The OPEC Monthly Oil Market Report (MOMR) for September 2025 was published recently. The last month reported in most of the OPEC charts that follow is August 2025 and output reported for OPEC nations is crude oil output in thousands of barrels per day (kb/d). In the OPEC charts below the blue line with markers is monthly output and the thin red line is the centered twelve month average (CTMA) output.
OPEC 12 output for July 2025 was revised down by 73 kb/d and June 2025 output was revised higher by 5 kb/d compared to last month’s report. OPEC 12 output increased by 478 kb/d with the largest increases from Saudi Arabia (258 kb/d), Iraq (122 kb/d), UAE (87 kb/d) and Kuwait (40 kb/d). Iran (27 kb/d) saw decreased crude output. All other OPEC members had small increases or decreases of 13 kb/d or less (net decrease for all other OPEC members of 1 kb/d).
The chart above shows output from the Big 4 OPEC producers that are subject to output quotas (Saudi Arabia, UAE, Iraq, and Kuwait.) After the pandemic, Big 4 average output peaked in 2022 at a centered 12 month average (CTMA) of 20849 kb/d, crude output has been cut by 1378 kb/d relative to the 2022 CTMA peak to 19471 kb/d in August 2025. The Big 4 may have about 1378 kb/d of spare capacity when World demand calls for an increase in output. Since April 2025 the OPEC Big 4 have increased output from 18346 kb/d to 19471 kb/d, and increase of 1125 kb/d in just 4 months (an average monthly increase of 281 kb/d.) Also the increase in Big 4 output over the past 4 months is 13 kb/d more than the OPEC 12 increase.
The chart above shows the most recent 36 month average annual increase of 134 kb/d for the Other 6 OPEC group which excludes the Big 4 and Iran and Venezuela. Iran and Venezuela have seen their output rise over the past 3 years at an annual rate of 405 kb/d which I believe will fall to zero in the near future. The chart below shows the long term trend for the OPEC Other 6 which is roughly flat (a small average annual rate of decrease of 8 kb/d over the past 11 years.) I expect the Other 6 will see their output return to this flat trend in the next few years.
No changes in the supply and demand forecast compared to last month’s report.
Refinery crude throughput is a measure of World demand for C+C (though imperfect because some crude is burned directly in power plants in the middle east.) OPEC data shows the peak was 81.68 Mb/d in 2018. For the most recent 4 quarters the average World refinery throughput was 81.52 Mb/d, the previous 4 quarters had average refinery throughput at 81.09 Mb/d and 2024 had refinery throughput at 81.03 Mb/d. If the recent rate of increase continues (490 kb/d in 9 months) we may see a new peak in annual average World refinery throughput in 2025 or 2026.
Preliminary July 2025 data show that OECD commercial inventories stood at 2,761 Mb, a build of
2.4 Mb from the previous month. At this level, OECD commercial stocks were 66.5 Mb less than the same time last year, 128.5 Mb lower than the latest five-year average, and 208.6 Mb below the 2015–2019 average.
OPEC has again revised its estimate of US tight oil output lower compared to the previous month. The 2024 estimate is 10 kb/d lower. 2025 is 20 kb/d lower and 2026 is also 20 kb/d lower than the August 2025 estimate.
Peak US C+C output may be 2025 based on the OPEC estimate through 2026, US total liquids continues to increase after 2025 due to increased NGL, biofuel, and other liquids output (210 kb/d)
Novi labs had an interesting Webinar on Permian well inventory entitled Permian Staying Power: Inventory Now and Through the 2040s, though a bit on the optimistic side I used information on expected tier 1 to tier 4 inventory through the 2040s from the presentation to create low, medium and high oil price scenarios for the Permian. Low price is $70/bo, medium is $85/bo and high is $100/bo in 2025 US$ with natural gas assumed to be $3/MCF in 2025$ and NGL at 30% of the crude oil price.
The low oil price scenario has a URR of 46.5 Gb with 119.3 thousand wells completed. The medium scenario has a URR of 55.3 Gb with 149.2 thousand wells completed. The high price scenario has a URR of 66.4 Gb with 194.4 thousand wells completed. With higher output from low cost producers and falling future crude oil demand as the transition to electric land transport proceeds along with pressure depletion higher water cut and other cost pressures, I expect the low oil price scenario to be the most likely future for the Permian Basin.
187 responses to “OPEC Update, September 2025”
Iver (from previous thread):
I had not read the linked Seppuku/IEA document. But did so after you inquired. Well…I started reading it. But it was long…and not “news”, so I shifted to skimming, after a bit.
Overall: nothing groundbreaking. There’s decline, if you don’t drill. No duh. No. Duh.
Yawn.
Anon
If you don’t read you stay ignorant
I’m thinking of buying Yergin’s new book. The first one, The Prize, informed me on the whole history of the industry, especially history of supply cartels. Felt like it gave me good perspective as I headed off to Desert Storm and later for an interview with Esso, I mean Exxon. It’s a true classic, and even if you’re a peaker, it’s just a great narrative history of the industry, globally, from Titusville to Desert Shield. Got the second one also. It was fine. But just not as magnificent. So, now, debating on the third.
“We can’t kill our enemy without eventually killing ourselves.”
You might want to try “Here Comes the Sun: A Last Chance for the Climate and a Fresh Chance for Civilization”
by Bill McKibben
We’re near the start of something about as big as FF, and the industrial revolution: conversion from burning things to using sunlight. Burning is getting more expensive, sunlight is getting cheaper.
And currently China is getting way ahead of the rest of the world, especially the US as it tries to hang on to FF.
Just look at the latest Chinese sodium batteries: sharply cheaper than li-ion or lead-acid, much longer cycle life, wider temperature operation. Just astonishing…
Anon
I have been more critical than anyone of the likes of ASPO and followers like Ron over the years. Claiming peak oil almost every year since 2012.
However oil is finite.
https://www.westwoodenergy.com/news/westwood-insight/westwood-insight-norway-state-of-exploration-2014-2023
The graph below is of Norwegian oil production, no matter how much they drilled production has been falling since 2001.
https://www.thearcticinstitute.org/norway-energy-cooperation-barents-sea/
The U.K. used to export one million barrels per day at peak now it is an importer.
Do you know how many countries are now in decline?
Russia and that was before the war, Egypt, Nigeria, Iran, Mexico, Libya, Angola, Venezuela, the list goes on.
There comes a point that decline rates of existing fields can’t be matched by production from infill drilling or new fields.
Nothing can replace Ghawar in Saudi Arabia and it has been producing for 70 plus years. Production tipping points are happening in hundreds of fields around the world each year.
It is just a matter of time and every field will end up like Brent.
Never claimed oil was not finite. And it’s a “duh” type comment to say it. Sort of along the lines, of ZOMFG, oil wells have a decline curve. No sh&#, Sherlock.
I don’t know if Ghawar can be replaced or not. There’s nothing Euclidean proof-like about that statement. I mean…it’s like 5 MM bopd (approx) of production. Maybe 6% of total supply. If it went away oernight, it would be a hella shock. At the same time, we’re talking about less than the recent growth in oil supply (from mid 70s to low 80s MM bopd, from the TOD days until now).
For that matter, Ghawar worry for a couple decades now…that has underperformed. Simmons and Staniford died well before Ghawar did. And prices are in the low 60s right now (like low 40s in 2010 constant dollars).
I mean…it’s more than 20 years since the Simmons calamity book (2004). Maybe time to ramp down the calamity concerns. TOD and ASPO are a lot more dead than Saudi production. (Which is higher than it was when Staniford was sounding the alarm with “Nosedive” article in TOD, c. 2007-8. So much for his imagined irreversible decline!)
http://theoildrum.com/node/2325
You can’t answer any simple question can you.
All you can do is stupid comments that a ten year old could make.
Stay ignorant, too much effort to get properly informed.
Nony,
The Article, Nosedive in the desert was about Ghawar and it has in fact declined from 5 Mbpd to about 3 Mbpd, the last time the Saudis released info on it. Saudi Arabia has replaced those losses with output from other fields. They have never produced for 12 months at claimed capacity of 12.5 Mbpd and probably never will.
Yeah, you’re right. I named one and linked a different one. Staniford had a bunch of different articles, saying slightly different things. But if you look at the one I linked (yes, NOT named Nosedive), he predicted, for Saudi Arabia overall:
“Saudi Arabian oil production is now in decline.
The decline rate during the first year is very high (8%), akin to decline rates in other places developed with modern horizontal drilling techniques such as the North Sea.
Declines are rather unlikely to be arrested, and may well accelerate.”
And an 8%+ decline rate, HAS NOT been seen.
Staniford’s bet is quite amusing in that he doesn’t make it on what he asserts in the bullet points, but on some way higher goalpost. Which the Saudis would be crazy to do, even if they can, because it would CRUSH the price.
FWIW, your point that other fields replaced Ghawar is very similar to mine…that worrying about that specific field can be a fallacy. What matters to the world is world supply. Not Ghawar, specifically.
Oh…and here we are 20 years past the Matt Simmons silliness. SA is doing fine. And oil price is well below those that Simmons warned about.
Even FOR peak oilers, I’d be WAY more worried about Texas’s coming decline than the Saudis. Texas has been worked over much harder and has much higher base decline rate, with more rigs required to keep production flat. Saudi Arabia is an incredibly large and oil rich area, that has still been much less explored than Texas has been.
The Saudi worries…man that is some old, failed peak oil wine. Heck…let’s worry about Ace’s charts next.
https://www.zerohedge.com/energy/renewables-movement-making-itself-wholly-unappealing
I never tire of being right , do you folks ever tire of being wrong?
Texas tea,
Making a lot of money at $60/bo? Demand for oil is likely to continue to falter as the World moves on to electric land transport.
TT. We have several solar farms in operation here now, several more under construction and still more permitted.
Rental rates per acre are $2,500-,$4,500 per acre, with 2% annual rental increases, lease terms 30-50 years.
It’s taken a decent chunk of prime tillable farmland out of production, which is my primary concern. But the rent amounts are just too high to pass up on.
I had thought maybe the BBB would stop this trend, but it hasn’t.
Keep in mind, I’m an all of the above guy.
As for oil, natural gas and grain, your buddy in the WH continues to stab producers in the back.
Hey, TT,
Always glad to see your now-infrequent posts.
As a bit of a follow-up to the above Nony/Iver/Dennis convo regarding future liquid hydrocarbon supply … folks might do well to closely monitor the dizzyingly rapid evolution/adoption of AI.
On a personal note, my last several months’ interactions with ol’ chatbot lead me to the firm conclusion that its impact/influence will surpass that of the cel phone … itself one of the most innovative devices in recent history. People with far more knowledge in these matters – investing hundreds of billions of dollahs- claim the impact of AI will rival the introduction of the wheel as regards human development.
Laughable hyperbole?
Perhaps, but they sure are putting their money where their ‘mouths’ are.
Relevance?
Power.
Energy.
The essence of which has long been delivered – in large part – by oil and it’s derivatives.
Change is quickly coming as the oh-so-critical need/demand for electricity is prompting a frenzy of pathways by which ‘the juice’ will be generated/delivered in the next 5 years.
Natgas is in the forefront with GE Vernova, Siemens, and Mitsubishi ramping up their manufacturing capacity of aeroderivative turbines by 50 to 100 percent. (Aside, a commenter on this site claimed the imminent demise of the 7HA model over 10 years ago. Wrong, again.)
Big, BIG push in the nuke world with the micros/SMRs leading the charge. We’re talking 4 to 5 years to deployment with the Canadians’ Darlington plant leading the pack.
Globally, the resurgence of coal continues to display its dominance in countries where electricity demand overshadows other factors. (Sipercritical processes actually overcome many of King Coal’s perceived environmental concerns).
Bottom line, the demand for energy derived from oil will continue to be supplanted – and in clearly rapid fashion – in the coming years.
This relative demand decline will ensure no need for scarcity concerns whatsoever.
As said over and over … the Stone Age did not end ‘cuz we ran out of stones.
texastea11
The problem with renewables, especially solar, for the fossil fuel industry is that it really is incompatible.
Solar is very different from nuclear, which also competes with fossil fuel but acts just like a coal plant, working best when it runs 24/7. Also a significant part of the cost of nuclear is operating expenses, meaning you can cut costs by turning it off. Nuclear works well with the industry idea that the cheapest electricity comes from the steadiest source.
Solar has no operating expenses. Once you install it, it just runs, so the marginal cost of electricity is zero. That means that it can undercut the price of any competitor. Furthermore, its output swings wildly, which undercuts the business model of coal fired plants.
Selling fossil fuel is going to get harder and harder as solar spreads and energy becomes a low margin fight for market share. It doesn’t matter if you “like” or “hate” solar.
Solar is what the tech bros call “disruption” — a new product that doesn’t just compete, but changes the rules in the market.
Also Tyler Durden is a crazy person, don’t read his stuff, it will rot your mind.
shallow sand
How does that $2,500-,$4,500 per acre compare to farm income?
in china’s remote desert like areas where solar farms are deployed, they claim the solar panels helped lots of semi-desert land to grow grass and economical plants. The local sheep farmers could have more grass grown under the solar panels, and therefore they could raise more sheeps and ducks, and therefore more income.
DC,
In 11 days the Federal gravy train for electrics comes to an end, we will learn then how much love there is for the Plug-ins.
Natural decline rates
https://iea.blob.core.windows.net/assets/0edbecab-acf7-4701-bcb4-3fa9927787fa/TheImplicationsofOilandGasFieldDeclineRates.pdf
It is worth posting again.
Natural decline rates without investment would amount to 5.5 mb/d each year. This decline rate has been increasing for decades. In the early 1950, the decline was about half a million barrels per day each year. One month of drilling would find enough oil to replace that loss. The fifties and sixties saw the highest rate of discoveries averaging over 40 billion barrels each year.
Today it takes 11 months of drilling and investment to make up for the decline rate. Global oil discoveries are a fraction of the 1960s high and decline rates in the tight oil sector are crazy.
If natural decline rates continue on the decades long trend then it will be less then 8 years before global decline sets in.
I believe water scarcity in Iran will lead to a revolution and during that time Iran production will decline considerably.
My guess is peak oil in less then 3 years.
1. You have to be careful about the discovery timing stuff. A lot of that sort of ends up backdating. For instance, if a field is discovered in the 40s…and with 40s technology has ERR of X…but with time, we learn how to do more secondary/tertiary recovery, the ERR might grow to 2X.
So…is the “extra X” counted as a discovery from when the field was found or from when the secondary recovery method was developed? Well…in those charts you look at, they backdate it. So…they actually “find oil” in fields that they already found. But if you were doing a Hubbert/Pratt style ERR calculation in the 40s (or 50s or whenever), you’d have had a much lower total resource. This is a very old story and is called “reserves growth”. It’s how, despite increasing rates of production along with “crappy discoveries since since the 60s”, we have actually had a quite strong (even at times growing) R/P ratio.
2. So base decline is higher…noted. It’s a long, long stated concern. Not news. I’ve heard it in shale gas a lot. Look at how fast the wells in Marcellus (or the overall field) decline. Way worse on a % basis than the tired legacy production of PA from 2005, for instance. Yet…we still managed to grow like crazy. Yes…even when the Patzek/Berman/Hughes crowd doubted it. Who cares if you have to add some rigs?
And for what it’s worth, rigs are actually way more efficient than in the past. The USA had 4000 oil and gas rigs running in the early 80s! Now we have what? 500? Less?
3. And how the heck do you get from some random, long, article about base decline to a global 3 year peak oil prediction? How do you know it wasn’t 2018? Or won’t be 2035? Or whatever. You just “feel it”. (And that’s actually not a bad rationale…just asking.) Oh…and were you able to predict 2028ish from 5 or 10 or 20 years ago? Like we had articles on base decline from TOD well before 2010.
Anon
When the Brent oil field was discovered in 1971 that is when you place the discovery date and amount. If you want to pee around with allocating reserve growth to each and every year they reappraised the amount of oil it actually had that’s up to you. It was not discovered 20 times.
Decline rate news! I never said it was news you twonk. The figures I recapped are to show that today $500 billion in investment is needed just to stand still.
In the North Sea billions is spent yet production is falling.
So your R/P point is meaningless. If production falls the number of years you can produce oil at an ever declining amount can increase. Just shows you don’t know the basics here.
Also my estimation is based on an appraisal of every oil producing country and from information deciding what it’s production could be going forward.
Between at best and worst scenario the median is the most likely.
DC method, decline rates, discovery rates and actual production are all aligned to a peak of around 2030 give or take 3 years.
Again do you know how many oil producing countries are in decline?
Let’s see your calculations or analysis
NO didn’t think so. Ha
If you insist on backdating reserve growth to show a higher discovery in the past, you will always have this phenomenon of seeing discoveries look like they’ve dropped recently. That’s fine if you like to keep score like that. But it also means the apparent drop in discoveries is not such an awful concern.
You didn’t answer how you got to three years.
Dennis, is the non-oil thread forum suspended?
We seem to be in a timeout lol
LeeG,
Yes.
Coffee:
Please have some skepticism about the new fission reactor hype. Been hearing that since the 70s with Jerry Pournelle. Yet…where the heck is my pebble bed.
P.s. Read this article. https://ecolo.org/documents/documents_in_english/Rickover.pdf
The “paper reactor” hype goes back to the 50s at least…don’t be Charlie kicking the football.
Thanks for Rickover letter. There are many regret quotes from his deathbed. How many carbon based political units know that ionizing radiation is from unstable radioisotope atoms? Radnet was censored post Fukushima. A doomsday extinction event if we are not mature enough to isolate life wreaking isotopes from the biosphere. 100% political issue – not a technical problem. The creator made risk with the reward.
anonymous
The Germans built a pebble bed thorium reactor in the 80s. They could never get it to work, and it got shut down after a few years.
https://en.wikipedia.org/wiki/THTR-300
Thanks Ali. Nice article.
I don’t really keep track of the “new fission” type reactor space that closely.
Seems in some way similar to the story of the USN sodium cooled reactor.
https://www.ans.org/news/article-1999/seawolf-tries-sodium/
They got it to work. In an operating ship. But it had too many problems and they scrapped it.
Rickover said it might work on shore, but was too non-robust for at sea use. There is a funny story that he supposedly said that if we had a planet of liquid sodium seas, some damned fool would want to try a water cooled reactor! 😉
———
Obviously 300MW is a pretty decent size. They call it a prototype, but that’s a decent sized prototype. I mean…enough to supply a small city.
It’s a little sad to hear that the FRG operated it so cavalierly. Especially given it had more danger, just in being more unknown. Kinda makes me wonder how badly they operate their conventional PWR reactors! Homer Simpson eating donuts and shimming rods. 😉
Rig Report for the Week Ending September 19
The rig count drop that started in early April when 450 rigs were operating added rigs again this week.
– US Hz oil rigs added 2 to 369, down 81 since April 2025 when it was 450. The rig count is down 18.0% since April. The two rigs were added in Colorado while the main LTO basins were unchanged.
– Both New Mexico and Texas rigs were unchanged at 79 and 189 respectively.
– Texas Permian was unchanged at 150. Midland and Martin were unchanged at 20 and 17 respectively.
– In New Mexico Eddy dropped 1 to 33 while Lea added 1 to 46.
– Eagle Ford was unchanged at 33.
– NG Hz rigs were unchanged at 104.
A Rig
Frac Spread Report for the Week Ending September 19
The frac spread count rose by 5 to 174. It is also down 62 from one year ago and down by 41 spreads since March 28.
A Farc
Anon
I did answer, I said I looked at every country and from information made an assessment of possible min and max production. Adding all the median’s value together gets about 85/87mb/d which will be around 2028-2030. However other things are at play one of which I mentioned.
This website is concerned with global oil production and every country in it.
What is your claim exactly? Will there be a peak of production?
When do you think it will be?
Or do you just laugh at people who were wrong but too gutless to stand by any kind of prediction of your own.
OK. (Still don’t understand how you get an equation out of that. Or how you deal with the ability to add rigs. But…at least you said something.)
Huh?
Yes.
Don’t know.
Personally, I think that one of the big problems with peak oil as a social phenomenon has been amateurs wanting to puff themselves up and pontificate with predictions. And not acknowledging uncertainty. So…no…I don’t have a prediction. First step of avoiding Dunning Kreuger is to be honest about what you don’t know.
That is not to say there is no value in guesses made on partial knowledge. I can see value in that. But with a lot less of the bluster we got from the peak oil crowd. Art Berman on the steps of the DOE. George Kaplan laughing at EIA on GOM. Etc. Yeah…I think that sort of stuff needs a strong corrective. You can’t learn how to be better until you learn how you were wrong (including how you indulged your biases, used flawed methods etc.) So…yeah…absolutely…faces need to be held to crow bowls.
Anon
If you bothered to do a bit of study, instead of spending all your time saying nothing of any value you may then write something worth reading.
If you bothered to study which countries are now in decline and when, the amount spent on drilling etc. If you knew anything about oil discovery trend. You would realise that, as in the North Sea, as in Australia, as in the entire Asia Pacific area and others, that after a certain point no matter how much drilling is undertaken decline cannot be prevented.
So far you have added zero to this website, a leach on people’s time and effort.
I’m old enough to remember when the US was post peak.
https://econbrowser.com/archives/2007/05/peak_oil_in_ame
And I remember Hamilton saying that once a region has peaked, new peaks are lower.
https://www.youtube.com/watch?v=D9z1Ikn84Aw
Hmm. How’d that work out for US oil production? Or Texas? Or North Dakota? Or Ohio? Or the App combined (NY+OH+PA+WV)? Or South Dakota? Or Colorado? Or Utah? [All have had new peaks.]
Heck…you can even see pre-fracking, a textbook example. Canada overall had risen above its old peak.
And the peakers will whine that that came from new technology and exploration. No duh. That’s what the cornies warned you about.
When you think decades in the future, you have to contemplate that we will have new technologies and discoveries. That’s sort of been the history of the industry. That is why Hubbert, Deffeyes, that guy in 1919 Tractor magazine, Campbell, Ace, etc. have gotten so much egg on their faces.
Sure…oil is a fossil fuel. Wow…what a middlebrow insight. Everyone in the industry knows that. But the history of the industry has been one of depletion fighting technology. We use up the stuff we know. And we find new methods of discovery and extraction.
This is why the ERR estimates are so silly. If you are going to look 50 years in the future, you also need to posit continued developments in geophysics and petroleum engineering. Do you know exactly what they are? No. But assuming “zero”, given the time scales involved, as well as the history of the industry is butt dumb.
I have zero clue what will happen in the next 50 years (and will be dead then). But…if you want to pontificate like Dennis with these sort of Hubbert-plus (TRR based estimates), you should do more than just take USGS (or Laherre or Rystad or whatever) and just shade them low. The uncertainty is much higher. I warned Dennis of this 10 years ago. I mean…I freaking warned him.
You need to do something like half the consensus industry estimate and 3X of it…to have a real idea of TRR. That encompasses the real uncertainty. Just look how egg on the face the 2013 era peak oiler LTO predictions were (and this was years after shale was a popular thing…not like some 2003 prediction, where it was still a gleam in daddy’s eye).
The problem is…that those kind of radical high possibilities are VERY unpopular on peak oiler chitchat websites. What people want are Ace graphs. They LOOOOVE that sort of thing. Heck, Dennis gets crap all the time for not being pessimistic enough. (And it makes me laugh when he responds with an even more shaded down scenario in response…like at least stand for something…don’t just draw more spaghetti graphs.)
One thing that ought to give the doomers amongst you a LOT of pause is the vast amount of shale in the rest of the world. Shale is the most common sedimentary rock. All the conventional oil had source rocks. There is a massive amount of shale in ROW. US isn’t even the biggest source rock country.
Is iLTO hard to develop outside of the US (and Canada) right now? Sure. Don’t have the same legal structures. Don’t have the same service industries. And many of the places (e.g. within The Kingdom) are places where the incentives are to restrict production of conventional…not to flood the market and crush prices.
But when you look on the timeframe of many decades? I am much more convinced by geological constraints from the Peak Oilers. The whole fossil fuel argument–who wants to wait another 100 million years for new conversion of bottom goop to Texas Tea? But legal and business constraints? Those are a lot more likely to be moveable. Especially when you think about decades down the line. Heck…look how people used to laugh at Asian manufacturers in the 60s. Who’s laughing now?
—
Heck…are you really so confident that 50-100 years from now, someone can’t figure out how to do synfuel style projects economically (the whole Green River, Carter-era hype). I mean…there is carbon and hydrogen there. We need the right chemistry and engineering to pull it out…economically. But this isn’t transmuting gold. Heck…we figured out how to get LTO out economically. We figured out how to get tar sands oil out.
I would be afraid…very afraid, tree lovers. That’s a lot of time for technology to develop…and on a not that hard problem (we can already do it uneconomically…just gotta move down the cost curve…but that’s chemical engineering…it’s not like discovery a room temperature superconductor).
I mean…look how solar cells or batteries…or rockets…have evolved over time. Do we get there right away? No. Do we take a break for 20 years sometimes? Yes. But…on the time frame of 50-100 years? That’s a lot of time for people to get better at things.
Nony,
Producing oil from kerogen is certainly possible, but from a physics point of view, it cannot be done with positive net energy, we do not have the option of bending the laws of physics to our will, time doesn’t matter in that case, whether it is 100 or one million years. That my friend is a fact.
Are you arguing that oil will never peak? If so I am not convinced.
So you are think of a high case for World C+C of 6000 Gb? I am highly skeptical of such an estimate (current TRR is roughly 1500 Gb, times 3 is 4500 with 1500 Gb cumulative). My guess is in the 2500 Gb to 3500 Gb range for 99% CI with the 95% CI from 2800 to 3300 Gb for World C plus C.
I think it likely that demand will be more of a problem than supply in any case. I no longer foresee high oil prices (more than $100/bo in 2025$ in the future). Competition from solar and batteries will reduce demand for oil and natural gas significantly over time.
On US tight oil the chart linked below(second link) has the EIA’s AEO 2013 estimate for reference case and the low and high resource cases. So everyone except you had this wrong in 2013.
Chart from link below
https://www.eia.gov/outlooks/aeo/data/browser/#/?id=14-AEO2013®ion=0-0&cases=ref2013~lowresource~highresource&start=2010&end=2040&f=A&linechart=~~~ref2013-d102312a.10-14-AEO2013~lowresource-d012813a.10-14-AEO2013~highresource-d021413a.10-14-AEO2013&sourcekey=0
chart(107)
I have no clue if it is 6000 or 4500 or 3000 or 4000 or what. I do know that the history of picking a TRR (e.g. Hubbert) has been flawed. We tend to estimate low, based on time now thinking, versus what we are capable of 50-100 years from now. Easy examples of improvements to date include:
1. Improved geophysics (seismic).
2. Improved secondary recovery.
3. Extraction of gas (later of oil) from source rocks, via fracked multi-staged, horizontal wells.
4. Extraction of oil from tar sands (by two different methods, one in situ, one with mining).
There is a vast history of Hubbert having too low of a TRR. Of Piccolo/Verwimp/Dennis Coyne (for LTO). Does it mean it’s unlimited? No. But danger, Will Robinson. Make sure you have a higher uncertainty boundary. I’m still aghast at you standing in front of a poster at AGU which had just been 2X invalidated in ultimate resource, a month before you presented. Given 50-100 years…and a planet full of area and depth, you don’t think we’ll find anything to make a bottoms up ultimate resource budget be low?
I’m absolutely not “pushing” the Carter Synfuel program. My point is that on the scale of 50 to 100 years from now, it’s a lot more likely something will be possible than the converse. Appalachian gas shales were a 1970s era DOE hope as well. Took a long time, but eventually worked out. Same could happen for Synfuels. Or maybe it’s something else the peakers didn’t count on and that I’m not thinking of (a Rumsfeldian unknown unknown). Like how LTO rocked their world and crushed their ASPO/TOD group fads.
If you have to put energy in, big deal. E.g. if moving from an ~1:1 carbon to hydrogen ratio, to 1:2, I can just consume natural gas as my source of hydrogen. In terms of energy, I could just burn (i.e. fully oxidize) some fraction of the kerogen (or final product). As well as some fraction of the methane. This is really not that different conceptually from what goes on with tar sands.
The bigger issue is not some kind of (false) worry about enthalpy. The issue is how much it costs to build the pots and kettles and mineshafts and conveyers, versus value of what you get coming out. I mean…electrical generation plants consume more energy than they produce. But we want the electricity more than the input fuel. Similarly LNG plants consume more natty than the output liquified gas. But since you can’t put gaseous natty on a ship, liquified is more valuable. And we just roll.
But again…I am far from pushing Synfuels as some current government program. Was happy when Reagan killed that Carter sweater-by-the-fire 1970s program (after OPEC fractured in the early 80s and oil prices dropped). I’m just saying that trying to predict we will be same technically 100 years from now is silly. The history of mining, oil production, and chemical plant operation is one of massive improvements over the last 100 years. It would be silly to think further improvements will be “zero” on a time frame of half to a full century.
Heck, look how much better we’ve gotten at mining lithium. Look how we found a synthetic substitute for cryolite (flux for aluminum refining) after the Greenland ore was depleted.
“EIA’s AEO 2013 estimate for reference case and the low and high resource cases. So everyone except you had this wrong in 2013.”
You’re actually making my point for me. The uncertainty was on the high side. You only want to take EIA or USGS or Rystad or whatever and shade it low. True uncertainty can be massive (was definitely massive then, as we learned). I’d actually have been fine with a set of scenarios that included something super pessimistic. But you also needed a much higher upside. Not these tiny shaded different scenarios.
If we look globally out to 2200, than you need to consider a planet full of area and depth. All kinds of different resources. And a lot of time for innovation and the like to occur. And no…I don’t believe in abiotic oil. And yes, I do think it takes ~100 MM years to produce commercial hydrocarbons from organic sediments. Just danger, danger, danger…on thinking you have it all figured out, bottoms up, from now.
And it wasn’t about me being right. I didn’t even make a prediction. Just wanted you to expand your window of uncertainty.
Heck…it wasn’t just production history you gaffed up. I still remember you would have models that showed price ramping to ~$180 or the like. That worked out pretty hella bad.
And it’s not about guessing what dice will show at the craps table. It’s about thinking about the problems and thinking about your methods….and thinking about your biases. There’s a reason why TOD made their Ace predictions…and why they are tits up in the ditch now.
Nony,
The USGS released the Permian Delaware Assessment on Dec 6, 2018
https://pubs.usgs.gov/publication/fs20183073
The 2018 AGU Fall Meeting started on Dec 10 and the poster had to be submitted to the printer to be picked up at the meeting by December 7, I had adjusted my scenarios to account for the new information at the event.
You are creating a false history.
By December 7, 2018 I had updated my scenarios see
https://peakoilbarrel.com/open-thread-petroleum-december-6-2018/#comment-660548
and comments that follow.
revised scenario for mean TRR Permian estimate when adjusted for ERR at link below, posted on December 7, 2018
https://peakoilbarrel.com/wp-content/uploads/2018/12/660549-1.png
The oil price scenarios were based on the EIA’s low, reference and high oil price scenarios of the day. Often they were wrong, predictions about the future often are incorrect. There is much uncertainty about future prices which could be between zero and infinity.
I thought before you said poster was already printed with the wrong numbers. (Still not clear if that was actually the case, that the poster had wrong numbers on it. Are you explaining why it happened or saying it didn’t happen?)
If the numbers were wrong…then did you draw by hand on the poster (like with a sharpie) and correct it? Or clarify somehow with a sheet you handed out (of correction) to participants.
Or were you able to reprint the poster with a correct version?
So…even if you weren’t in front of a crap poster, it still remains that you had a massive 2X change in TRR happen, in late 2018. Not 2013. 2018. That ought to give you pause. You were ready to go out with something that flipped 2X on a dime. What if it happened in JAN2019? Your method is dependent on USGS, which has a history of occasional sudden (in a day, but every few years) radical changes. And at least in late 2018 (before their adjustment), had still not “converged” to relative stability. And you were ready to use it…despite it not having “converged”. You hadn’t run some check (like waiting for two in a row estimates) to see that they had things figured out, relatively well.
And this goes to what I’ve always thought was a flaw in your work. A lack of consideration of the resource estimate itself and what it’s level of uncertainty was. That’s actually the more interesting and important thing than how you draw the lines of production over time through some “shock” mumbo jumbo. I actually have more value for Pratt/Hubbert/Laherre or their modern equivalents trying to figure out the TRR…than your “shock” pretensions. I mean they end up being wrong also…but at least they are trying to understand the resource.
Nony,
The wrong chart was published one day before it needed to be submitted (not enough time to redo the analysis and meet deadline for poster printing), new scenarios printed on paper to show those interested.
There were no previous estimates for tight oil resource in Permian, there was a midland basin Wolfcamp estimate in 2016, a Spraberry estimate in 2017 and a Delaware estimate in 2018, the only information I had, in November of 2018 there was no Delaware estimate so I guessed based on relative output levels from the two sub-basins at that time. My guess was too low, when I had more information I adjusted the model. I do not have access to the geological data that the USGS, EIA, and oil companies have. The databases are expensive.
I notice when I use your words to create a TRR, you walk it back and become much less confident. So you don’t like the 3 times TRR leading to about 6000 Gb for TRR for a high case and start throwing around random numbers. Your previous 1/2 TRR to 3 times TRR would be 2250 Gb to 6000 Gb, I agree URR will likely (99.9999999% probability) be between those numbers. The shock model adds past rates of development and assumptions about future rates along with past extraction rates and discovery rates to estimate future production, future rates of discovery, devolopment, and extraction are of course are unknown. Those future rates can only be guesses. I agree there is much uncertainty.
I’m glad that you handed every person who viewed the poster a correction. I think a red marker handwritten correction on the poster would have been even better. Or just withdrawing it entirely and having fun walking around the poster session. But I’m satisfied with the printout corrections you had. As well as looking them in the eye and telling the story of what happened. (It’s not just the number itself that is interesting…it’s the dependency of estimates that is interesting about that story.)
Of course the timing was very unfortunate. But it remains that it could have even happened a few days AFTER the meeting.
The points isn’t even that you were right or wrong. And sure, estimates can change over time. The point was the EXTREME dependency of your work on USGS. At a minimum, you should have noted how it was prone to revision. Described the uncertainty. After all previous revisions had happened.
Yes…I know you can’t do your own geological compilations. But. It then means that your work is strongly dependent on USGS, to the extent where I really just think of it as “warmed over USGS”. It’s incredibly strongly correlated. Not a new/different view really. And that’s sorta whether you shade the USGS down (from economic considerations) or shade it up (for technology development), so long as the changes are moderate.
It’s not at all clear to me that other people in the industry like Rystad or EIA are just taking USGS as baseline. They probably have some level of independence to think differently. They may even share my view (the horrors!) that USGS is conservative, leading to underestimates and growth over time.
Doesn’t mean they are right. But means, they are contributing true alternate views. It’s not like Moses came down and said use USGS (for one thing, he’d have to come down when revisions happened!) Or maybe midstream companies building expensive pipelines are using the same “take USGS as the base” method for justifying investment.
As far as the .5-3X. uncertainty..yeah…it’s freaking huge. I’d actually be comfortable with a narrower band now. Or even some method where you graph how the USGS has changed over time and note some sort of “convergence”. You can actually make a decent case there, for the Bakken and Marcellus, now. But couldn’t in 2012. Umm….and you sorta needed that crazy wide uncertainty bar in late 2018. Cause the estimate came out an whipslammed you with a 2x change.
It’s just something to think about. So…for example, the Fayetteville might be the extreme end of “known”, where there’s not even any drilling any more. Something like the Bakken (relatively old for a shale play) might be relatively converged, although not to Fayetteville levels. Obviously the Permian had NOT converged yet and had extreme uncertainty in late 2018. It’s not enough to just say I take USGS as the gold standard…you need to think about how gold the gold is. And a model in NOV2018, using USGS as a base…wasn’t very golden at all. So…for instance, you really should have floor to ceiling error bars for the Utica for instance (some fraction under the most pessimistic and over the most cornie…yes, even if it looks crazy wide and you don’t like high numbers…those are the error bars).
Nony,
Yes estimates can change, I agree. I use the information available to me and do the best I am able. I am not an expert, nor have I ever claimed as much. As to humility, my impression is that I am far more humble than you.
The shock model includes quite a bit of uncertainty and uses maximum entropy probability distributions as its underlying basis along with some basic physics and mathematics.
Below is a cumulative probability distribution using a negative exponential PDF for remaining World oil resources with the assumption that 2P reserves (about 730 Gb) will be produced so that at least 2300 Gb will be produced (zero probability that World URR for C plus C will be less than 2300 Gb). For this chart the mean remaining resource (above 2300 Gb) is 800 Gb with a standard deviation of 800 Gb (note that this is not a normal distribution.)
maxent urr
1. “The shock model includes quite a bit of uncertainty and uses maximum entropy probability distributions as its underlying basis along with some basic physics and mathematics.”
This is my issue with the whole fetish of the “shock” model. We are not talking about a simple closed system in some physical chemistry course problem. What makes you think that all the entropy and stat mech blablablaBLA is anything other than technical mumbo jumbo when applied to a time series of production (open system, economic factors, technical development, time series economics, etc. etc.) It sorta reminds me of people talking a bunch of quantum mechanics terms or relativistic terms or the like, for an engineering problem that has nothing to do with them…and this isn’t really an engineering problem. It’s a complex future forecast with lots of open issues in economics and technology development…not even a sandstone reservoir fluid flow problem.
1.5. I actually have more value for the general DC approach of “take USGS as a base”, add some price deck, add some economic hurdles for a haircut, and then make a model (of whatever general kind) that shows some pimple with area under the curve equal to the shaved USGS number.
I mean there’s significant uncertainty about the area under the curve itself (look at the late 2018 Permian 2X jump). But the “how it plays out” line graph? It could be a faster or slower or different shaped pimple that gets the area under the curve. No good way to predict that. So…adding a bunch of 3-5 semiconductor physicist mathium (not really even applicable to the scale/nature of the problem)? That’s actually reducing how much I value the DC stuff. If you just got rid of “shock” math stuff and made it some “USGS plus economics” model, I’d like it BETTER. I.e. more Dennis and less stat mech.
2. You’re right…in a sense, I’m less humble. In that I retain more uncertainty. I’m far enough up the Dunning Kreuger to know that there’s a bunch I don’t know. If I knew less, I’d be more certain!
A said
Someone who is obviously frightened by math. The basis for the shock model is a fundamental mathematical formulation called convolution. Go take a look through all the research literature on petroleum engineering and anything to do with extraction and you will find virtually nothing on how to apply convolution**. Yet it works to describe what’s happening. Perhaps the best example is in applying convolution to describing how a cumulative extraction profile of a population of fracked wells plays out. It basically says there is one average extraction response which is *convolved* with a series of impulse events spread over time.
I first introduced this at The Oil Drum http://theoildrum.com/node/10163 in 2013
It does track the progress exceedingly well even though it is all based on a typical or average response. How effectively it works in the future is a separate issue, but the main point is that it does work in describing the actual behavior, which is the proper response to calling it “technical mumbo jumbo”.
** (as an aside, do the same with convolution and AI, and you will find millions of hits. Convolution is a fundamental tool in neural nets and LLM)
bakken_conv
Nony,
If you think statistics, mathematics, and physics are not applicable to how the future might play out, then there is not much need for further discussion. It has very little to do with statistical mechanics except the very fact that understanding how many small molecular interactions can be shown to play out in a macro sense as in classical thermodynamics, the same principles are very likely to apply in other contexts. Not seeing that shows very little imagination on your part.
How it plays out in the future is of course unknown, this is so obvious I leave it unstated in some cases, but assume the audience is intelligent enough to understand the obvious.
I do very often state there is much uncertainty, note for World C plus C URR the 90% CI for the probability CDF I presented earlier is 2340 Gb to 4700 Gb which is a fair amount of uncertainty. Though not the 2250 Gb to 6000 Gb you suggested earlier, which would be close to the 98% CI for my CDF (2310 to 6000 Gb). Also the 50% CI (2530 Gb to 3410 Gb) is similar to low and high shock model scenarios I have presented in the past.
The 2X jump in Permian ERR was simply an artifact of limited information and a poor guess by me. Note that the USGS Bakken/Three Forks estimates for TRR (when adding proved reserves and cumulative production to undiscovered TRR) have been pretty consistent over estimates in 2008, 2013, and 2019. We have not seen any updates to Permian estimates, but perhaps that may prove to be true there as well. If the estimates change, I will adjust based on updated information.
Yawn. It’s stat mech voodoo. I throw the BS flag. Minus 15 yards and a loss of down.
It doesn’t even apply to what we are doing here. Buncha fancy words, not grounded in a physical rationale. If Terrell Hill were still alive, he’d tell you the same.
It’s like trying to use the Schroedinger Equation for Fantasy Football analysis. Sorry…that equation i designed to tell you how the electrons dance around the nucleus (and other similar problems). But it has nothing to do with helping me decide which injured running back I should start that week. Yes…even if you blather about Slater orbitals and tell me go take a density functional theory class so I can draw spaghetti graphs about the GaAs.
Those things are applicable to THAT PROBLEM. Not to oil production, NFL strategy, meeting women, or many, many, many other topics! Shock “blabla” is NOT APPLICABLE.
You can’t just throw some fancy math and physics and assume that’s making something better, if it’s just fancy words and hard equations that are not lined up versus the problem.
And statistics? I’m fine with that. Applied analytically and properly to specific problems. Keep doing linear regressions. But drop the stat mech entropy talk and the PDEs. That’s silly. If anything it shows LESS understanding (of these admittedly hard topic) to misapply them in this manner. Again, more Dennis and less physicist mumbo jumbo.
Nony,
See
https://openlibrary.org/books/OL33718680M/Mathematical_Geoenergy
I am not particularly convinced by the mumbo jumbo claim, just because you don’t get it doesn’t mean it’s bogus.
Also try
https://en.wikipedia.org/wiki/Principle_of_maximum_entropy
from Wikipedia on maxent
In ordinary language, the principle of maximum entropy can be said to express a claim of epistemic modesty, or of maximum ignorance. The selected distribution is the one that makes the least claim to being informed beyond the stated prior data, that is to say the one that admits the most ignorance beyond the stated prior data.
Uck.
You’re asking me to slog through a buncha math in a (very specialty) anthology style academic book.* It’s not self publishing…but almost. I can imagine how (non) difficult the peer review was there. Couldn’t publish it in a good physics or geosciences journal. Phys Review would laugh at you. Economic Geology would laugh at you. Science or Nature would laugh at you. They would all tell you it’s voodoo science and they’re not even sending it out. You are way too proud of what is a miserable academic publishing record for shock theory of oil prediction.
And you don’t understand that math either, Dennis. Be real. It’s like me telling you to decipher something in 5th century Coptic. It’s not relevant and it’s just a game to try to obscure things.
The Wiki article…ugg. Look at the (2008!) box at the top saying artice needs improvement. That thing is a mess.
*The evil side of me has to ask. Did you include the Permian predictions in that 2018 book? And were you able to get the 2X jumped value in there or if I pull it am I going to see the vastly too low numbers?
Google Scholar shows 5 citations for that monograph, Dennis.
https://scholar.google.com/scholar?hl=en&as_sdt=0%2C47&q=mathematical+geoenergy+%3A+discovery%2C+depletion%2C+and+renewal&btnG=
That’s not exactly setting the world on fire, Dennis. A “textbook” should have a massive amount of citations. Five is very tiny. Heck, even just for an article, five is small.
It’s actually four, if we don’t include the one self citation. But I’m actually in favor of self citation. Shows you are publishing, helps link your works together for a reader. If anything the amount of self citation is too low also! 😉
Now in contrast, this GaAs surface science article (where the math, physics, and some experiments are right on track), has over 400 citations. Now that’s an outstanding reception for a paper!
https://scholar.google.com/scholar?oi=bibs&hl=en&cites=15307566675427048235&as_sdt=5
5 << 400.
GaAs = good.
Peak oil = bad.
Nony,
The model is based on math and physics, critique the model all else is BS. You can get the book at a University Library.
I don’t know enough math or physics to disprove Blackrock hydrinos either. Nor do I choose to even try. But I notice that Science and Nature aren’t publishing hydrino articles.
I think the point that your model has not transformed the industry or academic fields based on the lack of people citing it (and how it was published in a monograph versus a good journal…i.e. lower peer review hurdle) is a relevant one, for me not wanting to weed through the spinach.
I kinda held back on twisting the knife. You have a day job. And I appreciate amateur analysis. But you have occasionally patted self on the back for this massive academic publishing feat (and presenting a poster at a huge meeting). It’s just not mainstream academic success. It’s not crank science…but it’s a notch or two above it.
Again, the GaAs paper was a rock that had a lot more ripples. A NOVEL and USEFUL technique ought to have a buttload of citations. Instead people have ignored “shock model”, the same way I have.
And…again, the baseload estimate is what really matters. Like when your Permian estimate doubled in late 2018.
Or like Paul needing to crab back with an after the prediction was made caveat about world C&C (&NGL for one model!) doesn’t include LTO (and NGLs from shale).
The point is that ultimate resource is the key driver, key uncertainty. And the math/physics and abnormal pop science-y usage of terms from chemical thermodynamics is…window dressing. It’s snowing people with something irrelevant and overcomplicated. The guts of your predictions are the resource estimates.
The anonymous commenter is actually hitting a sore spot among Earth scientists — their obvious lack of curiosity when it comes to any new ideas. This thread is not the ideal place for this, but the reason I started to look at models of oil depletion was because of my realization that no one was applying the idea of convolution to resource extraction. I know because I did the literature search at the time. Noticed also that Anonymous mentioned Sam Foucher, who got interested in the oil shock model because he had a background in signal processing and knew the techniques of convolution inside and out. BTW, just looked and Sam’s top paper is “Convolutional Neural Networks for the Automatic Identification of Plant Diseases” (419 citations published in 2019). So it’s kind of obvious in retrospect that only outsiders to Earth scientists and geology were going to look at these approaches.
My interest in geosciences after that snowballed because I got involved in a multimillion $$ research project in modeling of environment factors, suitable for subsequent engineering design. In that project I got paid to find out how easy it was to pick out various natural behaviors in an environment that could use a revised mathematical treatment. That turned into a final report that was open-sourced to the public, and decided to publish that as Mathematical Geoenergy, which took a few years to finalize, teaming up with Dennis to improve the contents.
Yet, still not a lot of interest after a few years. Earth scientists are not like other scientists — they have zero knowledge of how to do controlled experiments, mainly because they can’t (!) and so don’t have much intuition for understanding how things actually work or in ability to finding patterns in data. There’s lots of other stuff in the book that supercedes that of the oil shock model, such as modeling of El Ninos, QBO, geophysics stuff such as Chandller wobble, all researched because Earth scientist, climate scientists, and geophysicists essentially know squat about how to do signal processing.
So yes, the Anonymous commenter is free to dream up fantastical ideas of what our motivations are, but it comes down to enjoying trying to solve challenging puzzles, which is the backend of natural human curiosity. Don’t speak for him, but Dennis also seems to enjoy putting all the numbers and pieces of evidence together. That’s all there is to it.
Nony,
I am not a part of the academy, nor have I claimed to be. I think it is useful to have an underlying model to make clear what the assumptions are that one is making about the future. I agree the resource estimate will determine the area under the curve, the shape of the curve in the future can be modelled using the shock model for World output. The discovery curve is convolved with a negative exponential PDF for various development stages (fallow, construction, and maturation), a final extraction stage enables “shocks” such as the oil embargo in 73, Iran Iraq war in 82-84 and the pandemic to be applied. Different assumptions for future discovery (and a consequent change in the underlying equation) or mean times spent in fallow, build or maturation stage can be tested and fit to the historical data and then projected into the future.
For tight oil the model is far simpler with an average well profile convolved with completion rates, future completion rates and future well profiles can only be guessed, and again I agree the resource estimates will determine the area under the curve, but many different curves could be drawn and creating a model based on a set of assumptions seems better than drawing random lines (which seems to be what you believe I am doing, your wrong there.)
Paul,
“The anonymous commenter is actually hitting a sore spot among Earth scientists — their obvious lack of curiosity when it comes to any new ideas.”
-and-
“Yet, still not a lot of interest after a few years. Earth scientists are not like other scientists…in ability [sic] to finding patterns in data.”
——
You know what this sounds like, Paul? Classic crank science behavior. Except you are not the good guy. You’re the bad guy. The professionals aren’t listening to you, so you blame them, not yourself.
And FWIW, geophysics has a long history of signal processing, physics models, and pattern finding in SEISMIC ANALYSIS.
https://en.wikipedia.org/wiki/Reflection_seismology
You wanna make some money and get some attention from Halliburton? Figure out some seismic testing or analysis improvement.
another anonymous dude said:
That’s probably why I had this section in the book.
Stand by my take. If they were actually curious, they’d be paying more attention to spectral analysis and other interesting ideas from signal processing. No worries though, as machine learning will find all this stuff out. And as it is in the world of LLM, it doesn’t matter what others think. An LLM can already reason about the oil shock model and do all the computations on its own. AI algorithms don’t worry about what is crank science and whether it needs to do gatekeeping to protect the interests of a narrow niche of researchers.
https://www.youtube.com/watch?v=Dk0HCJvCgrI
5 minutes
I’m pulling this forward from last thread.
Zeihan (worked for Stratfor a military think tank … needs to go back to previous haircut )
Is saying, the Ukraine has the ability to strike Russia pipeline infrastructure and is being encouraged by Europe to do it!!!!!
I know nothing about pipelines, but I would suspect they are hard to repair in Siberia in the winter?????
For the time being the Ukrainians are mostly attacking the refineries. I don’t see how that would reduce Russia’s crude oil exports, and his explanation seemed muddled and incorrect — but I am not an oil man. Maybe even crude oil exports have to pass through a refinery as well?
The Ukrainians are focusing more on refineries, and I suspect they are in the hopes of slowing the invasion by cutting off its fuel supply. Attacking Russia’s finances by reducing crude oil exports makes strategic sense, I guess, but seems less urgent.
Ukrainians attack oil export facilities also.
“The most effective sanctions – the sanctions that work the fastest – are firing on Russian oil plants, on their terminals, on their oil depots,” Zelenskiy said. “They have significantly restricted the Russian oil industry, and this significantly limits the war.”
“Zelenskiy singled out an attack last week on Russia’s northwestern port of Primorsk for the first time which he said had inflicted “significant damage. This has been checked.”
https://www.reuters.com/business/energy/ukraine-attacks-major-russian-refinery-with-drone-assault-2025-09-14/
Primorsk had to limit oil exports because of these attacks. Primorks’ crude oil export capacity is about 1 million bopd.
Alimbiquated
In Saudi Arabia they have Gas oil separation plants (GOSP) to remove the gas and water that comes with the raw crude. In some cases low levels of metals have to be removed.
I don’t know what happens in Russia. These special plants are required to upgrade the crude to a marketable product and they could be part of a refinery or a separate plant. In either case, bombing those special plants would reduce the crude exports
In the recent past railway data was showing that targeted refineries were supplying Russian military.
Documents are also available- for example, Kstovo Refinery had a contract to provide jet fuel to FSB for one of its aviation units in 2020
I feel that limiting military capacity is likely prioritized over attacking export capacity.
Things may have changed since recently. I may have seen faulty railway data.
Russia exports both crude oil and refined products and they probably get more income from the refined products so attacking those reduces income for the war machine. In 2024 Russia exported 2175 million tonnes of petroleum products, the sanctions mostly have targeted crude oil and Russia has been selling crude at a discount to India and China and others through the dark fleet, but at reduced net income. In March 2025 about 46% of Russian income from crude plus petroleum products was from the products.
See https://energyandcleanair.org/march-2025-monthly-analysis-of-russian-fossil-fuel-exports-and-sanctions/
Are the products mostly sold in Europe, or at least “close”? Traditionally, crude is for long distance markets. Products is for local. Smaller ships. Lower shelf life. Seasonality of the markets themselves (“summer gasoline”). I would expect them to send crude to India/China, not products.
It’s probably a lot easier to take out refineries than upstream assets. In general refineries are more complicated and expensive and concentrated than upstream or even midstream assets.
Ovi:
I think you are confusing some different words here. Gasoil (one word) is a “cut” of the distillation tower at the front of a refinery. Part of downstream. It is not a mixture of methane and crude and water. It is a high molecular weight fraction of the crude. When people talk about hitting parts of the refinery (I assume they mean the cracker), that is a very different thing from gas-oil-water separation of upstream crude.
Almost all the methane and water from ground production of crude is separated in the upstream. This is done by three phase separators, which are almost always at the pad/platform level. These are dispersed all over the place and are not high value, large volume pieces of kit. Much smaller than refineries and refinery kit.
Here is a video explaining how to separate gas, oil, and water at the pad level.
https://www.youtube.com/watch?v=hzwJStZmI_o
Essentially, just looking at a salad dressing cruet and seeing the water at the bottom, oil on the top, and vapor [in this case air, but would be methane and some NGLs in a hydrocarbon…but to get the idea]) will give you the basic concept.
But again…three phase separators are WAY cheaper and smaller than refinery components. Especially in Russia, which has a very dispersed oil industry. Not (in general) massive integrated upstream facilities like the Saudis do at Ghawar.
Nony
Why are you always dancing on a pinhead. I am not confusing different words here. Read it for yourself.
“The key production region for Saudi Aramco is the Southern Area, which is divided into North Ghawar and South Ghawar. Throughout this region are numerous gas oil separation plants (GOSPs), which provide the initial separation of water and gas from the oil prior to shipment to Abqaiq for further processing.”
https://www.yokogawa.com/library/resources/references/saudi-aramco-southern-area-gas-oil-separation-plant-control-system-upgrade-project/
Ovi:
I’m responding to this sentence of yours:
“These special plants are required to upgrade the crude to a marketable product and they could be part of a refinery or a separate plant.”
Gas-oil-water separation (whether dispersed as in Russia or the US or even at some big megaproject in SA) is part of UPstream. It is not part of a refinery. Certainly not a Russian refinery. Refineries are DOWNstream.
Note also that the mixed oil-water-vapor coming out of oil wells is NOT CRUDE. “Crude” is the oil fraction (the liquid hydrocarbon fraction, the “middle” phase) AFTER you leave the 3 phase separator.
Three phase separators do NOT separate “crude” into water, oil, and gas. They separate the mixed fluid (which contains bubbles of three phases, two liquid, and one gaseous) into water, oil and gas. The “oil” coming OUT of the three phase separator is the “crude’. Note that it is normal for more than 50% of the volume of the mixed well fluid to be water.
Simply put gas-oil separators (or three phase separators) do NOT process crude. Crude comes OUT of them. It does not go IN to them.
—
Note also the term “marketable product” also appears to show some confusion from you. The term “product” or “marketable product” is almost always used for the output of refineries. I.e. motor gasoline, diesel, jet, etc. So when we talk about a “product tanker”, we are talking about a smaller ship (with nice clean tanks) that moves motor gasoline or diesel or jet. We are not talking about crude tankers that are much larger and do not have to have clean tanks (a little residual crude from the last shipment is no big deal…but would be a disaster for moving low sulfur gasoline or diesel or jet).
——-
Clearly, you don’t understand the difference between field separation of oil/water/gas (from well fluids in the UPstream), with how refineries work and what they do and what they produce. (I.e. the DOWNstream).
—
P.s. It’s OK to have gaps in knowledge. We all do. But please…use this process to learn things. Don’t just remain in ignorance. Your confusion of upsteam and downsteam separation is very similar to the error I sometimes see of people confusing (drilling) RIGS and oil WELLS (or pumpjacks). Don’t get mad…learn.
Ovi (and others):
Oil refining at a very simple (short, non technical) level:
https://www.youtube.com/watch?v=EI3OM7ok8cI
Note…this is part of DOWNstream. This is not the 3 phase separator that sits in upsteam, on individual pads in the Permian. Refineries are massive billion dollar level integrated “chemical plants”. They are a completely different world than upstream dispersed on the thousands of pads all over the Permian (much smaller facilities, with less $ in them, less complex engineering, etc.)
The 3 phase separator (that does 99% of the extraction of water and methane from the mixed fluid coming of an oil well) is a cheap piece of kit at individual pads.
Going after three phase separators makes zero sense for Ukranians. It’s like hitting a fly with a Tomahawk missile. Different parts of a refinery make way more sense. We can argue if you should hit the tower or the cracker or the coker–all are important, expensive pieces of kit. But they are NOT 3 phase separators. They are not there for the purpose of doing bulk (or even “polishing”) level removal of methane or water. They exist to take crude and differentiate it into useful products…ideally maximizing the “middle” products of gasoline/diesel/jet, which are most valuable.
I see that the oracle for oily matters has once again spoken. It is a pity that you are not as learned as you think you are. Have you ever seen a GOSP, worked a drilling rig, worked in a refinery? I doubt it.
All you do all day long is watch Youtube and ask AI for your answers and then try an convince everyone you are an expert. Do you actually have a job of work? A piece of advice. AI is not very good at technical matters, and even worse with history.
KSA has a a rather different set up to the US. The number of producing oil wells is a fraction – like 1%- of the wells in the US. That means the GOSP’s and gas plants are generally larger than the US. Try looking into the Saudi Master Gas System. Much of the oil production is done in an area a fraction of the Permian basin. You will also know, if you have done your homework that Ghawar is made of of several producing zones that differ considerably in porosity and permeability. Look into Haradh III which is produced with multi-lateral completions and an intensive water injection capability.
When looking to disable a refinery the last unit to destroy is the coker. The crude distillation unit would be the prime target. It is big and runs the entire refinery. Disable the CDU’s and you take out the refinery. Even the FCC would not be a prime target as it is mainly a gasoline machine and it does not produce jet or diesel, which are essential for war mongering.
Refinery
This isn’t about me being an oracle. And absolutely, I get info from YT/Google/web. And am well aware that it comes with limitations. It’s actually not even a who’s oilier dick size contest. It’s about learning.
Crude does NOT go into a 3 phase separator. It is one of the three streams coming out of it.
And three phase separators are a standard part of upstream, not a piece of refinery kit.
The discussion of SA (and specifically Ghawar) was a diversion from the original issue, which was “where in the Russian refineries should the Ukrainian missiles target”. And FWIW, yes, I hella agree that Ghawar is a much richer oil source with massive bopd wells, different from West Texas. I think I even mentioned something about that. But…it still doesn’t make a massie 3 phase separator part of a “refinery”. It’s not. And it doesn’t change that crude comes out of the separator, not goes into it. (It’s mostly water going into it!)
The tower would absolutely have been my first guess also. (Although I’m open to some debate about the FCC or the coker…those are very expensive pieces of kit and key bottlenecks that the linear programming model tries to optimize.) But the beginning of the discussion was if they should target other areas. They were already targeting the towers. It was a discussion to think about if they should go after something else. And my point was that that discussion has nothing to do with upstream 3 phase separators! It just doesn’t. It’s like calling a well a rig.
https://mishtalk.com/economics/california-tries-to-stop-the-exodus-of-its-oil-refiners-oops-too-late/
California tries to halt exodus of oil refiners from the state.
Sorta reminds me of JenniG (Biden DOE Secy). She talked all kinds of bad stuff (and implemented regulations) when the prices were low. Then when prices were high, she begged the US industry to drill more. Well…you can’t turn that ship on a dime. The oil and gas industry looks at political risk when making multiyear, multibillion dollar decisions. Look at the massive amounts that companies lost on Constitution Pipeline or on leasing rights for shale in New York State.
Oil companies know good looking Gavin hates their guts. He just pivots (momentarily) when it looks bad. Like Harris…she was anti fracking in the Democratic 2020 primary…then pivoted, when she had to for political reasons (like trying to win PA in 2024). At the end of the day, companies see this and know that they are still in danger from Donks. Not so much danger they stop operating. But enough so they increase their risk threshold in financial models.
Enjoy those high gasoline prices, Californians. And I hope the Republicans beat the sh&% out of that issue in 2028 when we go Vance v. Newsome.
Nony
Nah. Trump Jr vs Newsome
Oil and gas prices are in dead slow spiral and Trump is desparately trying to cutting out Russian supplies to boost price by encouraging Ukrain to attack Russian oil and gas infrastructures.
I’m not able to post anything in the non-oil forum – there is no option to reply.
rgds
WP
I think the “reply” only works with respect to a reply to first level (i.e.. at second level).
So, if you want to reply to a reply…don’t…just reply to the first in the chain.
Note, there is also the option to just reply the the thread overall (at the very highest level). Helpful when things get too tangled.
—
[Edit] Never mind. I don’t see any option to reply in any way within that post! 🙁
I decided to shut down the Non-Petroleum thread.
Since people use the “it’s all connected” excuse to bring in the electric cars and the nuclear reactors anyways? Slippery sloping to Trump and stock market and organic gardening.
Dennis in light of the incessant political comments I think that that was a wise decision. I, and many others were tired of the the extreme comments by some.
DC, I can’t blame you. The site will be more focused, and I’ll choose to do more reading than commenting. Keep up your great efforts.
Thanks Mike B,
Your comment helped me to see the light. It had become more of a heat source and provided little light.
I hope no one is congratulating themselves or being congratulated for the improved readability, commentability, or functionality of this website.
I know Dennis works hard, thank you Dennis.
Every aspect of visiting, reading, and commenting is far far worse than it has ever been. I realize some help would be appreciated and money is tight.
But please don’t pretend that things have improved by dropping the non-oil thread and the new formatting.
Apparently Ukraine has been targeting the Crude Distillation Unit ELOU-AT-4 with its attacks on Russian refineries.
This is what I have on that:
Unit ELOU-AT-4 integrates the following sections: electric desalting, crude rectification, naphtha stabilization and re-distillation.
Also they have attacked a blending station where high-sulfer and low-sulfer crude is mixed to create Urals crude.
btw the Russian President recently said that the long term goal for Russian exports of crude is to drop the sale price by $1 per year. Its not clear to me if that is a $1 drop on the contract price, or if it is simply the effect of inflation on locked in long term contracts with adjustment mechanisms that don’t take inflation into account.
The Russian Federations published long term strategy is to continue on the path to becoming a diversified, self reliant, technically advanced economy with low debt and multiple state/private capital income streams.
The objective is to reduce the reliance on oil income (gas not so much). This makes sense as fading oil production means you must make big changes before major fields play out or be caught unprepared.
One of the long stated aims is also to help keep oil prices at a modest band that allows major economies to grow. In this way demand for gas (especially) is maintained. Major long term contracts allow Russia to do it its way. China and India benefit ‘bigly’ from cheap oil, whether crude or, to some extent, refined product.
But the Russian crude processed in India, China and elsewhere can be sold at a good price on world markets. At market prices. To the benefit of the refiner/reseller (India, China and elsewhere). And of course Russia does have some investment in foreign refineries, so they do pick up some income from foreign investment.
“It’s remarkable how Republicans in Congress acquire wisdom after they have announced that they’re not seeking re-election.”
First past the post voting and unbridled corruption(aka money id free speech) combined with sharper gerrymandering tools look like they are going to doom America’s limping democracy.
Novi Labs on Permian inventory remaining:
https://novilabs.com/resources/webinar-permian-staying-power-inventory-now-and-through-the-2040s/
About 20 years of Tier 1/2 at current rate of activity. Note…this does not conflict with the foreshadowed decline, as current activity is already down 20% from last year.
They don’t go all the way to a production outlook, but I’d assume something like a very slow rate of decline (dwindling plateau). But…no…not some sudden “they’ve used it all up” that makes rigs lie down cause there’s nothing left to drill.
Nony,
The Novi labs economic assumptions are quite dubious in my view. Note that the Permian scenarios in the post are based on the Novi tier one tier two estimates, with the medium scenario assuming all tier one and tier 2 inventory is drilled and the low scenario assuming only tier one and the high scenario assuming tier one, two and three. Much depends on prices of oil, NGL and natural gas as well as the ability to deal with increasing water issues in West Texas which will drive up costs and reduce profits in the long run.
The mighty Dennis Coyne has spoken. Nobody should even look at the alternative. 😉
For me, it’s not (just) about some neener, neener “I like this number because it’s big”. Or the converse for some Laherre estimate or the like.
It’s about taking a step back and being humble about future uncertainty. At least keep an open mind and realize that someone (not a complete idiot and with some experience…yes, still not an “oracle”) came up with this estimate.
Yeah…I could see OPEX going up, with the water an all. All that said, there’s a hella lot of old, very small (0-10 bopd), stripper wells running…and doing 10:1 water to oil. It has to get really bad, before it makes sense to shut a well in. I would be a little leery of too much peak oiler “hopium/copium” with the water worries. Kinda reminds me of all the blather about 15 year well lives (I think I remember hearing 10 year worries in the past!) And we can see places like the Barnett or Bakken clearly heading to 20+, even now. I’m not even making an argument of what WILL happen. Just saying, be careful, be open, be humble, retain some uncertainty. I.e. it’s possible you are right…but no reason to think you WILL be right or even that it’s PROBABLE that “Dennis Coyne the blog operator and 2018 AGU poster presenter” will be more right than the industry analysts. That’s why my easy view of window of uncertainty is something a fraction less than the most doomer prediciton and a fraction over the most corny prediction.
Also, even if the “numbers are too big…me hates them” sets you off, try to watch the presentation and get something out of it regardless. For instance the way of disaggregating multifactorial causes (geology, completion, pressure reduction from competition, etc.) Or the Delaware versus Midland. Or the operator insights. I.e. don’t be so obsessed on the peaker perspective that you don’t learn something. If all I got from your stuff (or Berman or whoever like that), it would be a complete waste. But I try to extract some insight/learning about the whole thing regardless. That’s why I look at what Ovi writes.
Nony,
I listened to the presentation, the scenarios are my take on what might happen, they are certain to be incorrect.
As far as 20 year Bakken wells, looking at Novi for pre 2006 wells there were 263 producing wells as of Jan 2006 and 157 producing wells in Jan 2023, so after 17 years about 40% of wells were no longer producing.
https://public.tableau.com/shared/BFPWGTT8Q?:toolbar=n&:display_count=n&:origin=viz_share_link&:embed=y
As usual, you’re counting idle wells rather than plugged ones. It ain’t over until the fat lady sings. If they ain’t plugged, they’re still available. In general, what happens is they cycle the wells. (Leasing and government regulations don’t generally allow more than 12 months of being idle.) So they run part of the year. So…that’s still an operating well.
We’ve had this discussion before.
Nony,
Also note that horizontal tight oil wells completed today are significantly different from wells completed before 2006, whether current wells will last 15 or twenty years is not known. Whether downhole repairs needed after many years will be profitable when old wells are producing perhaps 8 bopd will depend on water cut and oil prices. Experienced oil men suggest most wells needing downhole repairs after 15 years will likely be plugged, especially at $65/bo in 2025 $. Some operators will try to avoid plugging liabilities and leave wells inactive for as long as possible. If a well pumps one day out of 365, perhaps it is an “active” well, but is essentially no different than a well that has been plugged. There were 58 of pre-2006 Bakken horizontal wells plugged out of 300 horizontal tight oil wells completed by December 2005 or roughly 19%.
DC,
Given your prior interest in gas supply pipelines into NY with their possible impact on New England…
Williams is moving forward again with:
1. “…the Northeast Supply Enhancement Project, a key 23-mile underwater segment from New Jersey to New York City that would run parallel to its existing Transco pipeline and then connect to it to provide gas for state markets.”
https://www.enr.com/articles/60831-developer-confirms-revival-of-northeast-gas-line-projects-totaling-2b
And,…
2. “Williams remains committed to advancing the Constitution Pipeline project and has submitted permit applications,…”
https://ctmirror.org/2025/09/15/constitution-pipeline-natural-gas-new-york-pennsylvania-ct/
Williams site for Constitution shows a timeline with a target of April 2028 to be in service. No coincidence there.
https://www.williams.com/expansion-project/constitution-pipeline/
It was wise for you to close the other thread. I’ll miss what I learned there, but tempers were rising and at least one post appeared to be a direct threat to you personally.
We need a lot more oil and nat gas (data centers, export to allies) and not a dwindling plateau.
0. I “need” a pony. Doesn’t mean I’m getting one. Might have to do without. (Biking to work or shopping or errands is a great thing…don’t have to even always do it…just mix it in partially and it makes a difference…and gets you fit.)
1. I wouldn’t be too worried about natty. There’s several decades of the stuff, especially as you move to lower grade rocks. Lot of shale on shale competition has occurred. Ass gas is cheaper than Marcellus. Marcellus is cheaper than Haynesville or Deep Utica. (But bottlenecked by lack of transport out of basin.) And then plays like the Barnett have kinda turned off because of price point (drilled up a lot…but definitely not completely).
It’s kind of a resource pyramid for natty. And there’s a lot of lower tier inventory out there. If the price goes up, it can come on line. Or even if production costs drop (even small increases in efficiency over the course of many years). There’s really a lot more natty out there than oil. Always has been. Just not as premium a product as oil (Look at the massive difference in price/BOE versus oil.)
[And then even if we did run short of natty…the US is the “Saudi Arabia of coal” (tm Jimmy Carter)…and we could run a lot of electrical generation on coal if we wanted to.]
2. Oil? Donno. I sorta think we’ve hit the major LTO shale plays. Always a possibility of something new, given the time scales. But my gut feel is there is no “next” major shale play out there. Not a sizeable one (1 MM bopd plus, Bakken/EF sized). We’ll keep learning how to squeeze the grapefruit harder…but that has limits.
There’s probably still some major potential in AK (vast area) and in the VACAPEs (vast area). But we’re not allowed to explore/drill/lease there. Could change if things got dicey (price high enough to drive changes in the political hurdles).
Realistically, I think future world growth in oil volume has to be supplied by the Middle East. If they are willing. (I actually do think they have the oil…so rich out there.) But they will get more market power as shale declines and ROW (non OPEC+) declines. So their ability to dictate price will increase. And they will likely choose to drive price up, rather than volume.
There is also the dynamic of getting Russia to effectively “join OPEC” (de facto, not de jure). It’s hugely important to having a good cartel, to have them at the table, hobnobbing with the Saudis, versus wrecking prices.) Looking years/decades into the future, I think you have to posit the danger (if you are the Saudis) of “peace breaking out”, which will lead to lower restrictions on Russian exports and on Russian development/collaboration with the international majors (or by themselves). And the Saudis want to keep a lid on that jar. There’s a vast expanse of territory in Russia. It ain’t the Persian Gulf. But it’s still vast…like Canada and US together in ultimate potential, perhaps. And still containing areas that are underexplored, especially in the East and North of Siberia and offshore (granted in difficult environments). All of which says…if you are a Saudi prince, you want to keep Russia in the cartel.
Don’t worry Australia
Anon says in 50 years time new technology will be there to boost your oil production.
In the meantime for normal people Australia oil production peaked in 2000AD and after 25 years of improvement in technology is down to less than a quarter.
Joining 160 oil importing countries. And that list has been growing for 40 years.
Never say never. Look what happened with shale gas and LTO in NAM. Would have been crazy talk in 2005 to imagine the shale boom. But…it happened.
50 years is a long time…and Ozstralia is big. And even if Oz doesn’t develop, maybe Angola will. Or they’ll find natty in the Eastern Med (oops already happened). Or whatever. World is big. Time is big.
Be a happy grasshopper and enjoy the summer. The ants will turn to, when needed. Somehow, somewhere. They have a long history of doing so. Look at all the failed past peaker predictions!
Besides…oil is really a global product. Import the oil and export the coal and diamonds. Not the end of the world to be importing some natural resources while exporting others. It’s impossible for every country to be self sufficient in every mineral.
Anon
That is so simplistic I hardly know where to start.
Let’s take the U.K. as an example. Back in the 1990s when it was exporting a million barrels per day. That vast amount of money coming in supported all sorts of things. There were plenty of dentists. Doctor appointments were easy to get and hospitals had reasonable waiting lists.
Today millions can’t get an NHS dentist because of funding, hospital waiting list and A and E are horrendous.
Taxes are higher than ever. When the U.K. had plenty of gas electricity bills were very reasonable. Today the electricity cost from wind is crippling industry many shutting down.
So your view is not based on reality as usual.
Shall I tell you what happened in Syria when oil production started to fall sharply.
Cheap energy is the basis of all prosperity and for many countries it is running out.
“All true wealth comes from the ground.”
I actually sort of half agree with you. I’m not as dogmatic as the Berman/Martensen “energy is the economy” crowd. But…I do think the “service economy” neoliberal hypers are underestimating the importance of energy and mining and manufacturing. They think you can just do a bunch of websites and financial derivatives and that will feed the kids. And we saw how that worked out with the dotcom and Wall Street crashes of 2000 and 2008.
So…I guess I’m in the squishy middle. To the right of the AI/ML/dotcom hypers. But to the left of the peak oilers. (On some imaginary spectrum of virtual versus real.)
For the technology champions.
https://ourworldindata.org/energy-mix
Oil has provided more additional energy over the last 25 years than the additional wind, wave, solar, nuclear and hydro combined.
Coal has provided more additional energy than wind, wave, solar, nuclear and hydro combined.
Gas alone has provided double the additional energy of all those others combined.
India has confirmed it intends to increase its use of the coal it has to provide cheap and reliable energy to bring the Indian people out of poverty. They intend to hit a target of 1.4 billion tonnes of coal per year.
https://www.bbc.co.uk/news/articles/cpd184g6dj8o
Iver,
True, but that’s just primary energy, not energy services. Oil is a lousy source of energy. It’s much too expensive to compete with coal, so it is only used to generate electricity in a few niche markets, especially islands.
Oil is mostly used as a method for storing energy in a moving vehicle. It’s very handy for that, but hopelessly inefficient. If you spend a buck on gas, you’re spending 80 cents to heat your radiator, and another few cents to drag it through the airstream to cool it back down.
So there’s plenty of oil sloshing around, but it’s mostly just producing waste heat, not actually doing anything useful.
Alim
We all get your point, but the most efficient ICE can drive five people on a hundred mile round trip for the cost of 5 cups of coffee.
Or it can take 600 people from New York to Europe in 8 hours at a fuel cost of £400 per person. Nothing else can do that and never will.
The first link is very dense and seems quite good, parsing it.
Second link is a bit more simple.
It is interesting to note how there seems to be a very Western bias (even an unconcious bias) to thinking of Clean Coal in the context of CO2 emmissions only. For countries like China or India, it’s probably much more important to them to think of reducing smog type pollutants (via scrubbers and filters at power plants, and via elimination of home heating with coal furnaces).
I have seen a lot written about how China is trying (or even has accomplished) cleaning up some of their city smog, by “cleaner” burning of coal in the old sense (reducing NOX/SOX/metals, not carbon capture).
1. I don’t think I ever read or commented in the non-petroleum threads. It’s possible I did, during the whole history of the blog…but I can’t remember doing so. Didn’t even lurk/read. Just ignored them. Was I missing out? I figured there was enough drama in the O&G threads…why borrow trouble and look for more!? (Or did I miss some amazing insights and high IQ debates?)
2. I’m a little worried the “don’t want to debate O&G analytics, want to talk about the Orange Man” hoi polloi will now move into the grownups section of the blog, Dennis. Wasn’t that the whole purpose of having a “kiddy table/penalty box” for them?
The non-petroleum threads had quite a bit on climate change, which is more than peripherally related to oil depletion. The back-and-forth on this topic is what is actually a bigger driver for alternative energy adoption — climate change or peak oil? I’d say 1/2 and 1/2 — 1/2 climate change for marketing and 1/2 peak oil driving the bottom-line. And everyone knows talking peak oil for marketing is a non-no, like mentioning Fight Club.
Paul,
Are you saying that the most intelligent creature ever on planet earth can not be told that finite resources will peak and decline, and the only hope for respite from massive atmospheric pollution and the adoption of alternative energy sources is appealing to AGW?
Interesting article on refinery outlook for next 20 years.
https://peakoil.com/business/more-than-20-of-global-refining-capacity-at-risk-of-closure
They see continued shrinking of the total refinery capacity. I sort of don’t worry about this. It’s a conversion industry. And has been (in some ways) shrinking for years now. There’s still a lot of weak sisters out there (crappy local refineries). And some areas where countries have a refinery for “energy security” reasons, not for economic ones. Sort of like how every country has to have an airline, even though the world doesn’t need an Austrian Airlines or a Jordanian Airline or the like.
The chart shows total margin (profit) for the industry overall. It actually tends to drop when there are short periods of too many refineries (see 2022-2024 part of chart) versus when there are periods of low capacity. It;s not as extreme boom or bust as oil production. But still…they do better when refining supply is temporarily constrained, as the benefit in price is much more substantial than the percent drop in volume.
BTW, haven’t looked at those PO.com comment sections in a while. What a wasteland of AI spam! Thanks for keeping those monsters away here, Dennis.
Anonymous
(I don’t even know you good enough to make sense of why you are referenced as “Anon”. )
You are actually right about quite a few points regarding the refinery industry. It is a hard industry to make ends work.
I am more interested in the value of mathematical and statistical models for oil and gas. To dismiss the work as not essential is a little bit dishonest to me. What I can say is “bullshit in gets you bullshit out” output. The model is only as good as the input variables allow for. And that is the root of the problem in my opinion. There are several expansionist arguments on the table like technology development in itself or transfer of technology to newer or frontier regions. Some friendly or most likely unfriendly regions around the world are simply landlocked, not explored enough or has not been/ are not as of now accessible. I still rate efforts to make sense of output (graphs) given inputs. Dennis & Co have had to continuously delay peak oil visually due to more promising data input. I have no problems with that, and I am still curious of how the scenarios look like given data revisions from IEA, EIA, DOE and OPEC.
Thanks for the friendly comment.
FWIW, I was quite amused by the talk page comment on the back of the Wiki article on Maximum Entropy:
“Heh. Be aware, there was a big dispute earlier this year on WP; apparently, many chemists utterly dislike the way that mathematicians and info-theorists have hijacked entropy. So now we have Entropy (general concept), Entropy (information theory), Entropy (statistical thermodynamics), and many others, and the cultural divide seems to be too great to leap. Similarly, on the statistics side, there is a large body of work that sticks to “their” definition, and rather dislikes the intrusion of physics into conversations. linas (talk) 22:16, 3 September 2008 (UTC)”
I haven’t even been a part of that kerfuffle, but I totally GET how the traditionalists don’t like the hijacking. 😉 Just feels like trying to use fancy words and equations to dress something up.
And yeah…I really think the INPUT (i.e. USGS TRR) is way, way, way more important than the shock equation (whatever that equation is) versus some other mechanism for converting the TRR into a time-based line graph.
So talking about the input more (how stable is it, has it been in the past…how would output differ with a larger TRR from an alternate provider, etc.) seems much more useful than telling people to go take a stat mech course.
The Maximum Entropy concept is simple: If all you know is the mean, the best estimate for the standard deviation is to equate that to the mean. It’s essentially reasoning within the context of uncertainty. There are physical analogs to this with respect to the physics discipline of statistical mechanics.
I don’t see it on my Mollier chart:
https://en.wikipedia.org/wiki/Enthalpy%E2%80%93entropy_chart#/media/File:HS-Wasserdampf_engl.png
If you want to make some statistical Bayesian argument, make it. And use the stats words too. I don’t think the chemists OR the statisticians want you mucking around calling stuff entropy. It’s too pop-physics-y.
ND production essentially flat (up .02MM bopd):
https://x.com/staunovo/status/1970529906934403246
I didn’t bother getting on the MS Teams call, so will wait for the recording, for any “color”.
My poor little Bakken…still not up to 1.2 MM bopd. And winter is lurking!
North Dakota July Oil Production
Production rose by 2 kb/d to 1,161 kb/d. Down 169 kb/d from September 2023. Currently the average well is producing 61 b/d, down from 105 b/d in December 2014.
ND Oil Production Chart
A ND
Dennis
I can’t post a chart. It is a 50 k jpg file
Ovi,
Try another browser maybe?
Only links to charts will post. You could also try reducing file size by using either png or gif.
Not clear why there is a problem. You could also email it to me and I could try.
Hmm…reading that 2013 TOD link is interesting.
I kinda blew off reading the math/physics blabla (linked in a tiny text article) that was too hard for me. Guilty! Ooh…la…la.
BUT! I did look at how well old “shockie” did on predictions.
http://theoildrum.com/files/PU2013046.png
Looks like both shock and hybrid shock (HSM) got their butts kicked by shale. I mean…they had a bunch of fancy math that nobody followed (and probably doesn’t apply to the problem). But it didn’t drive good predictions.
Shockie said we should be doing 65 for C&C+NGLs, right now. In actuality we are over 100.
HSM said C&C would be less than 65 (low 60s) by eyeballing an extension off the chart) for C&C right now. In actuality, we are in the low 80s.
I haven’t checked every alternate method (there’s even some in other figures, that’s just a fraction of the total methods). but I don’t see anything special for shock or hybrid shock versus other methods in terms of performance. All the bottoms up methods did miserable also…peak oilers getting slapped by observed reality.
The BAU ones actually did decent. Were in the right direction at least.
I guess this explains why TOD and ASPO shut down. Shame, shame, shame!
shock model from 2018, I was using a higher estimate at that time for extra heavy oil (oil sands), today I have revised this to 200 Gb, Also tight oil estimate is somewhat lower at about 80 Gb today (from 100 Gb in 2018). Peak for this model is 2024 at 84.6 Mb/d, Covid pandemic was not foreseen.
shock 2018 c
The shifting sands of resource accounting. We always knew that the definition of oil would change to allow for poliitics. I wrote in 2013
Prompt for LLM: Need an accurate assessment of (in the narrow context of) global conventional crude oil production and what the peak level was in terms of daily production in millions of barrels.
Copilot: 70–74 million barrels per day is the best estimate for the peak of conventional crude oil production, reached around 2005–2006.
DeepSeek: The peak level of global conventional crude oil production was approximately 74.0 – 74.5 million barrels per day (mb/d). This peak occurred not as a single event, but as a “rolling plateau” that lasted from around 2005 to 2011.
ChatGPT: If you want a single, defensible number for global conventional crude oil (narrow sense): ~70 million barrels per day, around 2006 (IEA WEO 2010).
Claude: According to the most authoritative sources, conventional crude oil production peaked in 2006, with an all-time maximum of 70 million barrels per day WikipediaEncyclopedia Britannica. This assessment comes from the International Energy Agency’s (IEA) World Energy Outlook 2010
Grok: Global conventional crude oil production—referring to oil extracted from traditional onshore and shallow-water offshore fields using established drilling and pumping methods, excluding unconventional sources like shale/tight oil, tar sands, and heavy oils—reached its historical peak in 2005. That year, production hit approximately 73 million barrels per day (mb/d).
Gemini: Global production of conventional crude oil reached its peak in the mid-2000s, marking a significant turning point in the history of energy production. While pinning down a single definitive daily production number is challenging due to varying definitions of “conventional,” data suggests a plateauing of this specific type of crude oil at around 74-75 million barrels per day (MMb/d) between 2005 and 2008.
Perplexity:The peak daily production of global conventional crude oil was reached around the mid-2000s, with estimates converging on the year 2005 or 2006 and a peak level of approximately 70–75 million barrels per day. Since then, increases in global oil supply have been driven primarily by unconventional sources such as U.S. tight oil (shale) and Canadian oil sands.
—
That is what various LLMs say. So what happened with all those early projections that we did assuming only conventional crude oil, is that they always appeared to immediately drop from the peak. In retrospect, that happened because the truth was that conventional crude oil production had peaked and we were trying to squeeze blood from a turnip. Yet, now without a consistent metric for oil, the number is at a higher range, which is totally understandable with the unconventional sources of oil being included.
This is all from being objective with regards to the data we were dealing with at the time.
1. Sam Foucher was evaluating your model from several years before. Not from when you added the 2013 caveat.
2. FWIW, Bakken crude is almost indistinguishable from benchmark WTI. Permian LTO literally IS WTI. And has even been incorporated into the Brent futures “crude basket”. In terms of economics, it is smack in the C&C mainstream. An “economic substitute”.
Reminds me of the finding that the Iliad and the Odyssey were not by Homer but by another blind Archaic Greek poet with the same name.
Asking different AI engines the same question is only valuable if the AI engines are using different sources. Here, the engines are summarizing IEA data or other documents based on IEA data.
Is there a question you can ask that requires the AI to put together data from say, four different sources, each independent of each other? That will tell you which AI has the biggest training data set.
When i do that, i get a wide variety of answers, some of which are physically impossible, and others are clearly wrong, and a couple are close to being right.
That tailoring the question also makes it clear that the AI can’t reason, and any question asking it to do often fails to get you any further ahead.
Nony,
tight oil ramped up very quickly, it may decrease just as quickly, it is a very temporary solution to oil depletion and is now included in my shock model estimates along with extra heavy oil.
The chart below has the low, high, and my best guess scenario from a June 2019 post. The best guess scenario peaks in 2026 at 87.34 Mb/d and the low and high scenarios come from the 27 scenarios in the post linked below.
https://peakoilbarrel.com/oil-shock-model-scenarios-2/
None of the scenarios are likely to be correct ( a small sample of 27 out of infinity ).
shock2019
I think some of these anonymous comments can be classified as historical revisionism. The shock models were based on estimates of crude oil discoveries and the extrapolation of discovery tails. An anonymous commenter can make up stuff all day long
The bottom-line is that conventional crude oil never recovered, and that’s primarily because annual conventional discoveries continued to decrease, and secondarily reserve growth could not make up for that.
1. The Sam Foucher exercise (look at the graph and article that YOU linked) talks about your models from 2005 and 2007. So your 2013 caveat is actually the revisionism!
2. Even if you strip out “non conventional”, your model is still off, showing C&C ~62 versus what we have now at ~82. Unless you want to tell me that we have 20 MM bopd of non conventional.
3. The whole conventional, non-conventional is kind of an excuse. Who cares if it is non conventional if it supplies the market and benefits the economy. It means the “danger” peak oil is not a danger to people.
4. The industry has a long history of finding new oil and using new methods to do so.
5. LTO is actually rather “conventional”. More so than tar sands actually (which was what peak oilers in 2005 were worried about affecting their models). The formations are well known. Spraberry and Bakken had produced commercially since the 50s. And fracking and hz drilling were known technologies. The real change was the industry got better at doing them consistently and economically. But the oil is coming out of a well by pressure and later by submersible pump or rod lift. It’s hella more “conventional” than in situ or mined tar sands being steamed.
6. If you really wanted to give yourself a great caveat you could disallow new discoveries entirely. But then what’s the point of your global production predictions. Since they don’t include a normal mode of industry addition.
7. For that matter, in 2013, you hella SHOULD have been including LTO as a part of your predictions. Since it was a known factor with the potential to either underperform (as all the Berman, Rune, Coyne, Hughes types predicted) or the possibility of high impact (like the crazies like Leo Maggueri and Goldman Sachs and Scott Sheffield predicted…who by the way don’t look so crazy in retrospect!) But in any case, you absolutely should have made an attempt to include LTO.
8. And AGAIN…it’s the resource estimate the drives the result (good or bad, caveated or noncaeated). Not the shock and awe “entropy” pop physics window dressing.
A said:
That’s great. So i can just use solar energy as a replacement for conventional oil as it benefits the economy and is directly related to oil, since it provided the original energy source. Tom Hartmann referred to oil as Ancient Sunlight
Time to do a Peak Ancient Sunlight post, eh?
An interesting assessment of Saudi Arabia reserves
https://blog.gorozen.com/blog/saudi-arabias-oil-reserves
Thanks Iver.
I agree an interesting piece.
Many would claim that much of Saudi Arabia is unexplored, I an skeptical of these claims.
DC
A recent press release described how Saudi Arabia had discovered several new fields.
https://www.reuters.com/business/energy/saudi-arabia-discovers-14-oil-natural-gas-fields-state-news-agency-says-2025-04-09/
This tiny amount oil would not get a mention 20 years ago.
At the time of the audit release, Goehring and Rozencwajg had also published an article along these lines:
https://blog.gorozen.com/blog/analyzing-the-degolyer-macnaughton-audit-of-saudi-crude-oil-reserves
When one looks at the Aramco prospectus from 2019, the audit results also read 162Gb of oil proved reserves: https://www.aramco.com/-/media/images/investors/saudi-aramco-prospectus-en-051219.pdf (see page 646 of the pdf)
Hope springs eternal with the peak oilers. Here we are 20 years past Twilight in the Desert. What did Simmons predict? (I don’t have the book.) Wiki, citing TOD says this:
“…he raised the possibility that in the not too distant future, Saudi oil production will suddenly decline.”
https://en.wikipedia.org/wiki/Twilight_in_the_Desert
Well…here we are literally (and not how millennials use that word) 20 years past TITD publication. And Saudis are pumping fine. Actually up from 2005.
Now G&R have another Ace-like “about to crash” peak oiler graph. Don’t you guys ever learn?
Oh…and it’s Hubbert Linearization. HAHA!
https://www.youtube.com/watch?v=kDuLHEoS45s
This 2014 POB thread is kind of a hoot also.
https://peakoilbarrel.com/closer-look-saudi-arabia/
The actual top-post by Ron is not that bad. Mostly an explanation of how SA kept production flat (even slightly increased) over last 10 years since TITD (at that time).
But read the comments! Ah…so delightful.
1. Dennis Coyne predicting a lack of supply causing an economic crash in 2-10 years (i.e. 2016-2024), which would be followed within 5 years (i.e. 2016-2029) with a “Manhattan Project” style alternate energy effort. We haven’t had the supply shock* recession. And the Manhattan Project style windfarm thing ain’t happening either.
2. Pre-ShaleProfile Enno with some screwed up predictions and remarks. (He ended up upping his game significantly.) Kind of the one example of an amateur analyst from the peak oil community becoming a general (and quite good) oil supply analyst. But…he wasn’t showing that yet, in the comments on this thread.
*Note that Covid does NOT count since (a) it was a drop in demand, not in supply and prices crashed rather than skyrocketed, and (b) we’ve bounced back since.
From EIA’s International Energy Outlook 2011 we have the following for Saudi Arabia liquids output compared with EIA data. In 2024 the EIA’s IEO estimate was about 27% too high.
ksa 2011
Dennis:
Still doing way better than what the peak oilers predicted.
Also, look at the 2018 to 2021 run of numbers, followed by the 2022 to 2024 numbers. It is hard to think those sort of big changes are depletion driven. It was well reported in the press that they formally cut volume to prop prices. Remember the pandemic and low oil prices? Followed by the 2022 jump when Russia invaded Ukraine? Note the reversion back up from 2021 to 2022, rather than a continued “decline”.
Peak oilers keep dreaming, dreaming, dreaming about Saudi decline. They are doing fine.
Looks like a plateau since 2016 or so, a CTMA peak in 2016 at 10634 kb/d was surpassed in 2022 at 10649 kb/d, we will see what happens, no decline yet, I agree, it will be interesting to see what happens to the CTMA in the future, the most recent CTMA using EIA data is 9167 kbpd.
I agree the lower level in 2020-21 and 2023-25 was not due to depletion, but an attempt to increase oil prices. Lately they seem to have given up on that, but they may change course again if they find that OPEC demand estimates are a bit on the high side.
saudi ctma
Anon
Oil prices were $120 back 12/14 yers ago, why didn’t Saudi take advantage of such high oil prices?
Does it occur just maybe they can’t produce 12mb/d or anything close for a sustained period?
Maybe they can’t. Maybe they don’t want to. But they sure seem able to do ~20 years of flat or even slightly higher) production, since the mid 2000s worries. Since…uh, they did!
So, if you want to say neener, neener, they can’t do 12…again, maybe. Personally, I think they could have. After all if they did 20 years at 9-10, how hard would it have been to do a few years at 12?
But again, “inability to do 12” is a whole different kettle of fish from, SA is going to drop off a cliff and we are going to have to kiss our asses goodbye, since there’s nothing to replace it. (The whole Simmons/Staniford/TOD idea from mid 2000s.) Since we are literally 20 years down the pike. And the whole “Twilight in the Desert” DID NOT HAPPEN.
Oh…but don’t worry…G&R has a graph that would make TOD-Ace proud! [EDIT: Actually now that I look at it, it’s not…that bad. Look at the time scale on the chart. Nothing significant happening for next 10 years…no impact on near term investment decisions or opportunities. Production rates only down to what they were during Staniford’s 2007 worry wailing. I mean…things are going to really suck in 2187, 150 years from now. Then again, consider how different the world is now versus 1875!]
Nony,
Staniford said in March 2007 that the Saudis would never see sustained output above 10.7 Mbpd. If we define “sustained” as the 12 month average output, so far he has been correct.
A nice summary of Staniford’s thoughts on peak oil in August 2013 at link below
https://earlywarn.blogspot.com/2013/08/what-oil-drum-meant.html
Stuart passed away in October 2024
https://www.instagram.com/p/DBazeRdxCcp/
Have you ever read Simmons book?
His claim was that Saudi Arabia would not be producing 15 Mbpd of crude plus condensate for the next 50 years as was being claimed by cornucopians in 2004.
That claim was correct.
He also said they might not be able to maintain a plateau in output for many years and might start to decrease their output. These were not correct at least through 2025, though they were qualified as being lower probability rather than high probability predictions.
On 02MAR2007, Staniford said:
“Overall, I feel this data is clear enough that I’m willing to go out on a limb and conclude the following:
*Saudi Arabian oil production is now in decline.
*The decline rate during the first year is very high (8%), akin to decline rates in other places developed with modern horizontal drilling techniques such as the North Sea.
*Declines are rather unlikely to be arrested, and may well accelerate.”
How did that limb work out for him?
P.s. I actually told you all about Staniford passing away.
Here’s the Saudi record of production:
https://fred.stlouisfed.org/series/SAUNGDPMOMBD
From 2011 to present, it has been higher, every year, than it was in 2007.
So much for Staniford.
1. I already said upthread that I hadn’t read the Simmons book.
2. I did find this article about old Matt Simmons and the Saudi Chicken Littles:
https://reason.com/2017/02/21/remembering-peak-oil-saudi-arabian-produ/
‘Even so, Simmons’ main assertion in his book Twilight in the Desert was: “My research has convinced me it is unlikely that Saudi Arabia could sustain any higher oil output than it now produces, and that even the current production rate may be too high.”‘
And…SA in fact HAS averaged SA higher production, OVER THE LAST TWENTY YEARS NOW, than it did in 2003-2004. (See the St. Louis Fed curve.)
Nony,
The EIA has Saudi Arabia’s C plus C output at
https://www.eia.gov/international/data/world
In 2005 when the book was published, average annual Saudi C plus C output was 9.8 Mbpd and average output after that year has been 9.7 Mbpd, using monthly data Jan 2006 to May 2025.
The synopsis of the book is inaccurate, I have the book in my hand, reading from pages 336 to 339 of the conclusion.
Nony,
Yes Staniford was wrong about Saudi output continuing to decline, the OLS on monthly data from Jan 2006 to May 2025 has the average annual increase in Saudi output at 40 kb/d. With oil prices at record high levels in 2008 it was a bit surprising to some that Saudi output of C plus C declined at an average annual rate of 189 kb/d from Jan 2006 to December 2008 while oil prices were rising from $57/b at the end of 2005 to $132/b in mid 2008 for Brent crude.
Note that from Jan 2010 to May 2025 the average annual rate of increase in Saudi C plus C output has been an impressive 11 kbopd.
If we use EIA annual C plus C data for Saudi Arabia from 2011 to 2024 and use OLS to look at the trend we have an average annual decline rate of 25 kbopd. Longer term (2005 to 2024) there is an average annual increase of 39 kbopd. Average output from 2005 to 2024 was about 9700 kbpd with most years within 500 kbpd of the average. Pretty much a plateau.
One last note on Simmons. In 2005 he said a plateau might be maintained for 10 years and after that we may see Saudi output decline.
In fact this was not that far off as there was roughly a plateau (slight increase in output at annual rate of 100 kbpd) from 2005 to 2016 (peak Saudi annual output of 10.6 Mbpd). From 2016 to 2024 the average annual rate of decline in Saudi output has been 131 kbpd. A gradual rise and a gradual fall (basically flat output.)
Very long term (1975 to 2024) over 50 years the average annual rate of increase in Saudi C plus C output is 75 kbpd. So an average annual rate of decrease of 131 kbpd over past 9 years might be significant.
Peak year is 2022 (not 2016) at 10.64 Mb/d, 2016, 2018, and 2022 are all within 46 kb/d of each other (10634, 10598, and 10644) the sustainable capacity seems to be about 10644, it will be interesting to see if it will be surpassed.
Dennis, Saudi production averaged higher in the last ten years than in the ten years before that. (And it was significantly higher during the “possible plateau in next ten years” than it had been prior to it.)
Average production:
*From 1999 to 2004: 8.3 MM bopd
*From 2005-2014: 9.3 MM bopd
*From 2015-2024: 9.7 MM bopd
Source: https://fred.stlouisfed.org/series/SAUNGDPMOMBD
Source for data is EIA at
https://www.eia.gov/international/data/world/petroleum-and-other-liquids/annual-petroleum-and-other-liquids-production?pd=5&p=00000000000000000000000000000000002&u=0&f=A&v=mapbubble&a=-&i=none&vo=value&t=C&g=none&l=249–199&s=94694400000&e=1704067200000
for chart linked below.
Average output from 1995 to 2005 not all that relevant Simmons was asking if output could be maintained at the 2005 level, there was an increase from 2005 to 2014 and a bigger decrease over the next 10 years.
saudicc
I can’t get that data from EIA, for some reason, Dennis. Something about their site glitching. If you look at average production from 2005-2014 and compare it to 2015-2024, the second decade was higher than the first one. I think you will see that with the EIA data also.
You’re talking about comparing a time series to a plateau. And putting a lot of leverage on an outlier high year (2015) at the start of your trendline.
You’re comparing apples and oranges and bananas, Dennis. Just look at average production first ten years and average production second ten years.
Anon
I have an interview of Simmonds on DVD, he clearly says Saudi Arabia could produce 12 million barrels per day. He also said their are voices in the industry there that say they should never attempt to go above that as that would jeopardise their reservoir integrity.
He does not talk about falling off cliffs.
Do you know how much storage the Saudi’s have?
Even when they tried to destroy the tight oil industry they could not produce more than 10.6 in a twelve month period.
So maybe they are in decline from the 12 mb/d from when Simmonds spoke. They have never shown otherwise.
Here’s the history of Saudi oil production:
https://fred.stlouisfed.org/series/SAUNGDPMOMBD
It didn’t crash. Who cares if it wasn’t 12 MM bopd either?
Here we are 20 years after Twilight. And there was no collapse of our way of life. And oil is selling for $37 (2004 dollars), right now. Not the $200-$300+ that Simmons talked about. And the Saudis are pumping more now, than they did in 2004 also. So, what was there to worry about?
https://reason.com/2017/02/21/remembering-peak-oil-saudi-arabian-produ/
‘Even so, Simmons’ main assertion in his book Twilight in the Desert was: “My research has convinced me it is unlikely that Saudi Arabia could sustain any higher oil output than it now produces, and that even the current production rate may be too high.”‘
Nony,
Simmons said if we had to create the current (2004) oil system from scratch (all oil infrastructure including pipelines refineries, the entire system, that $200/b might be too low.
He also assumed demand would continue to grow, it hasn’t grown as fast as 1985 to 2004 over the 2004 to 2024 period and he did not anticipate the the rapid growth of oil sands and tight oil.
Here is World C plus C minus extra heavy oil (oil sands) minus US tight oil. Roughly a plateau from 2005 to 2019. In 2005 output was 72.32 Mbpd, in 2018 it was 72.82 Mbpd and peak was 2016 at 73.54 Mbpd, 2024 was 68.42 Mbpd, and 1985 was 53.72 MBpd.
world conv
I say again
Simmons’ in his book Twilight in the Desert:
“My research has convinced me it is unlikely that Saudi Arabia could sustain any higher oil output than it now produces, and that even the current production rate may be too high.”
—
He was wrong.
Dennis says:
“Simmons said if we had to create the current (2004) oil system from scratch (all oil infrastructure including pipelines refineries, the entire system, that $200/b might be too low.”
———
That’s kind of a bizarre “if” remark, Dennis. Like IF I were missing a leg, I’d take a lot longer to do my next triathlon.
Oh…and FWIW, Simmons actually made a wager that oil would be $200+ in 2010. And it was NOT dependent on “if” rebuilding all the infrastructure. It was just a straight prediction.
https://en.wikipedia.org/wiki/Simmons%E2%80%93Tierney_bet
Nony,
Simmons never said that in Twilight in the Desert, it was made up in the book review in Reason.
Simmons did say (based on his research of SPE papers on Saudi Oil Fields that some individual fields had been over produced with potential damage to the reservoirs. He also accurately stated that we have very little verified information about Saudi output. So what level of sustainable capacity Simmons believed existed in 2005 when the book was published is unclear.
In his preface he says the sustainable capacity was about 10 Mbpd, in 2016 output was marginally higher at 10.6 Mb/d, but his main point was that the 22 Mb/d oil output capacity for Saudi Arabia in 2025 projected in the EIA’s IEO 2004 was not likely to be correct. This indeed seems to be the case.
Personally I doubt the Saudis ever get above 12 Mbpd for annual output of C plus C before 2035. It might not ever occur, but forever is a long time so we will wait and see.
Dennis:
1. Yeah, I already gave you props for the quote and correcting the Reason author. Take a second victory lap. 😉 You were right. I was wrong.
2. I’m fine if EIA was off high and Simmons off low. But, I don’t think Simmons was only trying to counter EIA though. He was peddling high price and disaster to way off like stuff also. Remember the Tierney wager? The chatter about lifestyles declining and people needing to carpool?
3. This is minor, but I think there’s some confusion of “total liquids” versus C&C, also, with that old chatter. I think you (as do I!) prefer to talk in C&C. But it was sort of common to talk about total liquids (including NGLs) back in those olden days. Not huge difference, but FYI, if you’re citing specific numbers, good to check if they said “C&C” or if they said “oil” (and if they said “oil” you need to check what they meant with it).
4. It feels a little “tree in the forest sound” to talk about “capacity”. We can’t see it…so people can even now argue pessimistically and optimistically about “capacity”, above the actual production record. For one thing, how much time/investment are you allowing when saying “capacity”?
I sort of prefer to just look at production. That’s why Staniford fell off his tree limb…from a production forecast, not a “capacity” claim.
Nony,
I must not have been clear, you were right. The quote is in the preface of the 2006 edition of Twilight in the Desert, I have the 2005 hardcover which doesn’t have that quote. Your link from the Globalist clearly says it is from Twilight in the Desert copyright 2006. Maybe you don’t read the links you post?
Nony,
In the International Energy Outlook from 2004 (from when Simmons was writing his book) the EIA only project oil production capacity for Saudi Arabia (probably total liquids it is not very clear).
See https://www.eia.gov/outlooks/archive/ieo04/pdf/appd1_d6.pdf
In the text to the report they say they expect output to be 90% of capacity so for the reference case this implies Saudi projected output in 2025 of 20 Mbpd, even if this is total liquids it is pretty far off from reality (2022 total liquids was 12.2 Mbpd and 2024 was 10.7 Mbpd) I am pretty confident they will not reach 20 Mbpd in 2025. The high oil price case has lower Saudi output at 14.4 Mbpd in 2025 and the low oil price case has higher Saudi output at 28.4 Mbpd in 2025.
Simmons was suggesting these projections were not very realistic, that seems correct based on the history.
OK, I guess…agreed on AEO 2004 counterpoint.
Simmons price predictions were pretty wacked though. Look at that bet he lost. As were his conspiracy theories about the Saudis producing from tank farms to cover up declines. As was the running around with peak oiler calamity worries.
I admitted to not having the book. But are you telling me all his sturm and drang was consistent with Saudi production having this record through 2025?
*From 1999 to 2004: 8.3 MM bopd
*From 2005-2014: 9.3 MM bopd
*From 2015-2024: 9.7 MM bopd
His whole shtick was just “gonna be flat”?
Up thread it was asked how solar rents in my area compare to farming the land.
This year, as well as last year, farming is a money loser for the farmer. Cash rents for landlords range from $100-$300 per acre.
So annual rental of $2,500-$4,500 per acre per year, with a 1-2% annual increase, is one heck of a deal. Life changing money if it pans out.
Shallow sand,
Wow, thanks. for the info.
Rather than going full bore (utility scale) solar on farmland many farmers will move toward a smaller scale deployment. And they will find that it makes a big difference to their viability and economic independence.
Sometimes it works out, and in a few cases it works out spectacularly.
https://x.com/mikecasey/status/1969243408695701948
Old Matt Simmons:
“If it is unrealistic for Saudi Arabia to substantially grow its daily oil output on a long-term sustained basis, a more urgent energy question arises: How easy would it be for Saudi Arabia simply to maintain flat production of seven to eight million barrels a day?”
https://www.theglobalist.com/saudi-arabia-oil-bust/
———
I guess it WAS “pretty easy”. Since it happened for last 20 years. Oh wait…7-8 MM bopd for last 20 years isn’t even accurate. It averaged over 9!
https://fred.stlouisfed.org/series/SAUNGDPMOMBD
In fact, I think I got why Matt Simmons had his panties in a bunch. Look at that FED chart (linked above) from 2000-2002. Scary, scary, huh? Dropping from 8.1 to 7.1.
But flash forward and it was just ANOTHER example of peak oilers getting all wound up because of a few years…and extrapolating that to a general worry. And then the facts come in and make you look bad when volume reverts.
Long, long story of the peak oilers. Keep falling for that Lucy football trick.
Oh well. At least TOD and ASPO had to shut down from terminal embarrassment. Been over 10 years since we had an ASPO convention.
Simmons, Staniford, Whipple, Campbell, Deffeyes all gone. Seems like we reached peak Peak Oilers before we reached Peak Oil.
The Matt Simmons quote comes from your link above, which I had not read previously, it is an excerpt from a 2006 version of Twilight in the Desert, I have an earlier 2005 edition where that quote does not appear.
Also note how supposedly Saudis and others were claiming capacity could be increased to 15 Mbpd and maintained for 50 years (that is 273 Gb of output).
I am skeptical of such claims.
1. OK, noted, you’re skeptical. The way you say it sounds like you are passing judgment, a little. I will take it as more like…”your opinion, man”. I mean you’re a commenter and an amateur analyst. It’s fine…and I’m interested in your hunches. But it’s not like Tom Brady opining on football strategy. And he ain’t 100% either.
2. I wouldn’t muddle those two things. Max production for a year is a different thing than total resource. Maybe they only have 150 BBO instead of 250. But that’s a different thing than could they (with some work) raise production for a year or two.
I mean…look how the US grew. SA will be easier to grow than the US. And the US went from 5.7 to 12.3 MM bopd in eight years.
And SA would be easier than the US. No? You gonna make an argument that new wells in the US have higher IPs and lower declines than in the Middle East? Huh? I thought the whole peak oiler thing when you throw shade at shale is to point out how crappy the wells are compared to the ME! Now you want the opposite of that side of the argument? Huh?
3. And of course (a) it would take some work/investment and (b) it would suck the beer bottle dry earlier.
P.s. I feel like I’m in opposite world trying to remind you that the ME has better wells than West Texas. And how a total resource area under the curve works….you can drink the beer faster, but there’s still so many ounces in the bottle. I’m taking the traditional peak oil side here!
Nony,
I am skeptical of 15 Mb/d for 50 years, full stop. I agree they probably could increase output, but will only believe they have the capacity when it is demonstrated. They have claimed 12.5 Mbpd capacity for many years, maybe the real sustainable capacity is 90% of that (11.25 Mbpd) so far they have never reached even 10.7 for a 12 month average based on EIA monthly data from Jan 1973 to May 2025 (see link below).
https://www.eia.gov/international/data/world/petroleum-and-other-liquids/monthly-petroleum-and-other-liquids-production
I’ve never driven my car as fast as it will go either. 😉
It should be crushingly obvious that running the depletion at 15 MM bopd will mean that it expires in 2/3 the time that running it at 10 MM bopd. But if you can do 10 for 20 years, you can (with investment in drilling and pipes) do a year or two at 12, which was your previous hurdle! Not 50 at 15.
I said, I think the instantaneous capacity is about 11. And they could do 12 with a moderate amount of work (time, money). Of course 15 would take a lot more investment.
I don’t see why this is such a hard concept. After all, if the US can physically add 2 MM bopd, going from 10 to 12 (as they did within 2018), why is it so crazy to think SA could add capacity and just suck the straw faster if they chose to? It’s clear that wells come in way stronger and have lower decline in SA than in the Permian.
Of course they may not have instantaneous capacity to do 12.5 And they certainly don’t for 15. Why they heck should they build a bunch of pipes for a flowrate they don’t normally use? They’ll do it if they ever need to and choose to. Heck…they want to keep prices up also and not just crash them.
Anon
If you are so clever why can’t you answer a single question put to you?
What is Saudi capacity today?
Up to 11, almost overnight. Up to 12 with a few months work and the “how to” already identified.
Higher than 12 would be possible also, but would require significant drilling and infrastructure development.
They aren’t geologically limited per se. I mean if you do 20 years at 10 MM bopd, that’s like doing 13.3 years at 15 MM bopd. It’s just sucking the straw harder. You peakers taught me that.
I think the reason why you don’t see higher production is not geological constraints. The Persian Gulf is way richer geologically than West Texas. The Saudis just don’t want to crush price. You know, OPEC and all.
Anon
You are dreaming.
Please tell us which fields will produce what to get up to 12 million barrels per day
Basis for my guess is Fesharaki’s video, with a high level comment, about top line.
I don’t know the fields.
Nony,
Perhaps the Saudis can produce more, but so far their C plus C output has been 10.64 Mb/d for the highest annual average (in 2022).
Your “quote” from Simmons seems to have been created by the author of that linked article, it is not in the book.
1. OK, thanks for the info on the quote.
2. I mean…sure we can’t prove that the Saudis can do 12+ in a Eucidean fashion unless we see it. (Nor can we prove the converse either, of course!) But from a logical fashion, looking back at c. 2005, if the Saudis did 20 years from 2005-2024 at ~9.5 bopd (and they did), why is it so hard to believe they couldn’t have done 12 during that time, for a year or two? I mean the oil was in the rocks. Sure, it drains the reservoir faster. And sure you gotta drill more wells and expand the pipelines. But it is not an “area under the curve” constraint. I would think a peak oiler would understand that especially well.
Nony,
Perhaps it is possible, but it may be putting the infrastructure in place to increase capacity to 12 Mb/d is not easy nor inexpensive to accomplish. I will believe the capacity exists only when it has been demonstrated. I am not claiming that it cannot be done only that it has not been accomplished to date.
Of course. Fesharaki said up to 11, easy. 12 with a few months work.
I mean was it easy or cheap to add the infrastructure for the US to add 3 MM bopd in the two years from 2017 to 2019?
No.
Did it happen?
Yes.
Was it physically possible?
Yes. (Since…uh, it happened.)
————
Is it going to be easier/cheaper to add temporary production in the US or SA?
SA is going to be way easier. The average well (new or operating) has way higher production than the average new well in the US.
———-
Besides, why are you even worrying about how much investment is needed to bring the oil in? Isn’t what matters to peak oilers the geological constraints? So if you have 20 years at 9.5 MM bopd, why is it so hard to think that (with investment in drilling and pipes) they couldn’t have sucked the straw a little harder and done 12 for a year within that?
I didn’t know Mathew Simmons but I followed him religiously. Like many of us today, he didn’t take into account the advent of new technology and also, perhaps, the determination of man in his quest for oil. For example, fracking was known at the time of his “Twilight” book, but for shale gas in the Barnett, and it was a giant leap for man from the fracture of shale to release a tiny hydrocarbon CH4 to the release of longer chain and even aromatic hydrocarbons. It was also difficult for him because new drilling in the KSA didn’t impress anyone. The big Safaniya field was in existence and he took careful note of it, but it produced at <150,000 bopd for the first ten years of its development. No one would ever suspect that by 2020 it would produce at a rate of 1.5 mbopd. And Mr. Simmons would have been shocked to see that seafloor sand was dredged up in close-by shallow water Manifa Bay. From that, engineers would create 30 drilling platforms connected by causeways, converting a shallow-water field in an environmentally-sensitive area to a stunning exhibit of how to do it right. He learned as much as he could about the Saudi fields but grossly underestimated their extent. For example, in 2010, the year he died, there were about 100 rigs working in the Kingdom. Today, AI tells me (using Baker Hughes as its humanly source) that there are 230. That's a whole different ballgame than the one Mathew Simmons was watching.
So what's coming next? Well, EOR, for one. The Saudis are collecting CO2 from their big gas plant and piping it to an oilfield particularly amenable to miscible formation with the larger hydrocarbons stubbornly trapped in formation. I suspect someone will figure out how to use CO2 in the horizontal shale oil wells too, and which ones will likely respond favorably. And refracking will result in a few more molecules. Despite these maneuvers to collect the dregs, we're standing pretty much where Mr. Simmons stood two decades ago, predicting the end of the lifeblood of civilization, the commodities that gave us the Industrial Revolution and more. There are certainly plenty who will point out that these molecules also created the anthropomorphic era–the first "age" that truly changed the climate. They would be accurate in that, but without oil and gas most of us wouldn't be here, so there's that.
Still, as we must learn the hard way, there is a geologic limit, just as Mathew Simmons and others warned about. I have no idea of the year we'll reach that, and neither does anyone else. I am confident that when it comes, the world will have mostly gone on to an alternative source of energy, which will probably be SMR's. I'm not accusing anyone of disparaging Mr. Simmons' authenticity or besmirching his standing, but he was truly one of the giants warning the rest of us that the single biggest boon for civilization had to be protected. And we haven't done that.
The problem is the doomers lost a lot of credibility sounding the alarm bells so hard, so early. If they had said there will be an issue more than 20 years from now, everyone would have yawned at them.
And the doomers wanted the clicks. People got their rocks off on the Ace predictions. The more doomer the better for the wingnuts reading the articles. So the TOD/ASPO/Simmons/Deffeyes and the ilk cried wolf and did it hard. Now if we get a real wolf in 2030, nobody will pay attention.
https://peakoildebunked.blogspot.com/2013/07/429-oil-drum-bites-bag.html
P.s. Look at Matt Savinar pivoting from peak oil grift to astrology! Is that “science based”, Dennis? 😉
Anon
If you don’t know which Saudi fields will produce your 12 million claim then it is simply a guess.
People who guess know nothing and people who know nothing are usually obvious by the childish way they make fun of others.
Iver,
All projections about the future are guesses.
I’m fine, calling it a guess. I’m not proud. We are living in a world of uncertainty. I’m not trying to sell certainty.
You don’t know for sure that they CAN’T do 12 either. It’s all guesses.
For that matter, even if I did (or you do) know a bunch of field names, neither I nor you have the internal secrets of ARAMCO on what is going on at those fields.
I mean George Kaplan had this MASSIVE (and impressive) list of field names in GOM and a lot of info on their history and characteristics. And he STILL had a peak oiler prediction that embarrassingly (he ran away for three years after flubbing it) underpredicted GOM production. And the EIA which he thought was smoking crack and didn’t give detailed field names ended up being pretty close to correct. I have no idea what their process was or is…but I know who had a better record on that prediction and hence forward who I trust more.
And again…I don’t GET why it’s so hard to imagine 12+ MM bopd in SA for a year or two. (With some work, of course, to drill more and have transport pipe increases.) I mean…it’s in the rocks. It’s just sucking the straw harder and being done with the drink earlier. Duh. I mean…this is a peak oil concept!
Nony,
I agree in principle that if Saudi Reserves are as large as they claim, they should be able to produce more. They have not done so in any sustainable fashion (I define sustainable as 12 months minimum). The Saudis tend to exaggerate so I only believe there is x capacity when x is produced on average over 12 months.
And yes there is obviously uncertainty about the future and even about the past in the case of Saudi Arabia because very little can be verified after 1982 or so.
By the tone of your comment I rather suspect that you missed an important set of numbers.
In 2010 the Kingdom of Saudi Arabia was running about 100 rigs to produce an average of 8.5M bopd. Currently a rig count of ~230 is required to produce an average of 9.7M bopd.
IOW, 100 extra rigs are required for the KSA to hold its place in the global market. It would appear that shale oil producers are not the only ones suffering from the Red Queen Syndrome.
It’s actually a little complicated (and confusing) to look at Saudi rigs on Baker Hughes. They have different definitions for what is a rig (e.g. including workover rigs) in SA versus ROW. And different standards of “active” in ROW (turning to the right most of the week) and “operational” in SA (listed as non-idle during the month). And the definitions and usages have changed over time, since 2019 for ROW and since 2024 for SA…and they say they are not backdating the data with new definitions!
FWIW, I just pulled the Baker Hughes worldwide data downloaded the spreadsheet, filtered for SA:
Third link at this page: https://rigcount.bakerhughes.com/intl-rig-count
The sheet shows a SA total of 81 rigs in 2024 and 76 in 2025. (In four categories total: off and onshore categories, with oil and gas subcategories for each.)
I’m comfortable that it takes a lot more rigs to maintain production in the US than in SA, no matter the correct numbers. Yes, even now. And I’m sanguine if a few more rigs for SA are needed now than in 2010. Since it is STILL easier than in the US, even at 230!]
[I have no clue how the definitions work, exactly…But I DO CAUTION you about Googling and ChatGPTing for SA rig counts, if they include workover rigs, have different definitions over time, etc. There is an interesting press release from BH sort of “clarifying” that the 230 number includes workover rigs. (My guess is the 76-81 numbers do not.)]
A new post is up
https://peakoilbarrel.com/steo-september-2025/
“Congratulations on being one of our top readers globally – you’ve read 415 articles in the last year”
I ned to get out more—-
A detailed article on the gasoline situation across Russia at the moment (John Helmet, former US administration official)
https://johnhelmer.net/russias-petrol-supply-this-is-what-is-happening-between-the-refinery-and-the-gas-pump-and-why/
[…] maintain idle reserves to increase oil production if needed. These reserves are estimated at 3 to 6 million barrels per day, bringing the total global reserve capacity to more than 10 million barrels per day. This overhang […]
[…] maintain idle reserves to increase oil production if needed. These reserves are estimated at 3 to 6 million barrels per day, bringing the total global reserve capacity to more than 10 million barrels per day. This overhang […]
[…] maintain idle reserves to increase oil production if needed. These reserves are estimated at 3 to 6 million barrels per day, bringing the total global reserve capacity to more than 10 million barrels per day. This overhang […]