget, and the same hyperscalers whose load underwrites the pitch are increasingly signing directly with Talen/Constellation/Vistra nuclear or Bloom fuel cells that never touch the utility meter.

My take ★★ Very interesting ★ Interesting − Maybe ✗ Not interesting

Snapshot

Dominion Energy is a Richmond-based investor-owned electric and gas utility with FY2025 revenue of $21.8B, ~$58B market cap (August 14, 2026, at $68.76/share) and roughly 17,500 employees. Its regulated electric business — Dominion Energy Virginia, which serves ~2.8M customers across the Commonwealth — sits atop the densest data-center cluster on earth: Loudoun County alone accounts for the majority of Virginia’s data-center footprint, and data-center load reached 28% of Virginia Power’s electric sales for the year ended December 31, 2025. On paper, that is the best load-growth tailwind in US utilities. In practice, the Virginia SCC’s November 25, 2025 biennial-review order and its June 2026 large-load orders have systematically re-priced that upside away from Dominion shareholders and toward the hyperscalers actually driving the load, while a $11.65B, now-2027 Coastal Virginia Offshore Wind build absorbs its second tariff-driven overrun and hybrid debt piles up to fund the $65B 2026-2030 capital plan.

Founding story

Dominion is not a founder story; it is a Richmond franchise story. On June 29, 1909 the Virginia Railway & Power Company was organized to run Richmond’s electric streetcars and streetlights. In 1925 a Stone & Webster-led syndicate rebranded it as the Virginia Electric and Power Company — VEPCO — and by mid-century it was the Commonwealth’s regulated electric monopoly. Dominion Resources was formed in 1983 as a holding company above VEPCO to enable non-utility diversification; that experiment ran hardest through the 1990s and 2000s (Consolidated Natural Gas in 2000 for ~$6.4B; Millstone nuclear; merchant generation across the Northeast; interstate gas pipelines) and produced, by 2017, a sprawling energy conglomerate rebranded as Dominion Energy.

The modern story is the retrenchment. In July 2020 Dominion agreed to sell 100% of its interstate gas transmission and storage — 7,700+ miles of pipe, 900 Bcf of storage, plus 50% of Iroquois and 25% of Cove Point LNG — to Berkshire Hathaway Energy for $9.7B ($4B cash and ~$5.7B assumed debt). That transaction jumped 18% of US interstate gas transmission into Berkshire’s hands and left Dominion, in management’s language, a “pure-play state-regulated” utility. The same week, Dominion and Duke cancelled the Atlantic Coast Pipeline after regulatory delays. The consequence: a 33% cut to the quarterly dividend, announced November 4, 2020 ($0.94 → $0.63), deeper than the ~28% signaled in July. On October 1, 2020 the board elevated Bob Blue — a Yale-trained lawyer who had come to Dominion in 2005 from Governor Mark Warner’s policy office and worked his way through state and federal affairs, regulation and law, and Power Delivery — to CEO, with a mandate to finish the pivot, deliver CVOW, and convert Virginia’s data-center boom into rate base.

How it works

Dominion is a rate-regulated utility holding company. State commissions — primarily the Virginia SCC and the South Carolina PSC — grant each operating utility the exclusive right to distribute (and, in Virginia, generate and transmit) electricity in a defined franchise, in exchange for tariffed rates set on rate base times an allowed ROE, plus recovery of debt costs, opex, depreciation and (in Virginia) a menagerie of rate adjustment clauses that pass through fuel, environmental compliance, and specific capital projects. Every dollar of qualifying poles, wires, transformers, substations, generation, tree-trimming, undergrounding, grid-mod IT and offshore-wind capex becomes rate base and earns the allowed ROE thereafter.

Unlike restructured Ohio or Pennsylvania, Virginia is a vertically integrated jurisdiction: Dominion Energy Virginia both owns generation and delivers it. That is why offshore wind is on Dominion’s balance sheet at all — CVOW is a rate-base build, not a merchant IPP. The physical mechanics of the growth story are Loudoun County: hyperscalers request a service address, Dominion designs an interconnection (typically at 230 kV or 500 kV via a new substation), and PJM’s transmission planning slots the upgrades into its Regional Transmission Expansion Plan. As of 2025-2026, Dominion has told regulators it is planning for tens of gigawatts of contracted large-load demand — reportedly ~70 GW in various filings — a step-change against a 2025 DOM Zone summer peak of ~24 GW.

Product and business overview

Four reporting segments: Dominion Energy Virginia (regulated electric — ~2.8M Virginia and northeastern North Carolina customers; the earnings core); Dominion Energy South Carolina (regulated electric and gas — SCE&G, acquired via SCANA in January 2019); Contracted Energy (long-term-contracted renewables, including the Stonepeak-partnered CVOW project); and Corporate and Other (parent-level financing, Charybdis vessel operations, non-utility). The signature build is CVOW — 2.6 GW, 176 turbines and three offshore substations in a ~113,000-acre lease area ~27 miles off Virginia Beach — supported by the first US-built Jones Act wind turbine installation vessel, Charybdis (delivered from Seatrium AmFELS in Brownsville, TX in 2024 after multi-year delays; chartered to Ørsted and Eversource for Revolution Wind and Sunrise Wind before returning to CVOW). Signature gas build: the 944 MW Chesterfield Energy Reliability Center peaker, approved by the SCC on November 25, 2025 at an anticipated cost of ~$1.47B — the first new gas plant approved in Virginia since the 2020 Virginia Clean Economy Act.

Business model and pricing

Base rates are tariffed; there is no price list. What matters is the biennial review — Virginia’s periodic true-up between allowed and earned returns — plus the individual RAC filings for each capital project, and, now, the GS-5 large-load rate class. In the November 25, 2025 biennial-review order the SCC granted Dominion a $565.7M annual base-rate increase for 2026 and a further $209.9M for 2027, against Dominion’s asks of $822M and $345M respectively — a 31% and 39% haircut, per SCC filings. Authorized ROE landed at 9.8% against a 10.4% request, per Virginia Mercury (November 25, 2025). Every 100 bps of ROE is roughly $11M of base and $13M of RAC-adjusted revenue at Dominion Virginia, per the AG’s Office of Consumer Counsel testimony.

The GS-5 class, also approved November 25, 2025 (effective January 1, 2027), captures customers with demand of 25 MW or more — essentially all data centers — and requires them to pay for a minimum of 85% of contracted distribution and transmission demand and 60% of generation demand regardless of usage, per the SCC news release. In June 2026 the SCC further ordered that GS-5 customers post $1.5M/MW of collateral (Dominion’s original ask; the data-center industry had lobbied for $450K/MW), per Forbes (June 9, 2026). Both provisions are explicitly designed so that if a data-center project shrinks or cancels, the ratepayer body is not left funding a stranded substation. In FY2025 Dominion posted $21.8B of revenue, operating EPS of $3.42 and GAAP EPS of $3.45, per the FY2025 10-K. FY2026 operating EPS guidance is $3.45-$3.69 (midpoint $3.57); long-term operating-EPS growth target is 5-7% CAGR through 2030. Annual dividend has been held flat at $2.67/share since the 2021 reset.

Traction over time

Market analysis

US electricity demand grew ~0% for two decades. AI ended that. PJM’s January 2026 load forecast puts summer peak demand in the DOM (Dominion) Zone growing at 5.4% CAGR over the next decade — the largest absolute increase of any PJM zone — trimmed from 6.3% in the 2025 outlook, per Utility Dive. The DOM Zone summer peak was 23,905 MW in 2025, already 23% above 2019, per PJM. Every US utility would trade for that curve. Dominion is on top of it.

The problem is that “on top of it” is not the same as “captures it.” Two structural forces are re-routing the upside. First is the regulatory response: Virginia’s Attorney General’s Office of Consumer Counsel, Clean Virginia, the Southern Environmental Law Center, the Piedmont Environmental Council and the Virginia Poverty Law Center have collectively spent 2024-2026 pushing the SCC to force data centers to pay for the incremental generation, transmission and stranded-cost risk they create — culminating in the GS-5 class, the 85%/60% minimums, and the $1.5M/MW deposit. Second is bypass: hyperscalers demonstrably prefer to sign directly with Talen (AWS at Susquehanna, 1,920 MW to 2042), Constellation (Microsoft at Three Mile Island Unit 1, 835 MW for 20 years) or Vistra (Meta options January 2026), and to buy Bloom fuel cells (Brookfield’s $25B commitment on June 30, 2026 for on-site AI power) — all of which sit outside Dominion’s rate base entirely. And in April 2026, Virginia enacted a law requiring an independent review of Dominion’s load-forecasting methodology, driven by Google and other hyperscalers arguing Dominion double-counts contingent projects and thus inflates the transmission it wants ratepayers to fund.

Competitive intel

Regulated utilities do not compete for retail customers; they compete for capital and for regulatory permission. Dominion’s peers all beat it on one axis or the other.

NextEra is the industry’s growth benchmark: an ~$150B market cap built on execution in renewables — the specific muscle Dominion has failed to prove on CVOW. Every $500M tariff-driven CVOW overrun makes NEE’s un-fumbled solar and storage roll-outs look easier to underwrite. Duke is the direct Southeast investor comp at ~$95B market cap, ~8.6M customers, a $103B five-year capex plan and cleaner rate-case outcomes in NC and SC. Southern Company owns the only new US nuclear (Vogtle 3 & 4 at Georgia Power) — an actual scarce asset hyperscalers now pay for; Dominion’s four Virginia nuclear units (Surry, North Anna) are the same 1970s-vintage fleet everyone else has. AEP wrote the data-center tariff playbook Ohio uses; the GS-5 class is Virginia’s answer in the same shape. Where AEP proactively negotiated hyperscaler cost allocation, Dominion had it imposed on it.

The genuinely disruptive competitors are not utilities. Constellation, Talen and Vistra are converting existing merchant nuclear into hyperscaler direct supply — every 1 GW they lock up is 1 GW that never becomes Dominion rate base. Bloom Energy and behind-the-meter fuel cells, backed by Brookfield’s $25B, deploy in under two months versus Dominion’s multi-year interconnection queue — a 2030 forecast that one-third of hyperscalers plan on-site generation is a direct hit to the DOM Zone growth thesis. Base Power on the residential side and KKR / Brookfield on the wholesale side both raise capital cheaper than Dominion’s cost of equity, and can therefore price hyperscaler MWs off-balance-sheet.

History and evolution

• 1909-06-29 — Virginia Railway & Power organized in Richmond.
• 1925 / 1983 / 2017 — Rebranded VEPCO → Dominion Resources (holding company) → Dominion Energy.
• 2000-01 — Consolidated Natural Gas acquired (~$6.4B), pushing Dominion into interstate gas pipelines.
• 2018-07 / 2019-01-02 — SCANA/SCE&G acquisition announced ($15B) after the 2017 abandonment of V.C. Summer nuclear units 2 & 3; closed January 2019 with a $1.3B SC customer rate credit.
• 2020-07-05 — $9.7B gas transmission and storage sale to Berkshire Hathaway Energy announced; Atlantic Coast Pipeline cancelled with Duke.
• 2020-10-01 / 2020-11-04 — Bob Blue named CEO; dividend cut 33%.
• 2022-11 / 2024-01-01 — Steven Ridge named SVP & CFO; promoted to EVP & CFO.
• 2024-04 — Charybdis WTIV launched at Seatrium AmFELS (Brownsville, TX); commissioning delays follow.
• 2024-10-22 — Stonepeak 50% CVOW partnership closes; $2.6B proceeds.
• 2025-11-25 — SCC biennial-review order and GS-5 rate class approved; Chesterfield 944 MW gas plant approved.
• 2026-01-27 / 2026-02 — First commercial CVOW turbine installed; work resumes after brief federal pause and court relief.
• 2026-04-30 — Virginia law mandates load-forecasting review.
• 2026-06 / 2026-07-31 — SCC imposes $1.5M/MW GS-5 deposit; Q2 2026 discloses another ~$300M CVOW overrun (total $11.65B) and pushes completion to end-2027.

What people say

The case for. Sell-side is broadly Hold-to-Buy. Wells Fargo maintained Overweight at $68 (May 15, 2026, Shahriar Pourreza); BMO market-perform, target $70 (July 22, 2026). Bank of America’s Ross Fowler is at Hold, $61. The Q2 2026 print — operating EPS $0.79 versus $0.75 consensus on $4.48B revenue — supported management’s reaffirmation of FY2026 guidance ($3.45-$3.69 operating EPS, midpoint $3.57) and the 5-7% long-term CAGR. The pure-play thesis is intact: the November 2025 GS-5 approval, however haircut, still creates a mechanism for data-center-driven capex to earn a regulated return, and the Chesterfield gas plant approval (also November 2025) restored the reliability capex track. The Stonepeak partnership legitimately de-risked CVOW: Stonepeak funds 50% of remaining spend, so tariff overruns hit Dominion at 50 cents on the dollar.

The complaints. They are pointed. The Virginia AG’s Office of Consumer Counsel argued through 2025 that data-center load growth was the primary driver of rising fuel and purchased-power exposure at ratepayer expense, and the SCC broadly agreed — witness the 31% and 39% haircuts to Dominion’s 2026 and 2027 base-rate asks. Clean Virginia, SELC, Piedmont Environmental Council, Appalachian Voices, the NAACP and Mothers Out Front have all filed against the Chesterfield gas plant on VCEA grounds, and SELC is separately challenging the DEQ air permit in Richmond Circuit Court; the SCC suspended and then reasserted its approval in late December 2025. On CVOW, cost has moved from $7.8B at project inception to $11.65B as of July 31, 2026 — including a widely reported ~$800M-$900M in Trump tariff impact (per WVTF, July 31, 2026, and Bay Journal, 2026) and a ~$900M PJM transmission-interconnect increase — and completion has slipped six months to end-2027. On top of that sits the April 30, 2026 law mandating a review of Dominion’s load-forecasting methodology, filed after Google and other hyperscalers argued in the biennial case that Dominion inflates load by not adjusting for contingent projects. And Loudoun County residents and elected officials have been steady critics of the transmission-line siting required to serve the data-center corridor.

Outlook: well positioned or at risk?

At-risk. Not because Dominion loses its Virginia franchise — no US state has stripped a distribution monopoly in living memory — but because it is being handed a once-in-a-generation demand tailwind and the value is being deliberately re-priced away from it in real time.

Five vulnerabilities. First, the SCC has stopped rubber-stamping. The November 2025 order cut Dominion’s 2026 rate ask by 31% and its 2027 ask by 39%, gave it a 9.8% ROE against a 10.4% request, and — the more consequential structural move — created the GS-5 class with 85%/60% minimums and a $1.5M/MW deposit that shifts data-center cost risk to hyperscalers. That is a regulator explicitly designing the tariff so Dominion’s residential and small-C&I ratepayers do not fund stranded data-center substations, and it does so at the cost of the shareholder growth story. Second, CVOW keeps overrunning. From $7.8B to $11.65B is a ~50% cost blowout, six months late as of July 2026, and Stonepeak’s 50% partnership only takes the shareholder pain from 100 cents to 50 cents on the dollar. Third, the 2020 dividend cut is not repaired — five years later the payout remains at $2.67/share versus the pre-cut $3.76 run-rate, a public reminder of a balance sheet that was overcommitted to gas pipelines that no longer exist. Fourth, data-center bypass is real and accelerating — Talen/AWS, Constellation/Microsoft, Vistra/Meta and Bloom/Brookfield are actively pulling hyperscaler load outside the utility meter, at a moment when the incremental utility MW faces multi-year queues. Fifth, the load forecast itself is under review by state law — an outcome in which the number Dominion has been telling investors is 40-50% smaller than presented is not remote.

The bull counter is Bob Blue: an insider-lawyer CEO with deep Richmond political relationships, backed by a CFO from JPMorgan energy banking, delivering to the guide (Q2 2026 beat), still growing operating EPS 5-7% and — if CVOW hits at the current cost — adding the largest offshore-wind build in North America to rate base. Four things all have to work at once for the bull case to compound: CVOW ends at $11.65B and not $13B; the GS-5 tariff mechanics actually convert the 5.4% DOM Zone forecast into 6-7% EPS growth rather than a redirect to hyperscalers; the load-forecast review lands without a major writedown of planned transmission; and the next SCC biennial review does not tighten ROE further. Each is plausible. Requiring all four at once is the definition of at-risk.

How a challenger would attack it

Nobody attacks the franchise; they attack the meter. Dominion’s distribution monopoly is legally safe, so the challenger playbook — already running — is to make the incremental data-center megawatt never touch the utility at all. The economics now push that way from both sides: the SCC’s GS-5 class saddles any 25 MW+ customer with 85%/60% contracted minimums and a $1.5M/MW deposit, while Dominion’s interconnection queue runs years against Bloom fuel cells deploying in under two months. A challenger — a Brookfield- or KKR-financed on-site-generation platform — sells hyperscalers speed plus escape from Virginia tariff risk, priced with capital cheaper than Dominion’s equity. The second front is the existing fleet arbitrage: Talen, Constellation and Vistra are converting merchant nuclear into direct hyperscaler PPAs; every gigawatt locked to 2042 is rate base Dominion never books. The third is credibility: Google has already forced a statutory review of Dominion’s load forecasting for double-counting contingent projects, and the CVOW record — $7.8B to $11.65B, six months late — gives every intervenor and competitor the same argument: this utility cannot build to budget, so route around it. The attack requires no new technology, only Dominion’s own queue times, overruns and tariff structure doing the selling.

Same playbook, new buyer

Dominion’s actual playbook — convert concentrated large-load growth into regulated rate base — is being replayed better elsewhere, and the open variants run away from Virginia. The first shift is geographic: Loudoun’s saturation, transmission-siting fights and the GS-5 deposit regime are pushing data-center developers toward jurisdictions that want the load — and utilities in those states can pre-negotiate AEP-Ohio-style tariffs on their own terms rather than having Virginia’s imposed on them, capturing the growth story Dominion is watching get regulated away. The second shift is the buyer inversion: instead of selling regulated power to hyperscalers, sell hyperscalers the full off-balance-sheet package — generation, interconnection, collateral structuring — as an independent power platform; that is precisely the Bloom/Brookfield and merchant-nuclear position, and it prices below Dominion’s cost of equity. Third, the Charybdis asset points at a real niche: the only US-built Jones Act wind installation vessel is already chartered out to Ørsted and Eversource, and a services business built on scarce Jones Act marine capability monetizes offshore wind without owning construction risk. Dominion can follow none of these: it is jurisdictionally captive to Virginia, its balance sheet is committed to a $65B five-year plan plus CVOW overruns, and its dividend — still at the 2020-cut level — leaves no capital for adventures outside the rate base.

Sources and further reading

• Dominion Energy Q2 2026 slides: data centers surge, offshore wind 81% done — Investing.com, July 31, 2026.
• Dominion Energy reports Q2 2026 operating EPS $0.79, GAAP EPS $0.37; reaffirms 2026 guidance — TradingView, July 31, 2026.
• SCC Issues Order on DEV Biennial Review 2025 — Virginia SCC, November 25, 2025.
• SCC approves Chesterfield gas plant and Dominion rate hike, creates new rate class for data centers — Virginia Mercury, November 25, 2025.
• Virginia Now Makes Data Centers Post $1.5 Million A Megawatt — Forbes, June 9, 2026.
• New state law mandates review of Dominion’s load forecasting, as data centers raise concerns — Virginia Mercury, April 30, 2026.
• Cost increases to $11.2B for Dominion offshore wind project, with tariffs to blame — WAVY, 2026.
• Trump tariffs added over $800 million to Dominion Energy’s offshore wind project — WVTF, July 31, 2026.
• Dominion Energy, Stonepeak Announce Closing of Sale of Noncontrolling Equity Interest In CVOW — Dominion IR, October 22, 2024.
• Berkshire Hathaway buys Dominion Energy natural gas assets in $10 billion deal — CNBC, July 5, 2020.
• Dominion Energy Cuts Its Dividend by 33% — Nasdaq, November 4, 2020.
• PJM trims near-term load forecast on stricter data center vetting, economic outlook — Utility Dive, January 2026 (5.4% DOM Zone).
• Dominion Energy Form 10-K FY2025 — SEC, February 2026.
• Bloom Energy Lands $25 Billion Brookfield Commitment for AI Data Center Fuel Cells — TechTimes, July 2, 2026.
• 2024 Virginia Business Person of the Year: Robert M. Blue — Virginia Business, 2024.