Weekly Storage and S/D Brief - week ending 7/10/2026

Headline: Modeled injection drops to +36 Bcf

A two-week power burn surge of +8.5 Bcf/d — driven by sustained mid-summer heat, coal-to-gas switching, and rising electric load — overwhelms modest offsets from seasonal heating declines and LNG maintenance.

Storage Summary

Injections have fallen sharply over three consecutive weeks — from +83 Bcf to +61 Bcf to a modeled +36 Bcf — a cumulative 47 Bcf/wk decline.

Primary Driver: Power Burn (+8.5 Bcf/d over Two Weeks)

The injection decline is overwhelmingly a power sector story. Gas-fired generation surged from 40.6 Bcf/d (week ending 6/27) to 46.1 Bcf/d (wk 7/03) to 49.1 Bcf/d (wk 7/10) — a cumulative +8.5 Bcf/d ramp over just two weeks. At 49.1 Bcf/d, power burn is running at summer highs and is 2.2 Bcf/d (+4.6%) above the same week last year (46.9 Bcf/d).

Weather: Same Average, Different Profile

A critical nuance: weekly average CDDs were nearly identical between the two weeks (12.5 vs 12.4), yet power burn jumped another +3.0 Bcf/d. The explanation lies in the daily distribution:

• Week ending 7/03 ramped into heat — CDDs started at 8.8 on Saturday and climbed to 15.3 by Friday. The cool weekend start dragged down the weekly average and limited the first few days of burn.
• Week ending 7/10 sustained heat throughout — CDDs never fell below 10.9, with every day registering as a high-burn day. The cooling load fully caught up to the elevated temperatures.
• South Central: Load boomed +12% but gas burn is actually down 2.2% on a weather-adjusted basis — solar generation surged +60% and wind +39%, absorbing all incremental demand and then some.
• Pacific: Gas burn down 11.3% weather-adjusted, as hydro (+23%) and solar continued to displace gas-fired generation.
• East: Gas burn up +1.5%, the one region where load growth (+4.3%) is translating to more gas burn — renewables were essentially flat YoY.

Deeper Dive: What's Behind the Power Burn

The raw year-over-year power burn increase of +2.2 Bcf/d (+4.6%) for the week ending 7/10 is itself a story of multiple competing forces. Analysis of EIA-930 hourly grid data and regional demand patterns reveals three distinct drivers:

1. Weather — The Dominant Short-Term Force

July 2026 has been hotter than July 2025, with population-weighted CDDs approximately 1.1/d higher on a month-to-date basis. Regression analysis suggests roughly 2.0 Bcf/d of the YoY burn increase is purely weather-driven. After adjusting for weather, national gas burn is essentially flat year-over-year (-0.6%), meaning the structural supply-demand balance for gas in the power stack has not materially changed.

2. Coal-to-Gas Switching — A Floor Under Burns

Henry Hub front-month prices are running ~$2.90/MMBtu, down 12.1% from ~$3.30 at this point last year. Cheaper gas has made it more competitive against coal, particularly in coal-heavy regions:

The Midwest (+7.7%) and Mountain (+11.8%) regions are seeing the most pronounced coal-to-gas switching, with EIA-930 data showing coal generation down 6.3% and 37.7% respectively in those regions through mid-July. Cheap gas is providing a structural floor under burn even in weeks where weather might otherwise allow it to ease.

3. Electric Load Growth vs. Renewable Displacement — The Tug of War

National electric load is up 3.9% year-over-year (571 → 593 GW average), reflecting ongoing demand growth from data centers, electrification, and population. However, renewable generation is up 14.8% YoY, capturing nearly all incremental load nationally. The regional picture varies sharply:

The net effect nationally: weather-adjusted gas burn is flat. Load growth and coal-to-gas switching push burns higher, while accelerating renewable deployment — particularly solar in the South Central and Pacific — pulls them back. The current heat wave is masking this structural equilibrium by layering weather-driven demand on top.

Secondary Factors

LNG feedgas eased 0.57 Bcf/d week-over-week to 18.68 Bcf/d, likely reflecting terminal maintenance activity. Res/Com demand continued its seasonal fade (-0.30 Bcf/d) as heating loads diminish. Production was essentially flat at ~110.5 Bcf/d, and Canadian imports provided a marginal 0.22 Bcf/d supply boost. None of these factors came close to offsetting the +3.0 Bcf/d power burn increase.

Model Performance

The SynMax S&D model has posted a 4-week mean absolute error of approximately 3 Bcf against EIA actuals.

See the full s/d analysis on the dashboard

Take a deeper dive into SynMax's power burn data here

New Dataset Release: US Demand

SynMax has released a new US Gas Demand Dataset for our Hyperion Clients.  It consists of a daily demand estimate, broken out by EIA gas storage region and by demand component.

It covers the four weather-driven end-use sectors (Residential, Commercial, Industrial, and Electric Power), built as an ensemble of pipeline flow data and weather-driven modeling, calibrated to EIA's monthly totals. It also includes LNG feedgas at all US liquefaction and regasification terminals, pipeline trade flows with Mexico and Canada, and supporting components like lease/plant fuel and pipeline/distribution use — giving a complete, regionally resolved daily picture of the lower-48 gas balance.

The data is currently out on query_datalinks and on Agents and will be rolled out to the SynMax frontend and the traditional API over the coming weeks. See here for overview and access methods, and here for full methodology and details.

As usual, contact support@synmax.com with questions.