Permian Completions Weekly

By job end date. Faded columns are recent weeks still filling in — operators have up to ~30 days to file, so the newest weeks always start low and grow.

Rigs drill the wells; frac spreads complete them. Both are active-unit counts on one axis — a widening gap hints at the DUC balance shifting. Note the scopes differ: rigs are Permian-only, spreads are U.S.-wide. · Sources: Baker Hughes; Primary Vision (via AOGR)

From NM OCD well records, horizontal wells in Lea / Eddy / Chaves / Roosevelt: 8,284 approved permits not yet spudded (the drilling backlog — includes some older unexercised permits) and 857 wells spudded 2023+ with no FracFocus disclosure yet — drilled or drilling, not yet frac'd (some are mid-completion or awaiting the operator's disclosure filing), of which 261 were spudded in the last 90 days — the wells hitting frac schedules next. Every well in the drilled column needs a frac crew and a drillout. Texas permit pipeline (RRC W-1 files) is the next data source to wire in.

From the RRC drilling permit master (current through early July): 4,862 approved horizontal permits issued since Jan 2024, not yet spudded or completed, in the 35 Permian counties. Texas receives spud dates late (typically with completion filings), so a drilled-not-yet-frac'd column isn't reliably possible from TX public data — the NM table above is the timelier read on that stage.

The other 54 operators holding 1,391 permits below the top table — the private and family programs where a new service company or a sales hand can actually get a look, instead of queueing behind incumbents at ExxonMobil or EOG. Majors and large independents are filtered out even when a slow permitting quarter drops them below the top table. Green dates = a fresh permit inside 30 days, i.e. actively planning wells right now.

Everyone outside the top 10 with 2+ completions since March 1 (31 operators) — the private and mid-cap programs that don't make headlines but buy sand, water, and services every month. Majors and large independents are excluded regardless of job count. Entities reflect current ownership (e.g., Double Eagle's retained DE Central position; Tap Rock's NM wells now under Jonah Energy).

Median job design over each operator's own most recent 90 reported days — dark filers (e.g., EOG, whose disclosures stop in Dec 2025) appear at last-reported design rather than vanishing · operators with ≥8 jobs reporting water · "n/d" = proppant mass not disclosed (EOG reports water volumes but not sand mass) · basin median sand concentration 1.07 lbs/gal

Median job design by true vertical depth band — a rough proxy for target bench (141 of 155 jobs report TVD, water, and proppant)

FracFocus does not disclose target formation or perforated-interval length. Formation names (Wolfcamp, Bone Spring, Spraberry…) and true per-lateral-foot intensity (lbs/ft, bbl/ft) are planned additions via Texas RRC completion records and NM OCD pool data.

Jobs ended in last 28 days · formation resolved for 101 of 155 jobs by joining state records (TX RRC wellbore filings, NM OCD pool assignments) by API number

Jobs ended July 2025 – present with a completed-lateral length on file (1526 jobs). State completion filings trail the frac by ~3–6 months, so this table reflects the settled record, not the newest wells. TX lateral = perforated interval span (RRC); NM lateral = MD − TVD − 600 ft curve allowance (OCD).

Inferred from disclosure cadence: a pad whose latest job ended within ~1.5× its own well-to-well pace is probably still being pumped or just finishing. 14 pads read as likely active as of the latest filings (through Jul 21), 20 more recently released. Disclosures lag the frac by up to 30 days, so this view sees roughly half of true live activity — treat as a floor and a lead list, not a map. Pumper marked * is inferred from the operator's dominant 2026 service company. Field confirmations will upgrade rows from "inferred" to "confirmed" in the production version.

Pressure pumper identified on 77% of the last 12 months' jobs: from FracFocus supplier records directly, plus inference validated on the directly-labeled jobs: chemical fingerprinting (pumper-specific product trade names and traveling chem/sand vendors, ≥99% holdout precision), pad siblings (same pad ±45 days, 99.9%), fleet continuity, and dedicated-fleet patterns (operators ≥90% one crew; op-county ≥85%, 98.3%). Fleet-equivalents from each pumper's observed pad pace over the last 90 days of reported jobs. "Not disclosed" jobs are genuinely unidentifiable from public data — largely self-sourcing operators and ProPetro's current work (identifiable only on pre-merger Endeavor filings) — so shares are a floor, not a census.

Fleet-equivalents = settled wells/28 days (last 3 fully-reported months, so slow filers like ExxonMobil are counted at true pace — marked ·· ) × the operator's own observed pad pace (days per well, multi-well pads) ÷ 28. This is an activity index, not a crew census — heavy simul-frac operators turn pads faster than their pad-window suggests, so their true crew count runs lower. Implied drillout units assume ~6 well drillouts per coil/workover unit per month. Basin total on this basis: 398 wells/28d ≈ 71 fleet-equivalents among operators with 4+ settled wells/month. NM data shows a median 34 days from frac end to first production (p25 19 – p75 57) — the drillout / cleanout / turn-in-line window.

Modeled from each job's disclosed sand, water, water-source mix, and lateral. Basin total from the last 3 fully-reported months: ≈ $1.99B per 28 days (1246 settled jobs), incl. ≈ $188M proppant and ≈ $190M water. Operator rows use each operator's own most recent 90 reported days, so slow and dark filers (ExxonMobil, EOG) are counted at last-reported pace — stale data is flagged. Estimates, not invoices — see assumptions below.

Assumptions (2026 Permian benchmarks + field estimates): sand $39/ton delivered (field-verified 2026: 100 mesh ≈ $38.9, 40/70 regional ≈ $39.9); produced/recycled water $1.00/bbl all-in; fresh $0.32/bbl sourcing; water transfer $0.13/bbl; pumping $50k/stage at 200-ft stage spacing; wireline & plugs $10k/stage; chemicals $20/lateral-ft; drillout & flowback $300k/well. Operators that don't disclose proppant mass (e.g., EOG) have sand estimated at the basin-median 1.10 lbs/gal of disclosed water. Model back-tests at ≈$424/completed-ft on settled wells, consistent with operator-reported D&C economics. AFEs are private; these are modeled market estimates.

Wells frac'd July 2024 – Feb 2026, 7735 with a clean 90-day production window. New Mexico: true well-level oil volumes from OCD C-115 filings. Texas: oil + condensate from RRC lease-level filings, allocated evenly across each pad's new wells, with legacy-producing leases excluded. IP90 = average daily oil over the first three reported production months.

Three signals, all read month by month. The number is the average month across an operator's last three fully-reported months; the bars behind it are the fifteen months leading up to it, with those three picked out in colour; the comparison is that operator's own preceding twelve months. Permits filed with the Texas RRC are intent, and lead the frac by six to twelve months (through Jun). Spuds in New Mexico are commitment, and lead by one to three months (through Jun). Completions are the work itself, and stop at Mar because FracFocus takes about ninety days to fill a month in — the three columns end on different months on purpose. An operator permitting hard while completions hold flat is building inventory you get called for next year; one completing hard while permits fall is drawing down what it already has. Where a series is too thin or too broken to trust, the direction is left off rather than guessed. Hover any row for the last six months in figures.

No completions figure for EOG Resources — it has not disclosed a job inside the settled months. Late filing, not idle iron; see the filing-lag table below.

FracFocus has no deadline it enforces the way a state agency does, and operators treat it very differently. That matters here because a company that files ninety days late looks identical, in the table above, to one that stopped working — and telling a client the wrong one of those is the kind of mistake you only make once. So this is measured rather than assumed: the pipeline keeps dated snapshots, and a well that was absent last week and present this week was disclosed in between, which gives a real number of days from job end to disclosure. Where the sample is still too thin to stand on, the estimate comes instead from how long an operator's months take to fill in against its own average, and it is described as such. Across the 70 disclosures this pipeline has watched arrive, half landed within 19 days of the job ending, 66% within thirty days, and 10% took longer than ninety.

Ranked by total base water volume

Thirty-six pages — the basin, then every county, with each well drawn as its lateral and coloured by operator. It is four hundred kilobytes, so it loads on its own page: open the map book.