I keep coming back to the same stubborn number. Henry Hub averaged $2.93 per million British thermal units from June through August, about 6% below the same stretch a year earlier, even though the Lower 48 just lived through its hottest July on record. That is not how the summer script is supposed to run. Heat should squeeze the power stack, burn more gas, thin inventories, and give prices a lift. This time the heat showed up. The lift did not.
What Actually Happened To US Natural Gas Prices This Summer
Average temperatures across the Lower 48 reached 77°F in July. Air conditioners ran harder. Electricity demand rose. On paper, that is a textbook setup for US natural gas prices to firm. In practice, the market absorbed the extra cooling load without looking remotely tight. I have watched a few of these summers now, and this one felt different in a quiet way. The demand was real. The scarcity was not.
Perhaps the most interesting aspect is how little drama the price tape showed. Traders did not get a late-July squeeze that stuck. Utilities did not scramble in a way that left a scar on inventories. The market just kept finding gas. That is the story worth unpacking, because it tells you more about the next two winters than any single heat-wave headline.
Record Heat Raised Power Burn, But Not Enough To Matter
Hot air still matters. Nobody should pretend otherwise. When overnight lows stay elevated and afternoon peaks linger, gas-fired plants stay online longer. Combined-cycle units that might have sat idle in a mild July get dispatched. Peakers fire. That incremental burn is the classic summer support for Henry Hub.
This year that support arrived, then got crowded out. Natural gas-fired generation did rise, by an estimated 7.5 billion kilowatt-hours from June through August versus the same window a year earlier. That is not nothing. It is also not the headline. Solar generation jumped by about 19.4 billion kilowatt-hours over the same comparison. Wind added another 9.3 billion. Put those two together and the increase from wind and solar was nearly four times the increase from gas-fired plants.
The hottest July on record increased gas-fired power generation without tightening the market enough to lift Henry Hub above last summer’s average.
I’ve found that people still talk about summer gas as if renewables were a rounding error. They are not, not anymore, not in the months when solar output is strongest. The extra cooling load did not land entirely on the gas fleet. A large share of it landed on panels and turbines that do not need a pipeline nomination.
Why The Generation Mix Changed The Price Outcome
Think of the power stack as a line at a busy coffee shop. Heat adds customers. If the shop hired extra baristas who work for almost nothing at midday, the old espresso machine does not have to run as hard. That is a crude analogy, sure, but it is close to what happened. Midday solar cut into the hours when gas plants used to enjoy a clean run. Wind filled other gaps. Gas still cleared the peaks and the evenings. It just did not clear as much incremental energy as the temperature anomaly implied.
In my experience, this is the part of the market that still gets underestimated in casual conversation. A 19.4 billion kilowatt-hour swing in solar is a lot of megawatt-hours that never become dekatherms. Convert that loosely and you are talking about a meaningful volume of gas that never had to be burned. Add wind and the displaced burn gets larger. The heat was record-setting. The residual gas call was not.
- Solar generation rose by an estimated 19.4 billion kilowatt-hours from June through August versus the prior year.
- Wind generation added about 9.3 billion kilowatt-hours over the same period.
- Gas-fired generation increased by only 7.5 billion kilowatt-hours.
- The renewable increment was nearly four times the gas increment.
Does that mean gas is disappearing from summer power? No. It means the marginal ton of cooling demand is less exclusively a gas story than it used to be. That distinction is everything for prices.
Supply Grew Fast Enough To Absorb The Extra Burn
Demand is only half the ledger. The other half was generous. US dry natural gas production averaged 2.7 billion cubic feet per day higher from June through August, a 2% increase. The Permian was among the biggest sources of that growth, which should surprise no one who has watched associated gas keep arriving even when oil prices wobble.
A 2.7 Bcf/d lift does not sound theatrical until you sit with it for a minute. That is a persistent daily surplus layered on top of an already large base. Official outlooks now point to dry gas production averaging a record 111.2 Bcf/d for the full year. Record output is not a weather event. It is a structural condition. Weather can still squeeze a structurally long market. It just has to work much harder to do it.
I keep thinking about how different this would have felt five or six years ago. Back then, a hottest-on-record July with rising power burn would have had a better chance of leaving a mark. Production was less elastic. Associated gas from the Permian was a smaller share of the national total. The system had fewer ways to shrug.
Storage Started Comfortable And Stayed Comfortable
Inventories walked into the April injection season with 1.906 trillion cubic feet of working gas, 4% above the previous five-year average. That opening balance matters more than people admit. A market that starts the fill season already a little heavy has room to absorb a hot July without looking desperate in September.
Monthly injections beat their respective five-year averages in every month through August except May. Read that again. The heat arrived and the industry still stuffed gas into the ground at an above-normal pace for most of the season. That is not the behavior of a tight market. That is the behavior of a market with surplus molecules looking for a home.
| Market Balance Piece | Summer Signal | Price Implication |
| July temperatures | Hottest Lower 48 July on record | Bullish for power burn |
| Gas-fired generation | Up 7.5 BkWh year over year | Mildly bullish |
| Wind plus solar | Up nearly 29 BkWh combined | Bearish for residual gas call |
| Dry gas production | Up 2.7 Bcf/d, about 2% | Bearish |
| Storage start | 1.906 Tcf, 4% above five-year average | Bearish |
| Injections through August | Above five-year average in most months | Bearish |
| Expected end-October inventories | About 3.985 Tcf, 5% above five-year average | Caps winter risk premium |
Official estimates now look for Lower 48 working gas inventories near 3.985 trillion cubic feet by the end of October, roughly 5% above the five-year average. If that landing zone holds, winter starts with a cushion. Cushions do not make prices immune to a polar vortex. They do make a routine cold snap less explosive.
LNG Maintenance Quietly Left More Gas At Home
Exports usually act as a pressure valve in the other direction. When liquefaction trains run hard, domestic balances tighten. This summer, maintenance at US LNG terminals moderated demand growth from the export sector. That left more gas available for power generation and storage precisely when air conditioners were working overtime.
Timing matters here. A maintenance calendar that overlaps peak cooling is not a crisis. It is a temporary gift to the domestic market. Feedgas that would have left the Gulf Coast stayed in the interstate grid. Some of it went to power. Some of it went underground. Either way, it did not leave the country in the form of cargoes.
I would not treat summer maintenance as a permanent bearish pillar. Trains come back. New capacity keeps arriving over the next couple of years. The point is narrower. During the exact months when heat should have been most bullish, one of the fastest-growing sources of demand was running below its potential. That coincidence helped keep Henry Hub from catching a bid.
Henry Hub At $2.93 Was A Verdict, Not An Accident
A $2.93 summer average, 6% below last year, is the market rendering a judgment. Heat raised consumption at the margin. Renewables, production growth, fat storage, and softer export pull raised available supply by more. Prices did what prices do when the second list is longer than the first. They sagged.
There is a temptation to call this a fluke. I do not buy that. The pieces that capped prices are not one-off curiosities. Solar capacity keeps rising. Associated gas from oil basins does not vanish because July was sweaty. Storage infrastructure is still large. LNG will grow, yes, but it grows in steps, and those steps can be interrupted by outages and work schedules.
Record production, strong storage injections and nearly 29 billion kilowatt-hours of additional wind and solar generation kept the gas market well supplied through the peak cooling months.
That sentence is the whole summer in one breath. You can dislike the implication. You cannot really argue with the arithmetic.
How Traders Usually Misread A Hot Summer
Hot weather headlines are easy to trade emotionally. The map turns red. Power prices pop in Texas or the Mid-Atlantic. Someone on a desk says the gas bid is coming any minute. Sometimes that is right for a day or two. This summer it was a poor guide to the three-month average.
The mistake is treating temperature as a sufficient statistic. Temperature is an input. The output that matters is residual gas burn after other resources have taken their slice, minus the production and import-export residual, stacked against the storage trajectory. That is a mouthful. It is also why a record July can coexist with a soft Henry Hub print.
- Start with the temperature anomaly and the implied cooling degree days.
- Subtract the extra energy supplied by solar and wind during those same hours.
- Add back the gas that still has to cover ramps, peaks, and cloudy or still intervals.
- Compare that residual burn with year-over-year production and LNG feedgas.
- Check whether injections are still beating seasonal norms. If they are, the tape has already voted.
Skip a step and you will overpay for a weather story that the balances have already digested. I have done that. Most people who have traded this market long enough have done that. The humility is useful.
The European Contrast Is Real, And It Does Not Rescue Henry Hub
It is fair to note that the Atlantic Basin is not one market with one mood. Europe can look tighter, more political, more exposed to pipeline politics and winter storage politics, while the US looks almost bored. That contrast is real. It is also a poor reason to force a bullish US summer narrative that the domestic numbers refuse to support.
US prices can stay modest while overseas benchmarks do something else, at least until liquefaction and shipping stitch those differences together more tightly. Maintenance, capacity additions, and shipping costs all sit in that seam. This summer the seam did not pull enough US molecules out of the country to offset the domestic surplus.
If you are sitting on a US producer book, that is frustrating. Heat without price is a missed opportunity. If you are sitting on a utility book, it is a gift. Cheap summer gas going into a well-supplied October is the kind of quiet win that never makes a viral chart and still shows up in the fuel budget.
What This Summer Says About The Coming Shoulder Season
Shoulder months are where surplus becomes visible. Cooling fades. Heating has not started. LNG either ramps or it does not. Storage either approaches the top of the working range or it finds a bid from winter risk. Right now the path of least resistance still looks like comfortable inventories into late October.
That does not lock prices at $2.something forever. A colder-than-normal November can still matter. A sudden train outage can still matter. A surprise drop in associated gas if oil activity slumps can still matter. The base case, though, is not a market that enters winter already short. It is a market that enters winter with room.
I’ve found that the best way to stay honest here is to watch injections week by week instead of arguing with the temperature maps. If fills keep beating norms after the heat has broken, the surplus was never a mirage. If fills suddenly stall while production holds, then the market is telling you demand found another gear. Watch the weekly, not the anecdote.
Production Growth Is The Quiet Villain Of The Price Rally
Everybody wants a simple villain. Heat should have been the hero for bulls. Production growth played the other role. A 2% increase sounds small until you remember the base is enormous. Two percent of a record-scale industry is a lot of gas looking for a bid every single day.
The Permian deserves special mention because associated gas does not wait for a Henry Hub signal. It shows up because oil wells are being drilled and completed. That makes the gas market partly a passenger in someone else’s vehicle. When oil activity is firm, gas supply can stay stubbornly high even if gas prices are soft. That is an awkward feedback loop for anyone hoping a cheap summer would ration output quickly.
Will producers throttle back? Some dry-gas weighted operators already think in those terms. The associated-gas share complicates the response. That is one reason prices can grind lower for longer than a tidy supply-demand sketch suggests.
Renewables Did Not Kill Gas. They Blunted The Weather Premium
Let me be plain. Gas remains the flexible workhorse of the US power system. When the sun drops and the wind dies and the load stays high, gas plants still answer the phone. That job is not going away this decade. What changed is the size of the weather premium attached to a hot summer.
A decade ago, almost all of a heat-driven load surge fell on thermal plants, and a large share of that thermal response was gas. Today a chunk of the energy, if not always the peak capacity, is met by resources with zero fuel cost. The capacity value conversation is separate and messier. The energy conversation is simpler. Extra kilowatt-hours from solar and wind mean fewer dekatherms burned for the same cooling outcome.
Summer balance, stripped down: Heat raises load Solar and wind take a larger slice than last year Gas still covers ramps and peaks Production and storage more than cover the residual Price stays soft
That sketch is ugly and it is useful. It keeps you from treating every red weather map as a buy signal.
Storage Economics When The Market Is Already Long
Injecting gas into a market that is already 4% above the five-year average is not an act of panic. It is an act of carry. Traders and utilities park molecules because winter still has option value and because selling everything into a soft summer tape looks worse. That behavior can persist until the working gas number starts to look like a logistical problem rather than a statistical surplus.
Are we near that logistical problem? An expected 3.985 Tcf by the end of October, about 5% above the five-year average, is ample. It is not an automatic overflow event in every region. Some basins and some storage operators will feel the weight more than others. Basis will tell that story more honestly than the national headline.
Still, ample national storage changes the shape of winter risk. You need a longer cold event, or a bigger supply outage, or a stronger LNG pull, before inventories stop being a shock absorber and start being a constraint. That bar is higher than it was on the first of June.
What Would Have Been Required To Lift Prices
It is worth asking the counterfactual. What would a hot July have needed to push Henry Hub above last summer’s average? A few paths come to mind, and none of them showed up together.
- A much smaller renewable increment, so that almost all extra cooling load fell on gas plants.
- Flat or falling dry gas production instead of a 2.7 Bcf/d increase.
- LNG facilities running full tilt with no notable maintenance.
- A leaner start to injection season rather than a 4% surplus.
- A heat event that lasted deeper into August with weaker wind.
Stack two or three of those and you get a different summer. Stack none of them and you get $2.93. Markets are rude that way. They do not pay you for the headline. They pay you for the residual.
A Note On How This Feels If You Live In The Physical Market
Paper traders talk in averages. Physical desks talk in constraints. A national surplus can still leave a local utility short on a 108°F afternoon if a plant trips or a pipeline segment is derated. Those local squeezes are real. They just did not add up to a national price event this time.
That gap between local pain and national price is getting wider as the resource mix diversifies. A solar-heavy afternoon can look easy on the gas system until 7 p.m., when the ramp hits. Those ramps still need gas, and they still need pipeline flexibility. Flexibility has a price. Energy volume is what failed to tighten this summer. Capacity and ramp products are a different conversation for another day.
In my experience, blending those two conversations too quickly is how people talk past each other. One side says gas is drowning in supply. The other side says the grid still needs every flexible plant it can keep. Both can be true in the same week.
The Investor Read, Without The Hype
If you follow producer equities, a soft summer after record heat is a reminder that volume growth can outrun weather. Realized prices matter more than the temperature graphic in the weekly note. If you follow midstream, throughput held up even if the commodity price did not. If you follow power, the renewable share of incremental summer energy is no longer a footnote.
None of that is a recommendation to do anything. It is a map of where the summer actually spent its energy. Cheap feedgas into well-supplied storage is constructive for winter reliability and less constructive for anyone who needed a price spike to justify the last increment of drilling.
Tax and basis details will vary by basin. Appalachia does not live the same summer as the Haynesville or the Permian. National Henry Hub is the scoreboard. Local cash is the game film. Watch both.
Winter Is Still Allowed To Surprise People
I want to leave a door open, because markets enjoy humiliating tidy summer essays. A harsh winter can still chew through a 5% inventory surplus. A cluster of LNG trains returning from maintenance at the same time as a cold blast can still tighten the complex. A production stall is not impossible if oil activity rolls over hard.
Those are ifs. The evidence from June through August is not an if. Heat arrived. Power burn rose. Prices did not. The reason was not a mystery. Extra wind and solar took a large share of the new load. Production ran ahead. Storage kept filling. Exports did not pull as hard as they might have. That combination is enough.
Weather can still squeeze a structurally long market. It just has to work much harder to do it.
That is the sentence I would tape to the monitor heading into November. Not as a slogan. As a filter. If the next cold snap is ordinary, fade the first spike until balances confirm it. If the next cold snap is extraordinary and LNG is running wide open, then the filter flips. Process first. Narrative second.
The Uncomfortable Lesson For Anyone Who Trades Weather
Weather markets reward people who update. They punish people who reuse last year’s playbook. Last year’s playbook said record heat plus power demand equals a firmer summer strip. This year’s tape said record heat plus a transformed supply stack equals a shrug.
Will next summer look like this one? Only if the same four conditions rhyme: strong renewable energy growth in the cooling months, ongoing production gains, a healthy storage start, and no runaway export pull during peak heat. Break one of those and the old correlation can wake up. Keep all four and the weather premium stays smaller than the headlines.
I do not find that boring. I find it clarifying. The gas market is not ignoring heat. It is pricing heat against a bigger machine. The machine got bigger. The heat, record or not, was not big enough to overwhelm it.
A Practical Checklist Before The Next Heat Headline
When the next map goes crimson, run a short list before you assume Henry Hub must follow. It takes five minutes and it will save you from a few bad longs.
- Is the heat concentrated in gas-heavy power regions, or is it sitting over areas with large solar fleets?
- What is the year-over-year change in wind and solar generation during the same weeks?
- Is dry gas production still running above last year by a couple of Bcf/d?
- Are storage injections beating the five-year average even while the heat persists?
- Are LNG terminals in maintenance or running near nameplate?
- Is the summer strip already pricing a scare, or is it still asleep?
If the answers look like this past summer, you do not have a shortage. You have a weather story competing with a surplus story, and the surplus story won. If the answers flip, you have a trade. Either way, you are no longer guessing from the temperature alone.
Closing The Books On A Hot, Cheap Summer
So here is where I land. The Lower 48 posted its hottest July on record. Electricity demand rose because people needed to stay cool. Gas plants did more work than they did a year earlier. And Henry Hub still averaged $2.93 from June through August, 6% cheaper than the same months last year. That is not a paradox once you line up the other columns.
Solar and wind added nearly 29 billion kilowatt-hours of extra generation. Dry gas production ran 2.7 Bcf/d higher. Storage entered spring already 4% above the five-year average and kept posting strong injections. LNG maintenance left more gas at home. Expected end-October inventories near 3.985 Tcf would sit about 5% above the five-year average. The hottest month did not get a chance to look like a shortage.
Maybe that leaves you disappointed if you wanted a rally. Maybe it leaves you relieved if you buy fuel for a living. Either reaction is human. The balances do not care. They already filed their report. Record heat failed to lift US natural gas prices because the rest of the system showed up oversupplied, flexible, and a little too good at absorbing a punch.
Winter can still rewrite the last chapter. Until it does, the honest read of this summer is simple. The thermometer ran hot. The market stayed long. Price followed the inventory, not the weather map.
