The EIA's August 2026 Short-Term Energy Outlook puts Henry Hub at roughly $2.87/mmBtu for Q3 2026 — down about $0.50 from the prior forecast — with prices expected to stay below $3.00/mmBtu through November. End-of-October storage is projected at 3,985 Bcf, running about 5% above the five-year average. By the inflation-adjusted historical record, this is the 18th percentile of months since 2010. Gas is genuinely, structurally cheap right now in the United States.
The conventional read: cheap gas is bad for nuclear. If gas-fired generation costs less to dispatch, nuclear looks expensive by comparison. Utilities optimize on marginal cost. Case closed.
That read is wrong — or at least, it's incomplete in ways that matter enormously for anyone thinking about where the grid goes over the next decade.
Cheap Gas Doesn't Stay Cheap — and the Infrastructure That Replaces It Takes Decades
Here's what the conventional read ignores: the same global LNG disruption that's suppressing U.S. Henry Hub prices is simultaneously creating a European energy emergency. EU gas storage sits at just under 58% full — the lowest level for this time of year in records going back to 2011, and 12 percentage points behind last year. European prices were running near €53/MWh as of early August, roughly double pre-war levels. Energy Aspects estimates that a cold winter with no Qatari LNG could push average day-ahead prices to €110/MWh across November through March.
The mechanism connecting these two facts: the U.S.-Iran war effectively removed Qatar — the world's second-largest LNG exporter — from global markets. Qatar's Ras Laffan loadings were running about 60% below year-ago levels as of August 11, even as the country began preparing equipment for a potential Hormuz reopening. Meanwhile, ADNOC Gas is actively exploring a new LNG export facility on the UAE's east coast specifically to bypass the strait — a project that, if it ever gets built, is years away from first cargo.
The point: the global LNG system is fragile in ways that domestic Henry Hub prices don't reflect. U.S. gas is cheap because it's landlocked by infrastructure constraints and temporarily insulated from the Hormuz disruption. That insulation is not permanent. And the nuclear plants being licensed and constructed today will operate for 60 to 80 years across many commodity cycles.
Nuclear's Actual Competitive Advantage Has Nothing to Do With Spot Prices
The capacity factor argument deserves to be stated plainly, because it gets lost in fuel-cost comparisons. The DOE reports that nuclear energy operates at roughly 92% capacity factor — nearly twice that of natural gas and coal units, and three or more times that of wind and solar. A 1 GW nuclear plant delivers roughly twice the actual electricity output of a 1 GW gas plant when you account for how often each runs.
This matters for the AI-driven demand surge that's reshaping the grid. Reuters reported on August 5 that data centers, electrification, and new manufacturing investment are accelerating power demand in ways that require reliable, around-the-clock generation. Gas remains deeply embedded in U.S. grid architecture — there are roughly 2,000 gas-fired power plants and around 3 million miles of gas pipelines — but the operational flexibility that makes gas attractive also makes it dependent on commodity prices that can move violently.
Data centers don't want cheap power. They want guaranteed power at a predictable price for 20-year contract horizons. That's a fundamentally different procurement problem than optimizing today's dispatch stack. Nuclear's high capacity factor and fuel-cost stability are features, not bugs, for this buyer profile.
The Retail Price Signal Points the Other Way
One more data point worth sitting with: EIA's May 2026 electricity data shows average retail revenues up 5.3% year-over-year to 13.83 cents/kWh nationally. Commercial sector consumption — the category that includes data centers — grew 3.4% year-over-year in May alone. Electricity prices are rising even as Henry Hub falls. The spread between cheap wholesale gas and rising retail electricity rates reflects transmission constraints, grid investment costs, and the structural demand surge that no amount of cheap gas can fully accommodate.
That spread is exactly where nuclear's long-run economics live. The fuel cost of a nuclear plant is a small fraction of its total cost structure; the capital is the dominant variable. Once built, a nuclear plant is largely insulated from Henry Hub volatility in either direction.
So yes, $2.87/mmBtu gas makes today's nuclear dispatch look expensive on a marginal-cost basis. But the civilizational question — which generation assets do you want running in 2045, 2055, 2065? — doesn't get answered by this quarter's NYMEX strip. It gets answered by the construction decisions being made right now, while the gas market is distracted by its own temporary abundance.
Watch for the EIA's next Short-Term Energy Outlook on August 26, which will update both the Henry Hub forecast and storage trajectory as the Hormuz situation evolves. If Qatari loadings recover and global LNG supply normalizes, the European price signal will ease — but the structural case for nuclear baseload won't move an inch.
