US Rig Count UP as We Compare Baker Hughes, Enverus, and WellDatabase

• Eagle Ford: 50 (−1 week, +5 year from 45).
• Granite Wash: 20 (+2 week, +5 year).
• Cana Woodford: 20 (−1 week, +2 year).
• OTHER: 101 (+3 week, +3 year) — the residual bucket is not shrinking.
• Williston: 34 (−30 week). Same caveat as North Dakota.
• Marcellus 23 and Utica 10: flat to slightly down year over year.

County-level heat on the WellDatabase map remains West Texas / Southeast New Mexico (Lea, Eddy, Midland, Martin, Howard, Reeves, Loving, Ward) with secondary clusters in Oklahoma and the Haynesville corridor.

Top operators (WellDatabase current / week / year-ago)

Unknown 34, Continental Resources 25 (+3 week, +4 year), Diamondback 18, EOG 18 (−1 week, −7 year), Pioneer 17 (−1 week, −9 year), Permian Resources Operating 16 (+9 week), Mewbourne 13, Hilcorp 11, SandRidge 11, Devon 10.

The operator tape is mixed: some large publics are leaner than a year ago (EOG, Pioneer), while Continental, Diamondback, Hilcorp, SandRidge, and Permian Resources show more activity. That lines up with Enverus research that private and smaller operators have taken a disproportionate share of incremental rigs this cycle.

DUC wells: five years of drawing the inventory down

Drilled-but-uncompleted wells are the buffer between a rig and first oil or gas. The last five years have been a drawdown, not a rebuild.

• Pandemic-era peak was roughly 8,900 DUCs in 2020.
• By December 2024 the EIA shale-region count was about 5,798.
• By December 2025 it was about 5,020 (−14% year over year; Permian down sharply from ~1,280 to ~821 in that World Oil recap of EIA data).
• April 2026 EIA estimates were about 4,972 — described at the time as the lowest in the series back to 2013 after 14 months of completions outrunning new drills.

August 2026 snapshots put the EIA-style national total near 4,385–4,919 depending on the aggregator, with Permian DUCs around 839 (+15 month-over-month in one OilPriceAPI/EIA cut). MacroMicro shows 4,919 for August.

From the ~8,900 peak to ~4,900 is a draw of roughly 45%. From late-2024 levels near 5,800 to ~4,900 is another ~15%. Completions have been using the backlog up. That is why production can stay high even when the rig count is only in the high 500s: longer laterals, faster cycle times, and a smaller but still-usable DUC queue.

Two caveats matter. First, EIA DUCs are the standard market series and are not the same as a raw regulator extract of every well that has a spud and no first production. Buckhead Energy’s September 26 state-regulator rollup across 11 mineral states shows 9,815 wells in that broader inventory (Texas and New Mexico each listed at 2,000 on that methodology). Older wells in those extracts may never be completed. Enverus has previously argued that excluding wells drilled more than two years ago produces a much smaller “live” DUC number. Second, some basins have started to replenish as prices and rigs firmed in 2026; Permian DUCs ticking up in August is consistent with that, not a return to 2020 inventory.

Are we using them up? Yes, on a five-year basis. The industry no longer has an 8,000–9,000 well completion overhang. The remaining inventory is a working queue, not a strategic stockpile.

Oil storage and production cross-check

Rig counts are a leading indicator. The EIA Weekly Petroleum Status Report is the lagging physical check.

For the week ending September 18, 2026 (released September 23):

• U.S. crude production: 13.939 million b/d, essentially unchanged from 13.944 the prior week and in record territory.
• Commercial crude inventories excluding SPR: 426.4 million barrels, +3.0 million on the week, about 2% above the five-year average for this time of year. Analysts had expected a small draw.
• morningstar.com

Cushing, Oklahoma: 23.7 million barrels, +2.3 million. Still below the five-year average even after the build.
SPR: 284.6 million barrels, −0.4 million on the week. That is historically low (roughly 43% of the 2020 peak near 656 million) and far below year-ago levels.

Refinery utilization slipped to 94.0% from 96.8%, which helped the unexpected crude build.

The cross-check is straightforward. Production is running near 13.9 million b/d with a 599-rig Baker Hughes count — efficiency, not a 2014-style rig army. Commercial stocks are no longer at the emergency-tight levels seen earlier in 2026, but they are not sloppy either. Cushing rebuilt off the floor and is still not overflowing. The SPR remains a structural drawdown story. A 4-rig weekly add does not flood the market next month; DUC drawdown plus 10k–15k ft horizontals is how 13.9 million b/d happens at these rig levels.WTI settled around $92.41 on September 25 per the Odessa American recap of the Baker Hughes week, down sharply on the week even as the rig count rose — a reminder that the rig tape and the price tape do not move in lockstep.

Bottom line

Baker Hughes is up. WellDatabase is flat-to-up year over year once you look through the North Dakota reclass. Enverus is higher still. Growth is concentrated in Texas, New Mexico, Oklahoma, the Permian, and Haynesville gas. Oil rigs still set the level; gas rigs set more of the year-over-year change. DUC inventories have been worked down by roughly 45% from the 2020 peak and are no longer a large spare-capacity reservoir. U.S. crude output near 13.9 million b/d with commercial stocks only modestly above the five-year average is the production result of that efficiency-plus-drawdown regime.

The weekly number to watch next is not whether WellDatabase prints 598 or 628. It is whether Permian and Haynesville hold, whether the ND/Williston classification settles, and whether completions keep converting the remaining DUC queue fast enough to hold 13.9 million b/d if prices stay choppy. We love our subscription to WellDatabase.com, as we can get the granular data we need when writing articles on operators or oilfield service companies.

Making Appendices Great Again

Check out the World’s Greatest Podcast Show Notes at EnergyNewsBeat.co or EnergyNewsBeat.com.

Appendix: sources and links