By Ovi
All of the Crude plus Condensate (C + C) production data, oil, for the US state charts comes from the EIAʼs Petroleum Supply monthly PSM which provides updated production information up to June 2025.
U.S. June oil production increased by 133 kb/d to 13,580 kb/d, a new high. The largest increases came from the GOM and New Mexico. July production is expected to decrease by 174 kb/d to 13,406 kb/d. Could June 2025 be the all time high before decline sets in?
The dark blue graph, taken from the August 2025 STEO, is the U.S. oil production forecast from July 2025 to December 2026. Output for December 2026 is expected to drop to 13,158 kb/d, revised down by 94 kb/d from last month. From July 2025 to December 2026 U.S. oil production is expected to drop by 248 kb/d.
The light blue graph is the STEO’s projection for output to December 2026 for the Onshore L48. June’s Onshore L48 production rose by 78 kb/d to 11,243 kb/d. From July 2025 to December 2026 production is expected to decrease by 313 kb/d to 10,835 kb/d. December 2026 production was revised down by 137 kb/d from last month.
U.S. Oil Production Ranked by State
Listed above are the 10 US states with the largest oil production along with the Gulf of Mexico.
These 10 states accounted for 83.1% of all U.S. oil production out of a total production of 13,580 kb/d in June 2025. On a MoM basis, June oil production in these 10 states rose by 69 kb/d. On a YoY basis, US production increased by 328 kb/d with the biggest contributors being New Mexico and the GOM.
State Oil Production Charts
June’s production increased by 11 kb/d to 5,723 kb/d. However May was revised lower from the previous report from 5,752 kb/d to 5,712 kb/d, a downward revision of 40 kb/d.
Texas production has rebounded since the weather related January 2025 drop. June’s production is essentially flat, small 11 kb/d increase, and still 109 kb/d lower than October 2024. The point to note here is both the production projection and the EIA production dropped after April.
The red graph is a production projection using the May and June Texas RRC data. Due to more production revisions and errors in the overall Texas data, the projection is a modification of the normally used methodology. Regardless the projection trend follows the EIA’s production up to May but not for June, a small drop vs a small rise. The projection is higher starting in February 2025 because of above normal MoM increases between the May 2025 and June 2025 EIA updates.
The blue graph shows the average number of weekly rigs reported for each month shifted forward by 10 months. So the 276 rigs operating in July 2023 have been shifted forward to May 2024. From February 2024 to July 2024, the rig count dropped from 312 in time shifted February 2024 to 256 in July 2024. That drop of 56 rigs had no impact on production up to October 2024 but November was the first month when the impact of the rig drop on oil production started to show up along with fewer completions. The rising production after January is difficult to explain. The small decline after April 2025 may be the first indication the drop in the rig count may be having an impact on oil production.
According to the EIA, New Mexico’s June production rose by 40 kb/d to 2,239 kb/d.
The blue graph is a production projection for Lea plus Eddy counties. These two counties account for close to 98.5% of New Mexico’s oil production. The difference between the May and June preliminary production data provided by the New Mexico Oil Conservation Division (ODD) was used to make the projection. A 1.5% correction was added to the Lea plus Eddy production projection to account for their approximate fraction of New Mexico’s oil production.
The projection estimates June production was 2,132 kb/d a decrease 7 kb/d from May. The decrease is related to the June production decrease in Eddy county and is discussed further down in the Permian section.
It is difficult to explain the 107 kb/d June difference between the EIA data and the NM OCD. The June New Mexico OCD update for Lea County had an 8 kb/d increase over May while Eddy had a 2 kb/d increase over May. All previous months had no changes. This implies the OCD June data is virtually fully up to date for those two counties.
Checking the EIA comp-stat-oil file Here provides a clue to explaining the difference. In January 2024, the ratio between the 914 Estimate and the 914 Survey Reported was 1.053 (See table below). By July 2024 it had risen to 1.084. The EIA has now started slowly dropping the ratio and for May it is 1.075. By my estimate the ratio should be closer to 1.048, i.e. closer to the January 2024 ratio of 1.053. Further the comp-stat-oil file also states the current NM State production for May is 2,145 kb/d vs my estimate of 2,139 kb/d.
Compare the Enverus data with the current EIA New Mexico data. In January 2025, the production gap between the EIA and Enverus is 43 kb/d. By May the gap increased to 64 kb/d. Why is the gap getting bigger? Also note that the Enverus and NM production are identical except for May.
Note: If the EIA’s NM June estimate is really overstated by 107 kb/d, US June production may really be closer to 13,473 kb/d.
More oil production information for a few Texas and New Mexico counties is reviewed in the special Permian section further down.
June’s output rose by 33 kb/d to 1,148 kb/d. Production is down 139 kb/d from the post pandemic peak of 1,287 kb/d.
The North Dakota Department of Mineral resources reported June production rose by 39 kb/d to 1,151 kb/d, which is essentially the same as the EIA’s estimate.
According to this Article “In May and June, certain operators had, because of the low price environment, curtailed some production in the state.” However production in August is expected to rise “as operators bring back online some production they had curtailed after oil prices sank earlier in the year, the state regulator said on Friday.”
“Given frac crews, completion numbers, summertime, I do expect that July will see an improved oil production level, and August has thus far been relatively consistent with those July numbers. So I do expect a couple of good months coming,” said Justin Kringstad, director of North Dakota Pipeline Authority.
There are currently 29 active rigs in the state of North Dakota, steady compared with July, the state’s regulator said. Meanwhile the frac crew count is currently at 14, up from 13 in July.”
Alaskaʼs June output dropped by 12 kb/d to 422 kb/d while YoY production increased by 21 kb/d. The weekly EIA report indicates that July and August production will drop close to the 350 kb/d level.
Alaska has recently brought new fields online to consistently have flat YoY and monthly production gains which have broken away from the earlier dropping production red trend lines.
According to this Article the Alaska Picca field is expected to come online in early 2026.
“Australian independent Santos’ 80,000 b/d Pikka oil field in Alaska’s North Slope region is now 91pc complete and will come on line in January-March 2026, the company said, ahead of the initial January-June target.”
At its peak, Pikka could produce up to 80 Mb/d from 45 wells. The production expected from these projects may be enough to flip Alaska oil production from a steady decline to a growth period.
Coloradoʼs June oil production dropped by 5 kb/d to 459 kb/d.
The biggest oil producing county in Colorado is Weld County and its production has been added to the chart. The two graphs are almost parallel since January 2024. Weld’s production dropped by 26 kb/d in June, 21 kb/d more than the whole state.
Colorado began 2025 with 7 rigs in January and February and then dropped to 5 in March/April/May. In August the average weekly rig count rose to 8.
It should be noted that Colorado drillers are facing increasingly stricter environmental rules according to this Article which could be reducing drilling locations.
Oklahoma’s output in June dropped by 3 kb/d to 403 kb/d. Production remains below the post pandemic July 2020 high of 491 kb/d and is down by 49 kb/d since May 2023. Output entered a slow declining phase in June 2023 and now appears to be range bound around 400 kb/d ± 20 kb/d.
In May Oklahoma had 51 operational rigs. However by July the number had dropped to 41. The dropping trend changed in August by adding 1 to 42.
California’s overall declining production trend continues. June’s production dropped by 3 kb/d to 259 kb/d, a new low. YoY production dropped by 44 kb/d.
Wyoming’s oil production reached a post pandemic high in February 2024 and appears to have entered a plateau phase around 295 kb/d. June’s production rose by 10 kb/d to 300 kb/d.
At the beginning of 2025 Wyoming had 14 operational rigs and they rose to 16 in May and June. The rig count In July dropped to 10 and in August dropped further to 8.
June’s production increased by 1 kb/d to 194 kb/d to a new high. Utah had 8 rigs operating from October 2024 through May 2025 but dropped to 6 in early June. In August 7 rigs were operational.
Ohio’s June oil production decreased by 3 kb/d to 139 kb/d. The most recent Baker Hughes rig report shows no oil rigs operating in Ohio. They have all been re-classified to NG rigs. There were 6 operating in January and rose to 10 in July and were down to 9 at the end of August.
GOM production rose by 67 kb/d in June to 1,915 kb/d. July’s production is projected to decrease to 1,896 kb/d.
The August 2025 STEO projection for the GOM output has been added to this chart. It projects production in December 2026 will be 37 kb/d lower than July 2025 at 1,859 kb/d.
A Different Perspective on US Oil Production
Combined oil output for the Big Two states Texas and New Mexico.
June’s production in the Big Two states increased by a combined 51 kb/d to 7,962 kb/d, a new high, and is 18 kb/d higher than October 2024. Clearly these two states were the drivers of US oil production growth up to October 2024. The rising trend has definitely slowed since October 2024.
Oil Production by The Rest
June’s oil production by The Rest rose by 27 kb/d to 3,281 kb/d and is 164 kb/d lower than November 2023.
Permian Basin Report for Main Counties and Districts
This special monthly Permian section was added to the US report because of a range of views on whether Permian production will continue to grow or will peak over the next year or two. The issue was brought into focus many months back by two Goehring and Rozencwajg Reports and Report2 which indicated that a few of the biggest Permian oil producing counties were close to peaking or past peak.
A more recent report was issued and can be reviewed Here. In this report they state:
“For years now, we have outlined with what we hoped was clarity, and what we now submit was prescience, the view that U.S. shale oil, that great source of modern supply, could not grow forever. It would mature, crest, and begin its long descent. That moment, by our models and measures, has arrived: shale has plateaued, and 2024 appears to be its high-water mark. And yet, investor sentiment has scarcely been more downbeat.”
This section will focus on the four largest oil producing counties in the Permian, Lea, Eddy, Midland and Martin. It will track the oil and natural gas production and the associated Gas Oil Ratio (GOR) on a monthly basis. The data is taken from the state’s government agencies for Texas and New Mexico. Typically the data for the latest two or three months is not complete and is revised upward as companies submit their updated information. Note the natural gas production shown in the charts that is used to calculate the GOR is the gas coming from both the gas and oil wells.
Of particular interest will be the charts which plot oil production vs GOR for a county to see if a particular characteristic develops that indicates the field is close to entering or in the bubble point phase. While the GOR metric is best suited for characterizing individual wells, counties with closely spaced horizontal wells may display a behaviour similar to individual wells due to pressure cross talking . For further information on the bubble point and GOR, there are a few good thoughts on the intricacies of the GOR in an earlier POB comment and here. Also check this EIA topic on GOR.
New Mexico Permian
The rig counts in Lea and Eddy county have started to move in opposite directions. However the overall total rig count in Lea and Eddy counties has remained close to 80 since June. In the week ending August 29, the rig count was unchanged at 80 and down 14 from the January 2025 count of 94.
Eddy county rigs stabilized close to 45 in May but then began a slow drop and hit the current and recent low of 32 at the end of August. Lea county, starting in late June, has added 12 rigs to 48 in late August. What is driving the increased drilling activity in Lea County?
Lea County’s oil production entered a plateau phase in May 2024 at 1,203 kb/d and the plateau continues to June 2025. June’s projected output rose by 2 kb/d to 1,200 kb/d. Preliminary June data from New Mexico’s Oil Conservation Division (OCD) indicates Lea County’s oil production dropped by 7 kb/d to 1,191 kb/d. There is virtually no gap between June’s preliminary data and the projected data, except for June, which indicates that the OCD data is essentially fully reported.
Production has been essentially flat since May 2024 as the rig count fell and rose. June’s projected production is another month in which the overall dropping rig count has had no impact on oil production.
The rising rig count starting in October 2024 may be a contributing factor to maintaining the plateau phase out to June 2025. The Lea rig count for real August 2025 is 47, roughly 3 rigs lower than the count for time shifted January 2025 to June 2025. Will Lea county production drop in July as the time shifted rigs drops again in July and August or continue on its plateau?
The blue graph shows the average number of weekly rigs operating during a given month as taken from the weekly rig data. The rig graph has been shifted forward by 8 months. So the 64 Rigs/wk operating in August 2023 have been time shifted forward to April 2024 to show the possible correlation and time delay between rig count, completion and oil production.
Note that rig counts are being used to project production as opposed to completions because very few extra DUCs are being completed at this time.
After much zigging and zagging, oil production in Lea county stabilized just below 1,100 kb/d in early 2023. Once production reached a new high in January 2023, production appeared to be on a plateau while the GOR started to increase rapidly to the right and first entered the bubble point phase in July 2023.
Since July 2023 Lea County’s production continued to increase as the GOR remained within a second semi-bounded region. This may indicate that additional production was coming from a new field/zone since the GOR’s behaviour since August 2023 to March 2024 time frame appears once again to be in a second semi bounded GOR phase accompanied with rising production.
The GOR moved out of the second semi-bounded GOR region in April 2024 and production hit a new high of 1,203 kb/d in May 2024. Since July 2024 the GOR was range bound between 3.35 and 3.45 but June’s GOR hit a new high of 3.52 while preliminary production dropped by 7 kb/d to 1,191 kb/d.
This zigging and zagging GOR pattern within a semi-bounded GOR while oil production increases to some stable level and then moves out to a higher GOR to the right has shown up in a number of counties. See a few additional cases below. The rising GOR in Lea county is another indicator that production is close to entering its decline phase.
June’s projected oil production decreased by 8 kb/d to 900 kb/d. Also preliminary production from the NM OCD decreased by 10 kb/d to 898 kb/d. Eddy county’s month over month production updates are very few and small and primarily occur in the last two months which indicates their preliminary production is very close to final. This is indicated by the red graph covering the green graph, i.e, there is little separation between the two graphs except for the last month.
Eddy County’s recent oil production rise and fall is related to the rise and fall in the rig count. From May 2024 to November 2024, production rose from 757 kb/d to 906 kb/d, an increase of 149 kb/d, while essentially paralleling the increasing rig count. Over that same time shifted rig period, 14 to 15 rigs were added to Eddy County as production rose. Was a new Tier 1 region/zone discovered to attract such a large increase in the rig count?
The blue graph shows the average number of weekly rigs operating during a given month as taken from the above weekly drilling chart. The rig graph has been shifted forward by 8 months to roughly coincide with the increase in the production graph starting in November 2023.
Clearly the production rise up to November 2024 is closely associated with the rise in the rig count and associated well completions delayed by roughly eight months. The rising production starting in February 2025 does not correlate with the dropping rig count. However that rising projected production trend reversed itself in April as it fell by 47 kb/d to 900 kb/d in June, paralleling the dropping rig count.
The Eddy county GOR pattern is similar to Lea county except that Eddy broke out from the first semi bounded range earlier and then added a second semi bounded GOR phase. For June New Mexico’s Oil Conservation Division (OCD) reported oil production decreased by 10 kb/d to 898 kb/d and the GOR increased slightly to 5.52.
Texas Permian
The rig count in both Midland and Martin counties started to drop in early August.
The Midland county rig count dropped to 17 rigs at the end of July and then added 6 to 24 in the first week of August before dropping back to 21 at the end.
Martin county’s rig count has been slowly dropping since March 7 high of 29 rigs. July and August saw continuing drops. August rigs dropped to a new recent low of 17.
Oil Production in Texas Counties
Comparison Chart: Midland County’s oil production chart from the previous report.
June’s projected production rose by 14 kb/d to 629 kb/d. I think the Tx RRC’s initial June production data looks reasonable and so is the projection. As noted in the previous post, the large revisions to April’s production in the May report turned Midland into a county with rising production, see comparison chart above. Note how the difference in the production trends of the two red graphs starts in January 2025.
The orange and green graphs show the oil production for Midland County as reported by the Texas RRC for May and June. The red graph uses the May and June data to project production as it would look after being updated over many months.
The blue graph shows the average number of weekly rigs operating during a given month as taken from the weekly drilling chart. The rig graph has been shifted forward by 12 months to better align with production. So the average 34.5 Rigs/wk operating in July 2023 have been moved forward to July 2024 to show the possible correlation and time delay between rig count, completions and oil production.
The 12 month rig time shift is much larger than the typical six to eight months used in other counties. It is not clear why there should be such a difference. If the twelve month shift in the rig count is approximately correct in that oil production can be tied to the rig count, oil production in Midland county should continue falling up to July 2025.
For June the Midland GOR ratio reversed direction and fell to 4.36 from 4.5 in May while reported preliminary oil production dropped by 8 kb/d to 577 kb/d.
With Midland county into the bubble point phase, oil production and the GOR stayed within a narrow range outside of the initial Semi-Bounded GOR region from March 2024 to February 2025. However the April, May and June 2025 GORs have broken out to new highs.
The oil production and GOR data shown in this chart are based on the RRC’s June production report. Note that while the last few months are subject to revisions, the January 2024 to December 2024 production data has been steady for a number of months.
Martin county’s projected June oil production dropped by 24 kb/d to 662 kb/d. The June preliminary production data along with the MoM updates look normal and the Martin projected production estimates are reasonable. Martin county’s production appears to be in a peaking type of plateau since May 2024 and appears to be range bound between 650 kb/d to 700 kb/d. The projected production increase starting January 2025 is real but then peaks in April before starting a new declining trend.
The red graph is a projection for oil production as it would look after being updated over many months. This projection is based on a methodology that uses preliminary May and June production data. The green graph shows the updated oil production reported by the Texas RRC for June and June itself is 34 kb/d lower than May’s production. Production since April may have begun to track the time shifted rig chart. The current real August average rig count is 18.5, 9 lower than time shifted June 2025. This could be an early signal that the projected production drop that started in May will continue.
The orange and green graphs show the production for Martin County as reported by the Texas RRC for May and June. The blue rig graph shifts the rig count ahead by 6 months.
Martin county’s oil production after November 2022 increased and at the same time drifted to slightly higher GORs within the semi bounded range. However the June 2024 GOR saw its first move out of the semi bounded region. The RRC’s preliminary June 2025 production for Martin County shows a decrease in production of 34 kb/d accompanied by an increase in the GOR to 3.07, a new high.
Martin county has the lowest semi-bounded GOR boundary of the four counties at a GOR of close to 2.60. The GOR is now clearly out of the semi-bounded region. Martin County has now entered the bubble point phase that should result in oil production possibly entering a slowly declining phase.
This chart shows the total oil production from the four largest Permian counties. Assuming current Permian production is close to 6,400 kb/d, these four counties account for 53% of the total. June’s projected production decreased by 16 kb/d to 3,386 kb/d and is the third month showing declining production.
The May and June initial production data are shown in the orange and green graphs respectively. The red graph uses the May and June data to project a more realistic estimate for June production.
Findings
– The preliminary June production data for New Mexico was good. Texas county data was generally good.
– The production charts for the four largest Permian County’s appear to be in different phases of their production life. Of the four, one NM Permian county continues in its plateau phase while the other one has entered its declining phase. The two Texas counties may have also entered their declining phase. Taking into consideration the price for a barrel WTI is stuck in the low $60s, the rig and frac spread counts continue to make new recent lows, when taken all together these considerations all point to peak production occurring in the onshore lower 48 within the next three to four months.
– Lea county entered its plateau phase in May 2024. While oil production is not following the rig count graph directly, the dropping rig count is resulting in Lea production currently being in a steady flat plateau phase.
– Eddy county’s production hit a new high in March 2025 but had a big drop in April while June saw a small decrease, possibly signalling the beginning of a plateau phase at a lower production level.
– Midland county’s production has been decreasing since November 2024. The increase in June production may be associated with an increase in the rig count. The addition of six new rigs to Midland county in August is an unexpected surprise and makes one wonder what it implies going forward.
– Martin county’s projected production has been affected by revisions to previous months. However I think the increasing production up to April is real along with the decline since then.
Texas District 8
The District 8 chart has been updated using the production data from January 2024 to June 2025 and is more realistic than reported in the previous post. In the previous post only production data after January 2025 was used because of errors in the data. The projected June production data indicates that production may have peaked in April 2025 at close to 4,015 kb/d. Since then production has dropped by 111 kb/d to 3,904 kb/d.
While revisions in the production chart affect the projection, it does not affect the GOR.
Plotting an oil production vs GOR graph for a district may be a bit of a stretch. Regardless here it is and it seems to indicate many District 8 counties may well be into the bubble point phase even though the GOR decreased in June to 4.49 from 4.5 in May.
Oil Production and GOR Charts for a number of Larger Texas Oil Producing Counties
Below are charts for six top oil producing counties in Texas. While the Texas June data is better than the May data, the projections are affected since they depend on the production difference between the two months and the time lag in data updating. As a result, questions can be raised for a few of the projections.
June’s projected oil production for Reeves county dropped by 48 kb/d to 473 kb/d. The production projection is reasonably close up to February 2025 but slightly optimistic after that.
Reeves county peaked in May 2024 at 525 kb/d and production dropped up to January 2025. The GOR chart indicates Reeves County initially entered the bubble point phase in December 2024 and then reversed back into the Semi-Bounded region. June’s GOR is 7.07 and at a record high.
Reeves county GOR is high because it is the number one Texas county ranked by gas production. The current C + C production is equally split between crude and condensate.
Loving’s projected production dropped by 2 kb/d to 478 kb/d in June. Loving production peaked at 499 kb/d in November 2024.
For June, the GOR increased to 4.28, a new high.
While Loving had 19 operational rigs in real June, they jumped by 2 to 21 in real time August. Loving has had roughly 20 rigs operating all year.
Howard county production peaked in July 2023 at 423 kb/d. The production projection fell to a new low of 260 kb/d in June 2025, which is the same as projected in the previous report. The general production trend appears to be following the dropping rig count.
Note the rapid movement of the GOR to higher ratios once it broke out of the Semi-Bounded GOR range.
The rig chart has been shifted forward by five months. Howard county had 10 rigs operating at the beginning of the year, July 2025 on the first chart. In the current real August there are 2 rigs operating.
Reagan county oil wells have a very high GOR and passed peak production in September 2024.
June’s projected production decreased by 8 kb/d to 179 kb/d. The GOR is still in the semi bounded region because the GOR is so high.
Glasscock’s production has been falling since April 2024. Increased drilling in February and March could account for some of the increased projected production reported in June. The GOR moved out of a very wide Semi-Bounded region recently. Note the rapid increase in the GOR starting from 3.90 in April 2024.
June’s projected production decreased by 2 kb/d to 166 kb/d while the GOR made a new high of 5.9.
Ward’s projected oil production has been in a slow decline since December 2023 while the GOR has been slowly increasing within the semi-bounded region. June’s GOR hit a new high of 3.53 outside the semi-bounded region. The projected June production drop to 157 kb/d is more realistic in terms of direction, i.e. a drop. The actual production 157 kb/d is questionable. The next report should clarify what is happening in Ward county.
The rig graph has been time shifted forward by 6 months. The drop in the rig count from 16 in time shifted January 2024 to 8 in January 2025 has had little impact on the projected oil output. However the large drop in the preliminary and projected June oil production may be the first indication the dropping rig count is beginning to have an impact on oil production.
Ward County began the year with 10 rigs. The rig count dropped to 3 in August.
Drilling Productivity Report
The oil production for the 5 DPR regions tracked by the EIA’s STEO is shown above up to July 2025. Also the August 2025 STEO projects production out to December 2026, red markers. Note DPR production includes both LTO oil and oil from conventional wells.
July’s oil output in the five DPR regions increased by 25 kb/d to 9,081 kb/d and was revised down by 46 kb/d from last month’s report. Production is expected to rise by 3 kb/d in August to 9,084 kb/d. Production rises slightly out to December and then begins to decline.
Production in December 2026 is forecast to be 8,881 kb/d a downward revision of 58 kb/d from last month.
The EIA’s August STEO/DPR report shows Permian July output rose by 11 kb/d to 6,569 kb/d. From July 2025 to December 2026 output is expected to drop by 155 kb/d to 6,414 kb/d. Note that December 2026 production has been revised down by 66 kb/d from 6,480 kb/d to 6,414 kb/d.
Production from new wells and legacy decline, right scale, have been added to this chart to show the difference between new production and legacy decline.
July output in the Eagle Ford basin increased by 3 kb/d to 1,121 kb/d. August 2025 production is forecast to drop by 1 kb/d to 1,120 kb/d.
Output in December 2026 expected to be 1,099 kb/d, a decrease of 41 kb/d from the July 2025 STEO report of 1,140 kb/d.
The DPR/STEO reported that Bakken July output rose by 11 kb/d to 1,158 kb/d. August production is expected to increase by 10 kb/d to 1,168 kb/d. The STEO/DPR projection, red markers, shows output rising up to December before dropping to 1,167 kb/d in December 2026.
This chart plots the combined production from the three main LTO regions. For July output increased by 25 kb/d to 8,848 kb/d. Production for December 2026 is forecast to be 8,681 kb/d, a downward revision of 94 kb/d from the previous report.
DUCs and Drilled Wells
The number of DUCs available for completion in the Permian and the three major DPR regions has returned to a dropping trend. July’s DUC count for the three basins dropped by 24 to 1,560. In the Permian the DUC count dropped by 14 to 980.
In the three primary regions, a total of 619 wells were completed in July, unchanged from June. There were 595 wells drilled in July, down 11 from June. For comparison, In January 2023, 688 wells were drilled.
In the Permian, the monthly completion and drilling rates have begun to drop.
In July 2025, 445 wells were completed and 431 new wells were drilled. This is the fifth month in a row in which the number wells drilled has dropped.
121 responses to “June US Oil Production New High”
Hi Denise . Your rework of the website is giving me a heartburn . Formerly I could read the new comments with the ” new ” tag so I did not have to scroll thru the comments that I had read in my earlier visits . Now do I have to scroll thru the whole page to see the latest comments ? I am not very computer savvy ( difficult to teach an old dog new tricks) . Assistance requested on the tricks .
P.S ; You and Ovi do a super duper job and thanks to Ron . Amazing he started this in 2013 taking over from TOD .
The “feature” of these blog apps is that they use plugins to do many of the convenience functions, such as a NEW comment highlighter. But if the main blog software gets updated while the plugin hasn’t been, it may not work together any longer. That’s why many blog sites tend to not want to update.
Would like to thank neuroatypical IT dweebs for QCing the software.
Do IT dweebs exist outside the corporate environment?
Use a Ctrl-F on the date to find comments that were left after your last visit. If you were here on the 6th, search for/06/ and later dates.
Ditto what HIH posted.
I realize, Dennis, that you have willingly taken on the challenging task of managing this site.
Hopefully you will be able to overcome the shortfalls in the new set up.
As always, much appreciation for the efforts of Ron, Ovi, and yourself.
Sorry everyone, doing the best I can. Often I have used free plugins to add features. I am not very adept at programming so when the old plugins no longer work, I cannot reproduce what I had before.
Maybe Paul Pukite can help me out he’s pretty good at this stuff. I am also trying to use Chat GPT.
Dennis –
Try the ‘Reeder’ app, add two feeds (main site and then comments feed)
Then for the comments feed, change settings to open in browser…new comments appear at the top of the feed…
Here’s what it shows:
IMG_4020
You’re doing fine. Don’t let the complainers get you down.
P.s. Don’t think I’m getting soft. I still reserve the right to squabble with you on peak oil versus horn-filled cornucopia. But I don’t like you getting down from website maintenance crap. Chin up! 😉
Although somewhat reluctant to post this …
If a financial outlay may be required/optimal in order to effectively put forth a ‘super duper’ format, perhaps your audience might be able to help in these efforts.
Might be worth considering.
I will get it sorted in time.
An article by our favorite (sarc) oil cornucopian Michael Lynch
https://www.forbes.com/sites/michaellynch/2025/09/04/shale-oil-pessimism-could-be-overdone/
Paul
I remember Michael Lynch countering the claims made by you and others that global oil production had peaked. That was 20 years ago!
People like yourself claimed you knew how much oil there was on the planet, you have not got a clue. There are more wells drilled in The U.S. than the entire world combined.
He was correct and you were wrong by a country mile.
Iver said:
20 years ago I had just started blogging. So please don’t lump me in with “others”. Yet even then, I was keen on developing something better than symmetric Hubbert curve analysis derived from the logistic sigmoid. The replacement was the oil shock mode,l which allowed for asymmetric long tails in the decline curve,
Sure enough, it was about the year 2004 that the world start on a somewhat lengthy plateau of conventional crude oil production that lasted for around a dozen years. Only the flash-in-the-pan introduction of unconventional light tight oil from fracking extended the crude oil peak to 2018.
globalcrudeoil
Paul
I know your work, it presumes a certain amount of oil in the ground. Which nobody actually knows.
Global conventional oil has not peaked, what has happened is U.S. shale oil has muscled in forcing conventional oil production to be cut back. Look at the international drilling figures! The amount of drilling is half what it was at $100.
We will only know what can be produced when prices are over $100 for several years. At the moment we are being flooded with oil, hence the price.
Iver,
The number that matters is how much oil is economically recoverable, and you are correct that nobody knows what that is (including Michael Lynch). Nobody knows what the future price of oil will be, you assume it will be over $100/b (let’s assume you mean in 2025 US$), we don’t know if that will be the case, right now that looks like a poor assumption. What about $200/b or $500/b do those prices (in 2025 US$) sound plausible to you? Will there be much demand for oil at those prices? My best guess is a URR of about 3000 Gb plus or minus 20%, at $500/bo in 2025$ maybe 3500 Gb or more, but I am highly skeptical of arguments that costs to extract oil will decrease significantly (more than 20%) or that prices will increase significantly (more than double). We will see.
Note that my estimate uses much of the foundation built by Paul Pukite.
Dennis,
In this context it helps to distinguish between peak demand and peak supply. Peak supply is the thing Peak Oil enthusiasts are concerned about. Peak supply can only be diagnosed when prices rise quite sharply and supply does not respond.
IIRC 2018 is peak demand…so far. It appears that we have not seen peak supply.
Dennis, the way to think about supply and demand is to assume a shift from one causes X price change and then what is the effect on the other.
So…if you are a producer (or an analyst analyzing supply, MMbopd), then it’s sort of irrelevent to you how we got to $200 (or $500). You just model how much supply comes on. And the answer is a lot immediately and probably a lot ultimately…in some ways it is a resource pyramid, with more crappy supply than good supply.
Of course if there is a supply shock (OPEC or “running out of shale” or the like that drives price up, you can then discuss how that reduces demand. Or even more realistically how a reduction in volume (MM bopd) affects price. Obviously some sectors roll off first. People without a lot of money joyriding stops before intercity trucking for instance.
IOW, you don’t need to wonder what happens to demand in your scenario if you are just modeling price as the externality (e.g. because of rest of world supply slacking or OPEC cuts). You can just say the price is X and how does that drive shale supply response.
At the end of the day, demand is somewhat inelastic. But that also means it is somewhat ELASTIC. We definitely saw reduction in retail gasoline and jet fuel demand when prices doubled a couple years ago. It’s not like half the demand drops. But some of it does for sure.
For a more tactical and near term example, Artem from Rystad (who has spent most of his life working in these analyses, getting paid for it, and is highly numerate) says a sustained move to $90 would bring rigs back in force and drive ~1MM bopd/growth rate. (And yes, on top of the 13.5 we are already doing now!)
Dennis
I did not assume anything. I said $100 plus oil would be needed for several years in order to develop all potential sources.
If you knew anything about the oil industry then you would not throw out silly numbers like $500. Nobody in the oil industry says $500 is needed to extract the most difficult oil.
A price of $100 to $120 would make the vast majority of global oil reserves economically viable.
At the moment there is so much oil that we don’t need $100 oil.
Iver,
I agree that currently we have an oversupply of oil which is driving oil prices down. My expectation is that once OPEC plus reaches its capacity output will no longer increase by much, especially at $65/bo or less in 2025 US$. Eventually demand may catch up with supply, at which point the price of oil may rise.
The price at which supply of oil and demand for oil will be roughly in balance in 2 or 3 years time (when OPEC plus may have reached capacity limits) is unknown. My guess is $80/bo short term and perhaps as high as $160/bo longer term (annual average price in 2025 US$). That is simply a WAG, nobody can predict this.
I also agree that most proved (1P) reserves (as estimated by Rystad) are likely to be produced at $100 to $120/bo in 2025 US$.
Rystad’s estimate for global 1P reserves is 435 G and about 30 Gb was consumed in 2024 (so about 14.5 years of oil at 2024 rates of consumption.)
Note that the 2PCX estimate of global oil resources (including yet to find oil) is about 1519 Gb according to Rystad, enough for 50 years at 2024 levels of consumption (URR would be 3091 Gb if all of the Rystad 2PCX resource estimate were produced).
https://www.rystadenergy.com/news/discovered-recoverable-oil-resources-increased-by-5-billion-barrels-despite-produ
To the above hint at production decline in the shale basins, get ready for a massive CO2 flood.
The 45Q section of the OBBBA pays EOR on par with CO2 sequestration @ $85/ton.
But the big money may come from DAC (direct air capture) @ $185/ton.
I mean, this is like being paid for catching moonbeams.
This really sucks.
Hopefully a solution can be reached.
I really have liked this site.
May have found a work around,
Download App:
‘Reeder’
Add the Feeds
‘Peak oil barrel’
‘Comments for Peak Oil Barrel’
I find that for now, turning phone 90 degrees works. I am on holiday for a couple of days. Will continue to work on it next week.
Hightrekker,
I assume you are using a smartphone?
If not, can you be more specific?
I have tried to improve the smartphone experience. Check it out if that was the problem, I think it is a bit better.
Yergin and Lynch are simply wrong. Crude-and-condensate are declining
https://www.resilience.org/stories/2009-09-03/michael-lynch-daniel-yergin-denizens-peak-oil-denial/
Peak Oil 2006 or 2008 advocates today claim they were wrong due to the unforeseen technological innovations surrounding tight oil. Fact is they have no idea how much oil could have been found or produced around the world if oil prices of over $100 were sustained for the past 20 years.
Also claiming Peak Oil when several of the most important oil producers are under severe sanctions is a cheap victory.
Peak Oil was and is, a global irreversible production decline regardless of price and in a totally open market.
All this has become a little irrelevant. People returning from visits to China are reporting that the Chinese people are sure they will take Taiwan by force. Within the next two years. The computer processing industry in Taiwan is so important that the U.S . Would have to protect Taiwan.
That article by Randy Udall of ASPO-USA is a hoot. It concludes with:
“…we wager $10,000 that all-liquids production won’t exceed 92 million barrels a day by 2020. This time, we hope Yergin and Lynch will put some money where their mouth is.”
I sorta doubt a wager was every finalized. And Udall died in 2013, well before the 2020 wager end.
But just looking at the data on what happened. We were at 93.9 MM bpd in 2020. I.e. well past his prediction.
https://www.eia.gov/international/data/world/petroleum-and-other-liquids/annual-petroleum-and-other-liquids-production?pd=5&p=000gfs0000000000000000000000000000vg&u=0&f=A&v=line&a=-&i=none&vo=value&vb=170&t=C&g=none&l=249–249&s=94694400000&e=1704067200000&ev=true
And that was with the pandemic. We were up to 100.4 MM bpd in 2019, the year before the pandemic. And are up to 103.1 as of 2024 (last full year).
In fact, we crossed the 92+ boundary, some time close to Udall’s death. 2013 averaged 91.4 and 2014 average 93.9.
“all-liquids”
Paul, that was Randy Udall’s metric. I was direct quoting him.
That wasn’t me. Wasn’t Lynch. Wasn’t Yergin.
It was the yardstick the ASPO writer (Randy Udall, son of Mo) specified. And on that yardstick, his prediction was wrong. Production exceeded what he thought it would.
Remember, Paul, that article came from 2009. Back then, ASPO thought we were about to have peak GAS as well as peak oil. (David Hughes did an hour long speech at ASPO conference saying that North America wouldn’t be able to build IMPORT (regas) terminals fast enough to meet demand!
The NGL rise has been insane…even happening through dramatic price crashes. Dennis had a wonderful analysis showing how not only has gas grown but the “wetness” (NGL content) has grown as well. It’s been an amazing double whammy. The US went from being the world’s largest propane importer to being the world’s largest propane exporter!
Oh…and yes…I know that propane serves non-transport markets and has a lower value than crude (or the major refined products from it). I took organic chemistry. I don’t know hydrocarbons as well as you know 3-5 semiconductors. But I know alkanes pretty well (kinda zoned out later in the course, but saturated hydrocarbon are nice and simple.)
Actually it was 2006 and a half hour David Hughes speech (senior moment, mean culpa).
https://www.youtube.com/watch?v=poRAEL7M9Ds
But the gist of my point remains. Uh…mazing change from the worries about Peak Gas to the “how do we get rid of it”.
The chances of a steep recession (or worse) in the next few years is increasing significantly regardless of what fed and treasury may try to do. A large swath of the population is going to be left behind. Increasingly so.
If so, the price of oil (in real dollar terms) will be sub $50 for an extended time, short of a shooting war.
Hey Hickory,
What makes you think we are headed towards a “steep recession (or worse)”. I agree, just looking for your logic.
Thanks
Some factors are at risk of escalating.
-Job displacement by AI will hit many vibrant economic zones pretty hard over the next couple years. It seems to be starting now. Unemployment could become very ugly.
-A real estate problem is brewing, with people wanting to downsize or relocate but not being able to find buyers. Cost of home insurance and maintenance is making it so that many people are having more trouble carrying the costs.
-“With trillions in commercial real estate (CRE) loans set to mature in the coming years, many borrowers are facing significant refinancing challenges due to higher interest rates and tightened credit conditions. Higher rates are making it more expensive to refinance, which could lead to a wave of distressed sales or foreclosures if loans cannot be paid off.”
-High costs for basics like electricity, food, medical care are becoming harder and harder to bear for the lower 3/4 of the income pie. Inflation has not been curtailed, or prices rolled back for most things.
-Helter skelter government policies have put a damper on business planning, and are raising costs. Tariffs are an additional tax that will get paid by Americans one way or another. Labor costs for all businesses are going up and up.
– There is a tide of increasing pessimism in many circles regarding the US standing in the world, which has taken a severe beating over the past 9 months. Even among our closest allies and neighbors.
– the stock market is likely far overextended, and many people will pull in any discretionary spending when it goes into the next contraction mode.
– purchasing power of most people is likely to continue to contract, maybe briskly as the federal government acts to drop the dollar value ( as a way to do a soft default on on debt payments)
These are some of the ones that have got my attention
Thanks Hickory.
I am no expert in economics, and observation suggests that supposed experts have poor track records too.
I don’t have enough familiarity with AI to understand how that is going to displace as many as some suggest, but it does seem to be impacting employment now.
S&P global notes that “the typical causes of recessions include policy errors, widespread financial stress, demand and/or supply shocks, military conflicts, and pandemics.”
US tariffs certainly seem to fit the bill of a policy error, at least in the short term. Also, I see that no less than AEI suggest US immigration policy will cut another 0.3-0.4 percentage points in US GDP growth. I would imagine that associated impacts on remittances will also hit US trading partners hard. Add in the impacts of OBBB and US energy actions to the list of policy mistakes.
Coupled with the continued growth in EV market share globally, it does look like headwinds for oil demand growth. Perhaps not so much natural gas?
ConocoPhillips is doing just great:
https://www.newsweek.com/texas-oil-giant-lay-off-quarter-its-staff-2124985
Rgds
WP
“Could June 2025 be the all time high before decline sets in?”
1. It could be. Who knows. Clearly, at these prices, growth has slowed and decline is looming. Still…if it is JUN or slightly later is hard to tell. For one thing, some of the peak lovers probably expected it even earlier and are unhappy with a new record.
2. I’m sorta sticking with my OCT/NOVish prediction. Just think we are sort of on a plateau and heading for decline soon. But there is some seasonality, especially from ND. So, it’s normal to have increases in summer…and setbacks in winter. That’s why I figured OCTish for a peak. But it’s a guess. And hurricanes for one, could throw it all out of whack.
3. I’d be a little wary of the logical implication you were making from the STEO. That is older than the 914 and based on data/models from a month (or more) ago. So, if we had a recent increase, it’s not crazy that STEO isn’t showing it yet.
Nony
Thanks for taking the time to respond and for that close look at NM.
The Title was rhetorical? I think we need a few more months of production to see what is happening in Lea. Totally flat and adding more rigs. Is the drilling enough to offset decline?
The next STEO comes out Wednesday. Check it out.
“It is difficult to explain the 107 kb/d June difference between the EIA data and the NM OCD.”
I’d be wary of this sort of thinking. There is a long, longitty, long-long history of people on this site being unhappy with higher EIA estimates. And in general, their concerns have been invalid. NM has much of the same issues as TX in terms of straggler reporting, necessitating the 914 survey in the first place. Efforts to just take state data (even knowing it straggles, and putting a multiplier on most recent month) have proved way worse than the 914 method, which after all does a direct survey, of most production.
Nony
“NM has much of the same issues as TX in terms of straggler reporting, necessitating the 914 survey in the first place.”
I totally disagree with that statement. Tx has many more reports to handle than NM and NM production is more up to date.
Attached is a table that compares the May and June reports for Eddy, Lea and Midland counties. The point is to compare the month over month changes for the last to or three month for the NM counties and Midland.
For Eddy, the May adjustment was only 2 kb/d. For Lea the May adjustment was 8 kb/d. For Midland the May adjustment was 22 kb/d and the adjustments go back to January. The NM adjustments are smaller than Midland both numerically and percentage wise.
As noted in the post the only difference between Enverus and NM was in May and my estimate was very close for May. In the posted table check out how wrong the 914 estimates were for the final production from January 2024 to December 2024.
Compare June 2024. 914: 2,010 kb/d — Final 1,975 kb/d. — Enverus/NM 1,975 kb/d.
Table
Click on Table to see it.
“In January 2024, the ratio between the 914 Estimate and the 914 Survey Reported was 1.053 (See table below). By July 2024 it had risen to 1.084. The EIA has now started slowly dropping the ratio and for May it is 1.075. By my estimate the ratio should be closer to 1.048, i.e. closer to the January 2024 ratio of 1.053.”
I like the analytical effort here.
But. (You knew there would be a “but”.) There is no reason to expect the multiplier to be static. They follow a process (explained in the 2018 methadology pdf) for how they develop this multiplier. If it changes, then that is what the statistics is driving. If anything, the issue is that they can’t rapidly compensate for radical changes in the multiplier, which is one reason, why you see the gradual changes going on.
It’s just as possible that they are underestimating as overestimating. But again, they are following a rational, mathematical method. And this is what the process outputs.
See https://www.eia.gov/petroleum/production/pdf/eia914methodology.pdf (especially pages 4 through end).
Nony
I think the info I posted above tries to answer this comment also. That ratio was totally wrong from January 2024 to December 2024 and it was just carried on into 2025.
DC
Why have you changed this website without testing to make sure it works?
Such as people being able to reply to comments
I am just testing the new website out. If this reply goes through, it is ok from my point of view.
Iver,
People can reply to comments, but the format was not good for smartphone use. It has been changed to only one level of replies to make it readable on a smartphone.
As far as changing website, there has been an update to software that has made the change necessary, I am not adept at this. If you would like to donate thousands of dollars so a developer can create a better blog you can email me at [email protected].
Just went looking for the latest peak oil review (or ASPO.net, the thing that Tom Whipple published, forget the exact name). Evidently Whipple died in late 2024. His last bulletin was in late 2022.
https://energybulletin.org/
While he was a hard corps peaker, that bulletin was a nice intel compilation of the details of publicly available sources. Solid stuff.
Yeah I know you think you’re on top of it. But Dunning Kreuger. Read the methodology. You are criticizing without understanding.
The ratio is established by the method in the pdf, that is too mathy for you. It’s a valid method. That doesn’t mean it is always right. It’s impossible to have perfection given the stochastic data along with trends in the reporting percentages. But the method they use is the best way to guess the ratio.
Nony
I am assuming this reply was misplaced and was supposed to be further up. The 914 methodolgy is pretty basic statistical analysis. The data they used in their analysis was from 2015 to 2018. The bottom line shown in table one is that NM oil production data lags by six months. That is exactly why the 2025 NM data is all wrong.
Let’s check the next EIA comp-stat-oil file and see if the EIA’s January production is revised down to the NM published production number.
I will try to remind myself to raise this issue with you a few months from now.
The commenting seems broken. (Could be me, but it seems some of the glitches are from the site.)
I know you like your “two county” thingamajig. And it’s probably fine to play with it and noodle around if you have some insight. But I think it would be crazy for EIA to use your method to determine the unsurveyed multiplier.
As discussed, they have a methodology (which I doubt you have read, have read twice, have read thoughtfully). ESPECIALLY pages 4 and 5 of the pdf! Honest, I suspect, on this site, Dennis is the only one other than I who have read it and thought about it, deeply.
Essentially, they go to older “frames” of reference and look at the gap between surveyed and eventual “semi-confirmed” amounts, that are a few months old. [Last “reference complete month” along with the six months before that.] (“Semiconfirmed”, because eventual state final records can change for up to 10 years!) They look at six months of such data (to avoid gyrations from one time blips).
And they use this methodology for all the surveyed areas (top 15 states plus FGOM). It would be insane for them to use the “Ovid method” in NM (occasionally, even) and then a statistical method in the rest of the areas.
It is essentially impossible to know how fast the unsurveyed population is growing versus the surveyed. They may differ in behavior. (E.g. small guys being more aggressive, trying to grow and get bought…or the converse, during a downturn, having less rigs on fixed contracts.) And it’s also an issue that there is constantly small amounts of land being sold between different players, such that a small guy could acquire wells from a big guy or visa versa.
The EIA has a sound statistical method to do the best they can. It’s not perfect. But (on average, over basins, over time) it’s as likely to undercount as to overcount. You will only know the perfect answer in hindsight. The records can get adjusted up to 10 years for oil and up to two for gas.
They are not fiddling with the multiplier to try to confound you. And given how they look at a six month comparison of frame, versus estimated, it’s actually somewhat reasonable to expect some trends (not snapping up/down).
For that matter, a 1.08 multiplier is not really obscene. (Doesn’t mean it is right…but it’s not crazy.) The EIA runs a process such that they survey at least 85% of each of the top 15 states (plus FGOA) production. But then these producers have to report on production in all (top 15 states, plus FGOM). In effect, we see some very small states (within the top 15) slightly below the 85% target, as they also have an allowance to not make operators doing less than 500 bm in any surveyed area, to not have to fill out the form. Many other larger states are a bit over 90%.
A 1.08 multiplier implies 92.6% coverage. A 1.05 multiplier implies 95.2% coverage. There’s as much reason to think 1.05 might be low as 1.08 high. Or even that both are high or both low. Or even that the multiplier really was changing over time. These guys are continually buying and selling acreage. And also different operators (surveyed and unsurveyed) can grow at different rates, even with no land sales.
I would really just roll with it. That’s what we get. If you think state data is the end all, be all, than there’s no purpose in even doing the 914. But it sure does seem to get revised itself, the state data, month to month, as well!
Also…I really don’t see why you have this whole fetish with the “two counties”. NM does have a report of statewide production, that comes out monthly. I was just on their site. It’s there…
Nony
I am surprised at the assumptions and conclusions you make.
Nowhere in my post do I suggest the EIA should alter their reporting process. My focus has been strictly focused on reporting detailed information for NM and Texas. After tracking them both closely over two years I think the NM data is at most one or two months behind being finalized whereas Tx is more typically 6 months to 1 year behind.
Attached is the comp-stat file for Tx. You can see the difference in the columns Final and State data. The differences between the two start in March 2024 with a difference of 29 kb/d. Compare that with the NM comp-stat file in the post. There are only 4 months in 2024 with a difference of 1 kb/d or 2 kb/d.
The only point I am making is that the NM production data is very up to date and I think it is more accurate that the EIA’s methodolgy. At the same time I am trying to make our followers aware that I think that the EIA is over predicting NM production. It is their choice to believe it or criticize it, fine by me either way.
I will keep checking the comp-stat file to see when the NM January 2025 production is changed by the EIA. Not sure how often the EIA updates the back data. When it changes, I will alert you.
As for my “thingamajig”. I am just trying to reduce the amount of data collection I have to do to put this US report together. I am thinking about cutting it back further. Any suggestions would be appreciated.
By the way, I didn’t pull the NM 1.5% correction out of the air.
A Table
Nony,
Can you explain your issue with the comments? I have only seen the legitimate complaint that on a smartphone for comments two levels deep the comments are not readable. That has been corrected by limiting to only one reply deep as I can’t seem to adjust the theme properly to accomodate a second nested comment.
Dennis,
I suspect it has to do with the email attached to my Anony handle. Seems like more likely to trip when I vary the email, than when I use the most frequent one. (In the past, seemed more the opposite.) But that’s just an impression…doesn’t seem 100% either way.
P.s. I use a nonsense (fake) email…hate the whole email collecting part of the Internet! After all, I don’t post real name either.
Nony,
You are not alone, the email is just an identifier so we know who has posted. If you did a johndoe type email address and just kept it consistent, you probably would have less of a spam problem, heck, use nony in the email address and I would know who you were and remove you from the spam.
You can tell who I am just by my style of posting. 😉 😉
https://www.lecturesbureau.gr/1/i-recognize-the-lion-by-his-claw/?lang=en
Nony,
Yes I can recognize you, but the spam filter cannot. A consistent email helps, one with anonymous in it might help less.
Ovi,
Note that the EIA does an adjustment to their data every August to reconcile the 914 estimate with state data, the August 2025 adjustment has been done, we will have to wait until August 2026 for the next adjustment. In some cases the 914 survey data gets updated/corrected and the 914 estimate will be revised, typically this happens only for the most recent 2 months of the compstat file.
Rig Report for the Week Ending September 5
The rig count drop that started in early April when 450 rigs were operating added a few this week.
– US Hz oil rigs increased by 4 to 367, down 83 since April 2025 when it was 450. The rig count is down 18.5% since April.
– New Mexico rigs dropped by 1 to 80 while Texas added 3 to 187.
– Texas Permian added 3 to 151. Midland dropped 1 to 20 while Martin was unchanged at 17.
– In New Mexico Eddy added 1 to 33 while Lea dropped 1 to 47.
– Eagle Ford was unchanged at 29.
– NG Hz rigs dropped by 2 to 103.
A Rig
Frac Spread Report for the Week Ending September 5
The frac spread count rose by 2 to 164. It is also down 56 from one year ago and down by 51 spreads since March 28.
A Frac
Dennis:
Another request, for the list (no rush).
Like to have some sort of monthly archives or the like. I sort of hack it by putting page numbers in the url (after hitting more, for older posts). But if I want to jump to some specific month/year, it is a hassle.
Nony,
Done for front page. I still can’s seem to get the comment layout the way I would like.
Also an archive has been added to the sidebar for the single post page.
Am I being moderated for:
1. Being a corny.
2. Being a mean guy.
3. New site software.
4. Other.
Not sure you are being moderated, sometimes stuff gets caught by spam filter and possibly an email address with anonymous may get caught by the spam filter.
Endless oil in US GOM,
I have been on a search of literature about the deep water Wilcox trend in US GOM, trying to find out:
1. why the oil in the depth close to 10km below sea level could still have oil, in fact still have low GOR &API oil. Whereas it is already almost all gas in the shallower part of US GOM. The regular marine source rock generation and migration models tells that such low GOR & API oil is dead oil with light components along with a lot more gas high graded to the shallower formations.
2. I read one article that has the Big White of the Wilcox trend bordering to Mexico with isotope data and GOR&API matching the hypersaline isotope-GOR-API. But besides that paper nothing else to support my hypersaline hypothesis. This hypersaline shale hypothesis could still generate oil instead of gas even under very high maturity where normal marine shale will only have gas left. Of course, the ultra-thick salt dome on top of Wilcox lower the temperature at 10km below sea level still at ~125C, not >200C as Bob Meltz corrected me and chatbot.
today, I read this paper and also several others that propose indeed the US GOM is a hypersaline deposit based on observation of sand and many other deposition properties (actually mentioned Art Berman), this correlates my observation from a totally different petro-geochemical isotope-GOR-API perspective.
https://www.researchgate.net/publication/340794375_Paleogene_Drawdown_of_the_Gulf_of_Mexico_Paleogene_Drawdown_of_the_Gulf_of_Mexico
This depositon along with the hypersaline oil and gas generation (not accepted by mainstream either, but quite obvious with major discoveries like Presalts on both sides Atlantic) models means that the chances of finding more oil, and low GOR&API oil with high productivity in US GOM is quite possible.
Of course, this proposal is not definitive and several others refute it or partially support by adding modifications.
https://explorer.aapg.org/story/articleid/56977/the-paleogene-wilcox-deepwater-play
http://www.searchanddiscovery.com/documents/2009/40418higgs/ndx_higgs.pdf
https://explorer.aapg.org/story/articleid/57465/it-is-still-viable
here is the short story I wrote,
https://www.linkedin.com/pulse/endless-oil-becuase-salt-lake-50million-years-ago-deep-sheng-wu-pc0sc
Rosenfeld also mentioned Art Berman in 2001/2002 noticed the drilling results from Unocal which first tapped the deep water Wilcox and found huge thick sand column, and this surprised Art Berman. And later in 2007 the two wrote a paper claiming this finding changed the GOM exploration paradigm.
Berman, A., and J. Rosenfeld, 2007, A New Depositional Model for the Deepwater Wilcox-Equivalent Whopper Sand: Changing the Paradigm: World Oil, v. 228/6.
Sheng Wu,
If I understand you correctly, you are saying that there may be vasts amounts of yet-to-be-discovered oil under the salt domes in the GOA?
Fascinating.
In a few months, there will be leasing auctions of some 80 million square acres in Da Guf, including deepwater.
With the ongoing rapid advances in seismic technology combining with the hard-earned experiences of the operators (both here and in Brazil), it should be interesting to see what develops.
Great work!
Hope you are correct.
Sheng+Wu:
I applaud your efforts but you do realize that there was tremendous runoff from the Laramide orogeny at roughly the same time that the world was in the PETM, with intermittent “hyperthermic” spells putting vast amounts of carbon into the oceans. A giant salt canopy was formed deep in the floor of the Gulf, but there are pockmarks all over where the salt was deposited in extremely thick domes, which acted as thermal protection against these high temperatures cracking the oil into NG. The shallower oil was cracked, because it had no protection. The deeper oil was cracked too, if it wasn’t under sufficient salt domes for protection. As I understand it, all the large pre-salt finds have been protected under extremely thick salt domes. Maybe Bob Meltz can help us; he’s the pro.
Thanks for the encouragements, Gerry and Coffeeguyzz!
I am not deposition geologists and actually had to learn from geologists trying to market my isotope machines/services.
The drilling results in the last 30 years serious changed the old models we had about oil and gas explorations, like the deep water Wilcox presalt metioned above by Gerry and shale drilling.
Like the now prevalent hypersaline crude which was only seen in limited locations before, mainly in China?, now a majority in the presalt trans-Atlantic, and even now in US GOM. This hypersaline crude generation model is serious different from the conventional marine and regular lacustrines, they are obviously thicker and high wax often make them look like premature or biodegraded fluid and often obviously hard to produce. That’s why when I logged the lacustrine Permian in China and it has similar maturity as North America Permian basin, and yet the crude in hypersaline lacustrine has API 22 and GOR less than 50scf/bbl, compared to marine at API42 and GOR >500scf/bbl.
But, in fact, often they are mature oil as mature as regular marine/lacustrine and produce just fine if not better with higher recovery often.
But Sheng+Wu, I suspect you’re correct. There’s probably enough presalt oil to supply the world, it’s just very hard to be sure where to find it and expensive to bring it up. I am amazed at how much more information is obtained with newer technology, so perhaps this will become old hat in time to save us from a massive supply shock. The GOM is fascinating because so much was happening at one time. The Paleocene-Eocene Thermal Maximus and its punctuations of hyperthermia explain the massive carbon sink that set the stage.
Some of the geology folks Sheng Wu has called upon are some pretty heavy hitters in regards to addressing these big scale geology issues, and they are out of my league.
Having said that, one of the reasons the drawdown model has been called upon is to explain the extraordinary scale (both in thickness and in aerial extent) of Wilcox deposition in the GOM. Conventional deep water depositional models have a hard time explaining that. If sea levels are lower in Wilcox times, then the Wilcox can be deposited in a shallow water setting where more widespread deposition is easier to explain. The John Snedden article challenges that saying that the whole core analysis and biostrat analysis done by industry on these deepwater Wilcox wells suggests a deepwater setting. (Larry Zarra, mentioned by Snedden, was a colleague of mine at Chevron and was recognized by industry as “the deepwater Wilcox stratigraphy expert”). That is my view.
Regarding the hypersaline source discussion and the possibility that it could result in a lot more oil being discovered: the Snedden paper addresses the hypersaline source and challenges that with his discussion about Cunningham’s geochem work. I support their conclusion that the source was not deposited in a hypersaline setting, and that is consistent with it being in a deepwater setting. Also, the oil in place is already fairly well established for the Wilcox fields that have been discovered. (In other word, you are not going to find lots more oil in the “already discovered fields.”)
The only way more oil is going to be discovered is if more structural features are identified, and as modern day seismic data continues to improve, that will probably happen. The December sale will be interesting.
Also, regarding the notion that the Paleocene Eocene Thermal Maximum (PETM) could be associated with a lowstand (low sea level) in the GOM – it just doesn’t make sense to me. The high temperatures associated with the PETM would equate to a high stand (high sea levels) in the GOM. That is in line with the notion that the Wilcox in the present day deepwater was deposited in a deepwater setting.
Thanks a lot for the notes, Bob Meltz!
I read Snedden paper on Explorer site first while searching “hypersaline”, “GOM”, and found that he is refuting the “hypersaline” idea. Then, the model by Rosenfeld mentioned in this paper led me to read the possible draw-down details.
There is a later paper by Cunningham and Snedden confirming there is ventilation to the outer atlantic ocean during the PETM, but the PETM did left the salinity higher
https://www.sciencedirect.com/science/article/pii/S026481722200112X
“with ventilation reduced but not eliminated by increases in salinity stratification due to runoff and reductions in deeper water entry into the GoM at gateways.”
Similar thing might happened in Vaca Muerta, the geology might not totally close the ventillation to Pacific, but salinity probably went up. Mainstream argue that Vaca Muerta is similar to Eagle Ford depositon, like the Wilcox trend oil sourced rock deposition is also assigned as Eagle Ford alike, but the oil and gas produced from Vaca Muerta shale is quite different from Eagle Ford shale, with low API and GOR, high wax yet highly productive.
I am trying to correlate the low API and GOR oil at deep >4km or inside or next to the source, where phase separation did not alter the original fluids generated, to the deposition and catagenisis environment.
One of the newer perspective of the oil and gas generation from source rock is that — besides initial deposition conditions which decides the kerogen input source and kinds, later catagenisis of oil and gas conditions from kerogen are also important. So, although there is ventilation to outside ocean based on fossile evidence, there is still higher salinity that could alter the oil and gas generation. This higher salinity really caused the oil generation to start earlier Ro<0.6 and last to very high maturity Ro~1.5, while suprressing the gas generation from normal marine's gas starting maturity Ro=1.0 to Ro=1.3.
There is also another later 2023 paper,
https://www.researchgate.net/publication/368411338_Carbon_isotope_and_biostratigraphic_evidence_for_an_expanded_Paleocene-Eocene_Thermal_Maximum_sedimentary_record_in_the_deep_Gulf_of_Mexico
Seems to support paleo-drainage and cause the sand sediments.
Sheng Wu – I salute you for your tenacity!! You have more energy around this topic than I do!
Do you think the source beds sourcing the Wilcox reservoirs are different than those sourcing the younger Miocene reservoirs? I bring that up because the Lower Miocene oils especially in the subsalt fields of southeast Green Canyon (from fields like Mad Dog, Atlantis, Tahiti and Shenzi) have similar characteristics to the Wilcox oils, low API and GOR. While in these same fields, the Middle Miocene reservoirs have higher APIs and GORs.
Is this just a possibility for the GOM? Or something that could apply, presumably in other areas also.
Sort of wondering if it could be like shale. That has been a NAM story to date, but geologically, there’s nothing that special about NAM. So, given continued advances in technology/access, could see shale develop in ROW. Certainly some signal to do so as prices increase, at least in non OPEC areas. (And as such, tends to limit the size/danger of extreme long-term price increases.)
Just wondering if this sub-salt stuff is similar potential. Sure, it’s hard to explore for it. But that’s the sort of thing that we might expect continued improvement over next few decades. After all, look how far geophysics has come so far.
Bob Meltz,
Thanks for your continued discussion on GOM oil and gas.
I have been close to blank paper to GOM O&G until 6 months ago until I read your writing on GOM O&G in this forum. But I was still confused with all the “Miocene, Pleistocene, Tertiary etc.”, and I actually believed Mad Dog is already Deep Water Wilcox (DW2). Chatbot seems to say DW2 is Tertiary which includes Miocene, so DW2 is below Miocene, actually should be Paleocene and Eocene ?
Now, which source rock generated oil in Tertiary DW2, Miocene (MadDog, Atlantis,Tahiti,Shenzi) ?—
I believe there could be several for different reservoirs, and if they are deep enough and no separation of oil and gas after generation and migration (separation will generate very low API and GOR fluids and gas pockets), then usually marine source rock already have API>35, and GOR >350scf/bbl; while hypersaline lacustrine source rock generated oil could have API <20 (Shengli China even have shale oil at API 14deg) and GOR down <50scf/bbl, e.g. Vaca Muerta has most prolific shale oil at API 25~30deg and GOR<350scf/bbl.
I have discussion also on this topic in the linkedin post I wrote above, and Dr. He says the DW2 oil or GOM source rock is " Tithonian, marine orano-facies A, similar to the Eagle Ford." I chatboted that Tithonian is latest Juarssic, much earlier than Tertiary and I doubt only this deeper souce rock could source the fluids in Miocene and Tertiary, assuming the PETM and other events already leaked much of the oil and gas generated.
Instead, I believe later source rocks could also generate the oils, e.g. this paper here shows the existence of shale in the Wilcox?
https://www.researchgate.net/publication/368411338_Carbon_isotope_and_biostratigraphic_evidence_for_an_expanded_Paleocene-Eocene_Thermal_Maximum_sedimentary_record_in_the_deep_Gulf_of_Mexico
also, this paper below here figure 7 & 8 shows that there is TOC shale in the Wilcox? and the TOC carbon isotope is lighter to the oil prone source rock? and mature enough in Great White?
https://www.sciencedirect.com/science/article/pii/S026481722200112X
just now, Dr. He confirmed that he believes the Wilcox in DW2 at 8~10km depth is not mature enough, and only the Tithonian is responsible for GOM O&G. “ Where Wilcox is targeted as reservoirs in deep water, the Wilcox is immature obviously. There are some TOCs in the Paleocene and Cretaceous in rafted sections. They are low and its contribution in but may have drawn out by the super rich Tithonian. It may be a better source rock in the shelf mixed oil and gas areas. ”
But I believe the literatures and drilling results are not what he said.
Anonymous,
The GOM O&G particulary the DW2 are quite diffrerent, and it is against normal convenitonal model, i.e. deeper drilling = higher maturity = higher GOR to all gas — GOM O&G has this conventional all reversed, and in a very significant way.
In my Linkedin post above, I have borrowed and modified Dr. He’s PPT on GOM O&G — you could see as it goes deeper, the API and GOR all gets lower. This PPT did not include the shallow to medium depth (<3.5km) GOM O&G yet, majority shallow to medium GOM O&G are gas and condensate only, i.e. much higher GOR and API than shown in this PPT.
Panama just announced plans to go forward with their much-discussed concept for an LPG pipeline to cross the isthmus.
Rough numbers …
Cost under $8 billion includes ~50 mile pipe (36″/48″ diameter), storage and piers both ends, all related infrastructure.
Currently about 4 ships transit the canal per day carrying propane/ butane.
By sending smaller, older propane-carrying ships (which are rapidly becoming obsolete) down from the Gulf to the proposed Caribbean-side terminals, piping product to the Pacific side, and then loading the (rapidly growing) Very Large Gas Carriers (VLGCs), significant cost reductions are expected.
Simply bypassing the bottle necked Canal would itself be a major plus as reliability and shipping schedules would not be at the mercy of droughts and maritime congestion.
Big deal, actually.
I’ll believe it when I see it. Every few years we hear about the development of the neutral zone, also. But it never happens.
So an announcement doesn’t mean jack. Even an FID doesnt. Since for Panama, it’s just an announcement. I need to see trees being cut down, before I believe.
I’m treating announcements like Lucy and the football. Like the boy who cried wolf.
Dennis,
1. My posts seem to be going through more.
2. Thanks for the edit feature.
3. Thanks for the archives.
P.s. Hope all this website crap is not a downer for you!
Thanks Nony,
Not a big deal, but thanks for the encouragement.
Glad things are working a bit better.
Sometimes your posts get caught in the spam filter, not sure why as it is a third party plugin, and obviously my programming skills are not top notch (or even close.)
Ovi:
1. I feel like the warden in Cool Hand Luke. (Failure to communicate.) I’m cool with playing around with an analysis and maybe having some implication of how EIA might be off.
Just cautioning that there’s been a lot of such speculations and it’s not clear to me that systemically, there is a better way to do things…and as such would be wary of picking out the occasional times/places where a different method gives an answer showing EIA “off”. It might be the ad hoc method that is off, this time. Hard to tell which to trust more.
And I’m not “accusing”, just raising an interesting past learning, here. The guy Dean Fantazani (name?) who used to always take RRC data and do some manual corrections to it and compare to EIA, never really demonstrated that it worked better than EIA…just makes me wary of ad hoc method comparisons. I mean…again, fine to play with it…but also, I have some skepticism/wariness, based on Dean’s work. At least you are doing more discussion/comparison to EIA (like your chart of multiple over time was a nice start).
Again, not exactly an accusation, but I’m wary of expressions on the POB (or TOD or PO.com) that EIA is (or may be) high. Just seems like these get highlighted more than the opposite. Also, there can be a bias to look for or talk about the “EIA might be high” more than the “EIA might be low” examples more on this site. Perhaps some general analysis of EIA revisions would be interesting…how much do they revise down versus up, say looking at 13 month view versus last month view (i.e. a year later). I might even do this. Hmm.
2. Also, even if you’re not proposing EIA stop using their current system and rely on NM state data instead, well…perhaps they should. I mean if you really think the (recent) state data is more to be trusted. After all, people have made this argument wrt ND data as well, where state data tends to be quite good. If it’s really a better metric, maybe they should do it all the time, at least for “good” states. (We know TX is not a “good” state…and I’ve heard very bad things about OK data as well.) But after all, they do use pipeline data for AK (or most of AK).
3. I’m not saying your % in top two counties is right/wrong. I’m sure there is a decent rationale for the value you pick. I just honestly wonder why you would use a proxy like that instead of the state reported totals.
Is there something I don’t understand? I went and looked on the site and there seemed to be a state reported total. I mean…it’s not a huge difference as 1% of 2.0 MM bopd is only 20 M bopd. And if you were off by say 0.5% on the split of two counties versus rest of NM, it would only be 10 M bopd difference. But still…just wonder why bring another parameter in, instead of using the state total. It’s even just partly a question, as opposed to criticism. Maybe there’s something I don’t understand when I look at what I think are state totals.
4. I seem to remember (but could be having a senior moment) that in the past NM had some sort of revisions up over time. Maybe a lot smaller than TX, but similar in nature (stragglers reporting in, or misfiled data in a pending log for checking or something).
It could be that they’ve gotten better…and this was an old thing. Or could even be that I’m remembering wrong. But I just sort of remember this.
I did try looking on their site for old versions of the monthly report, but I can’t find them. Might be I just don’t know how to navigate it. Like for ND, you can look at all the old Director’s Cuts (or old monthly reports). But for NM, I can only see the most recent one. So, it’s hard for me to see how much revision they do over time.
Again…it’s just an “I sorta seem to remember this” thing.
Nony,
I have older NM data and it does get revised higher over time, just much less than Texas. The data 6 months old and older is quite good (EIA does use this for their final estimate that is updated annually in August ), you can see this in the compstat file where 914 estimate is different from the 914 estimate.
Thanks, Dennis.
FYI, I’d be curious how much it changes. Something like the brown chart for RRC data on the TX chart.
Nony,
If we look at the compstat file for NM
https://www.eia.gov/petroleum/production/xls/comp-stat-oil.xlsx
The correlation coefficient for 914 estimate vs Final estimate from Jan 2015 to Jan 2025 is 0.995. Also if we take the average from Jan 2015 to Jan 2025 the 914 estimate is 1063.07 and final estimate (reported state data at end of July 2025 for these months) is 1066.48. Thus on average the EIA 914 estimate is slightly low by 0.32%.
click on chart to enlarge
new mexico 2509
DC,
That’s nice at showing EIA does a decent job.
But, I was asking how much the state data itself (not EIA) changes over time. Like how much “RRCing” NM does. Right panel of this chart:
https://www.eia.gov/todayinenergy/images/2018.03.26/chart2.png
Or even just a version of what Ovi shows as the brown line for TX. (i.e. just the most recent state report.)
Nony,
Did you mean the right panel of that chart?
The previous chart was to show that the EIA estimates are pretty darn good. In my view there is no need to reinvent the wheel as it were.
If I look at the most recent NM state data, it makes me suspicious that they have “RRC tendencies”. Maybe not as bad as TX, but similar in nature, requiring the use of a “projection” to compensate. This projection itself becoming a source of error/variation as we’ve seen from the Dean Fantazzani work. (If “projections” were more efficacious than 914 survey work, then we could save a lot of time/money/hassle for EIA and the operators, by just using “projections”, but Dean found, for TX at least, that there was no fixed multiple for “how delinquent” the state data was.)
[This is in contrast to ND, which, while it has some revisions, doesn’t seem to have a consistent undercount bias, that it firms up. ND state is very close to final…and seems like it gets revised up or down, not only up.]
Like I said, I don’t have access to old NM state reports, but just looking at the most recent one, makes me suspect they have “RRC tendencies”.
https://wwwapps.emnrd.nm.gov/ocd/ocdpermitting/Reporting/Production/ProductionInjectionSummaryReport.aspx
See final column of the table, headed “Total Oil”. Note this is in absolute bbls, but I opened the spreadsheet in Excel and divided by the appropriate number of days (varying by month) and got the following values, for JAN to JUN:
2,014,310
2,097,871
2,183,412
2,132,160
2,145,036
2,126,533
As you can see, MAR is the highest listed, which makes my Spidey sense tingle. If NM has “RRC tendencies”, then this is unlikely to endure as the report gets updated over the coming months, especially if NM is generally growing. I could confirm this, if I had access to old state reports. (But appreciate your comment to the effect that you do and have seen NM to have RRC “disease” albeit less extreme than TX does.)
DC, yes…right side, not left. The state vintages.
Nony,
Here is some data I found on my computer from NM for C plus C.
It looks like the most recent 6 to 8 months get revised higher over time, data more than 6 to 8 months old is pretty solid.
click on chart for better view
nm state data
Thanks, just doing some quick and dirty estimation:
Eyeballing that chart on the last month of the two older vintages, it looks like a 50-90 M bopd undercount of most recent month.
Adding that to the most recent state total report (2,126 M bopd) gets us to 2,176 to 2,216. Split the difference and call it 2,196. This is in comparison to the EIA value of 2,239.
It is a lower value, but more like ~40 low, not ~100 (Ovi’s view). Really to get something like 100 low, you almost just have to use the unprojected (as is) state report (2,126 versus 2,239).
I would probably just go with the EIA report anyhow. It’s not perfect, but it can be as likely low as high. It will get fixed eventually if it’s an overestimate. And might even be an underestimate. And the projection comparisons have been flawed in the past. Just am wary of the projection game, given how Dean showed, it doesn’t work very well.
Nony,
Keep in mind that the EIA is also simply making a statistical projection of future data fill in as state data gets updated over time. I agree that they do a very good job of it and I simply use the EIA data estimates. When looking at individual counties my method is to use the percentage of county output relative to state output (as reported by the state) and multiply this percentage by the EIA 914 estimate (for recent 6 to 18 months depending on the state in question).
Often it is interesting to look at the individual counties in the core area of the Permian Basin. For instance, there are about 10 counties in Texas and New Mexico where most of the drilling and completion activity occur.
Nony
First off thanks for your frank thoughts. I do have a thick skin as long as there are no personal shots. The participants on this site are good that way.
After I stumbled onto the Fantazini method myself, Dennis informed me that what I was doing was the Fantazini methodology that didn’t work. There is logic in the methoIogy. It depends on consistency and monotonic changes.
I looked at averages and least squares and finally decided to just compare the last two months. I then look at the differences and if they are monotonic and have no gaps I asses whether they are reasonable, optimistic or totally bad. If total bad I don’t publish. Check Midland in the post where I compare two months.
Bottom line is that sometimes the two month method works and other times it doesn’t. I then let the reader know my thoughts on how reasonable the projection is.
Attached is the June Midland projection which shows the methodology. The Diff column shows only 6 months of revisions and the data is monotonic. The NM projections in the post, green and orange graphs sitting on top of each other indicate very small month to month change except for the last month. For Midland there is a small gap.
A Tab
Nony,
Also keep in mind that Ovi and I have discussed this at length in private emails, we decided it is nice to have different approaches to estimating future revisions to C plus C output in various States.
Nony
The NM site I go to does not have a monthly to combined total for all counties. It show all of the counties separately. If you have one where it has the combined monthly totals, I would appreciate if you could provide that site.
Ovi:
It is the NM state O&G site (OCD statistics), with the ninth link in the center of the page, titled “Statewide Natural Gas and Oil Production Summary by month”:
https://wwwapps.emnrd.nm.gov/ocd/ocdpermitting/Reporting/Production/ProductionInjectionSummaryReport.aspx
Nony
Thanks. My report goes to the county level and gives the same result when the eight oil counties are summed.
Ovi:
I think most people are most interested in the new EIA numbers and the top 10 + GOM graphs.. This report is so long, it is hard to read it all and parse it all and I fully admit to tuning out the second part (sorry…but just being honest). It’s not just that it’s so long, but in addition, the stuff later in the report seems very intricate.
I would consider to either:
1. You could consider to do monthly posts on the EIA numbers report. And then just do the more elaborate/detailed stuff less frequently (and separately).
2. Or if you must do them monthly, consider to break this post into multiple posts.
It doesn’t kill me to have the longer post…I just read and comment on what I do. But…feel bad that you do the work and it’s not read, much. Also, if in some cases (e.g. the projections for NM), if there’s a need to understand the methods discussed later in the report, I seem to not be tracking it all.
P.s. Not in any way to be meant as a downer. Just trying to engage and being honest about the difficulty in grappling with the size of the article.
Ovi:
Mea culpa, but I’m better understanding now what you are doing with the NM Lea+Eddy. It’s not JUST two counties (from state data) plus a guess on the other counties (the 1.5% factor), but it’s ALSO a “projection”. Because the state data DOES have a lag.
NM lag may not be as bad as TX. But very similar in nature, what you are doing here. If anything, I’d be more worried about the “projection” aspect of the blue line versus the two county equals 98.5% aspect.
The “projection” game is the same thing Dean Fantazzani was doing for a long time (he may even still be doing it, on autopilot). And for TX, it seemed like it didn’t really work. Not only was there a lag/underreport, but the AMOUNT of that, tended to differ over time. Sometimes, he’d used a 12 month regression, sometimes playing with last month. And they were quite different…and there was no clear, good way of best estimating what the final (12 or 24 months later) result would be. Basically, the EIA 914, which is not a perfect predictor either, seemed to work better.
[Which sort of validates even having the 914. I mean, if TX projections work better than surveying, they could save a lot of time/work and not have operators filling out that complicated 914 survey, not bother chasing operators that miss a filing, etc. If projections are better, than just save a lot of time and money and use the state data that’s already being compiled. In a way, EIA sort of did that before they came up with the 914 survey…but they decided it was not good enough…at least when shale was growing, gyrating rapidly.]
So…you are actually doing a NM projection. It’s not the two county thing that worries me most. It’s the “projection”. Those things have a bad record. They’re not stable predictors. If you can show somehow that NM projections are systemically superior to EIA 914, I’d buy it. But just flagging this one time where they vary? Could be the projection being off, not the 914. You can’t tell which is driving the delta!
1. By the way, I just noticed you have a TX RRC projection also! Yikes. That is the failed Dean method. (Mea culpa, just noticed…like I said, I usually just look at EIA as “the score”.)
Also, that TX RRC projection is ~150,000 above the EIA. Whereas NM was ~100,000 under it.
So…if we really think projections from lagged state data are better than the 914 methodology, shouldn’t we credit both of them? In which case EIA total for the country might even be an undercount!
——–
2. Also, I don’t even know what this means: “The red graph is a production projection using the May and June Texas RRC data. Due to more production revisions and errors in the overall Texas data, the projection is a modification of the normally used methodology.”
Like…not even to immediately criticize, but I can’t even grok what you are saying, here. What is the “normally used methodology”? How did you modify it? And was the modification just for this month?
Nony
The change had to do with two months back when TX data and District 8 was totally wrong for 2024 but 2025 data was good. For that report I just used the 2025 data to update and said it was probably an under estimate. The May and June data the next month was better so the modification was to use both 2024 and 2025 Tx data to make the projection. The Tx projection is high because so many back months are updated.
Does anyone here track new drilling permits are a predictor of future activity?
Ovi
I really appreciate your work, very insightful. I think you have a point there with the NM numbers. Interesting to follow it up. As a whole, I would not change a lot. For new readers, it is good the article is complete and thorough so you can understand it in one read. For new reports, it is quite easy to track it for new data and updates once you know the structure.
Thanks, Juha
Juha
Thanks. Much appreciated. I do think of new readers when creating these posts.
China is Driving Nations away from Fossil Fuesl
Since the beginning of the industrial age, the global economy has required more and more fossil fuels — coal, oil and gas — to power growth.
It is increasingly clear, however, that China’s aggressive efforts to sell batteries, solar panels and wind turbines to the world is on course to bring that era to an end, a new report says. The Chinese dominance of clean-energy industries is “creating the conditions for a decline in fossil fuel use,” according to a report by Ember, a research group focused on the prospects for clean-energy technologies.
If Beijing is trying to wrest the future of energy from anyone, it would be the United States, the world’s biggest oil and gas producer and exporter. The Trump administration has eliminated almost all federal support for renewable energies and has pressured countries to purchase American fossil fuels as part of trade deals.
The falling cost of renewable energy, though, means that many countries, particularly poorer ones, have a strong incentive to reduce their reliance on fossil fuels.
Which country is on the right track?
https://www.omanobserver.om/article/1176268/business/energy/china-is-driving-nations-away-from-fossil-fuels-report
They are doing a great job in making solar and batteries, cheap.
But they, themselves are massive coal consumers (almost as much as ROW combined). And growing rapidly, recently.
https://ursaspace.com/wp-content/uploads/2024/09/Chinas-Coal-Production-Consumption-new-1024×681.png
P.s. I dislike the Trump finkery with asking people to buy US oil and LNG. That stuff will get sold at market prices anyhow. At this point, they are fungible. Are commodities. So, it sort of doesn’t matter if country X or Y buy our stuff. It’s getting dumped on the international market either way. That’s just Orangeman trying to take credit and be in the news/Twitter cycle. Like how he took credit for the India/Pakistan discussions. He’s a glory hog.
Nony,
China’s average annual rate of growth in coal consumption in exajoules from 2011 to 2024 was about 1% per year, from 2016 to 2024 it was about 2% per year. This rate may slow as they use more solar and batteries over time.
From 2000 to 2011 by contrast China’s coal consumption grew at about 9.6% per year on average. Data comes from the Energy Institute’s 2025 Statistical Review of World Energy.
china coal 2509
DC, 1-2% CAGR looks low. Perhaps this is an artifact of the source you are using, or more significantly that is is showing the LOG, not the absolute (and you aren’t deconverting for the log).
Using the source I gave, and eyeballing the chart, it’s about a doubling from 2016 to 2024, which is about a 9% CAGR. Not 2%.
Hi Nony,
The slope of the linear regression for the natural log of X will give the exponential rate of increase for X.
I will go with the Energy Institute’s Statistical Review of World Energy 2025 link below
https://www.energyinst.org/statistical-review
The chart you found looks wrong, but would need a link to the source article, maybe it is coal used for electric power only? From 2015 to 2024 electric power output from coal grew at 4.2% per year.
From 1985 to 2002 electricity output from coal power plants increases at an annual rate of 9.0%
From 2003 to 2014 electric power from coal in china grew at an annual rate of 9.2%.
china coal 2509b
Yeah, I downloaded that report and it’s 1+ % per year. The numbers themselves are higher in that report (80-90 exaJ). So maybe, it’s inclusive of metallurgical coal. Which is fine…you’re still using it. So your larger number makes more sense.
There’s probably also a phenomenon of coal moving out of home heating and into electricity plants.
And at least it’s a real report, not a random graph that I Googled. 😉
https://www.reuters.com/business/energy/private-chinese-firm-producing-oil-venezuela-under-rare-20-year-pact-source-says-2025-08-22/
China investing 1 billion in 2 Venezuelan oil fields
Dennis
Note that in the article it states the coal plants may be more for part time power or possibly peak power.
Also I wonder if all these coal plants are just another job creation effort.
Ovi,
Though the political system is terrible, China seems to be on the right track as far as energy policy, hopefully there will some day be democratic reform and greater political freedom in China, but probably not in my lifetime.
I think of coal plants as stereotypically baseload, not peaking. This is because steam turbines require warmup cycles that can take hours. Also some of the solids handling (the coal) is not an easy on/off. There can also be cooldown tasks needed, for a steam plant. Securing feed/condensate, air ejector steam, etc.
Natural gas (or diesel) generators are much more easy on/off and thus work well as peaking plants.
Nuclear plants are much like coal plants. Baseload. Even longer startup, shutdown.
Anonymous
Do you actually know someone who works at a coal fired power station?
Do you know someone who works for the National Grid responsible for balancing the grid?
It’s pretty much my surface, general impression. (And I admitted that!) Not an SME. I could easily be talking out of my butt. I do that sometimes. 😉
Yes, I do have friends in utilities, but haven’t really interviewed them. Again…just my general stereotypical impressions.
I guess I’ve also worked at a couple places that had coal power plants, but for onsite industrial power. (24/7/365 usage, with purchased city power as the site’s “peaker”, and with an unwillingness to supply the grid in reverse, from the coal plant). But I had nothing to do with them.
P.s. A Google search and AI response, gave this answer to “are coal plants used more for baseload of for peaking?” [my emphasis added, for “traditional”]
“Coal plants are primarily used for baseload power, designed to run continuously to meet the constant, minimum electricity demand, but they are increasingly being cycled to operate as intermediate or, less commonly, even peaking resources as newer energy sources emerge. While their TRADITIONAL role is baseload, economic factors and the rise of renewables have led some coal plants to operate more flexibly, though their slow start-up times make them less ideal for meeting sudden, high-demand peaks compared to other technologies.”
Anonymous
Your AI question twisted your original statement. Obviously coal power plants are mainly used for baseload demand. They are not switched off and on due to the damage that would do to all the furnaces and boilers. However that does not mean they can’t be used for peak demand.
The furnaces are fed by powdered coal blown into them, the blowers are fed via conveyors which can be speeded up at a moments notice.
So power can be increased in a few minutes.
This is information from someone who worked at a coal fired power station.
Those who manage the grid know with some accuracy how much electricity will be needed at peak. They give power plants plenty of notice to increase power, so coal can do this no problem.
How do you think we coped years ago when ALL power came from coal fired power stations?
Good point. And the same applies for nuclear, mechanically. You can just operate at less than 100% steam demand. There’s no cooldown/warmup issue then.
Of course, it doesn’t help if you’re “already” at 100%. (There’s no surge available.) But of course the grid generation needs to be sized for peak demand somehow anyways. Probably more surging with coal than nuclear. since the fuel/capital cost is higher for coal than nuclear. Once you got that nuclear albatross built, you want it running.
I sort of pictured it as coal and nuclear operating at 100%, with small gas peakers coming on quickly as needed. And I think this is how some of the grid operates. But yeah, you could just have the excess capacity, within the steam cycle plants.
An update to World and Non-OPEC Oil Production has been posted.
https://peakoilbarrel.com/may-world-and-non-opec-oil-production-flat/
A new Open Thread Non-Petroleum has been posted.
https://peakoilbarrel.com/open-thread-non-petroleum-september-11-2025/