Short Term Energy Outlook, November 2024

205 thoughts to “Short Term Energy Outlook, November 2024”

1. “Notice the slope of the production curve becoming flatter since 2023, this may not allow further increases in US exports of natural gas in the future, unless US natural gas consumption starts to decline.”

   I think the story is the reverse. Prices are low. That’s what’s stopping gas from growing. And we are exporting everything we can. (But you can’t just open an LNG facility on the spur of the moment.)

   There is a huge amount of very rich dry gas shale. And lots of rigs available. It’s just not worth mobilizing rigs into gas plays, given the prices.

   1. Only too obvious that charts can only tell you what has happened and not what will happen.

      https://www.aogr.com/web-exclusives/us-rig-count/2018

      Who knows how much gas will be produced when the gas rig count goes back to over 200.

      1. Loadsofoil,

         I agree, also the last two points on that last natural gas chart are forecasts by the EIA, those forecasts are often wrong and are revised monthly. The most recent EIA long term forecast for US natural gas was published in October 2023 (IEO 2023). Chart below compares that with data from 2005 to 2023 and STEO forecast for 2024 and 2025. The future is unknown, guesses is all we have. The chart is for dry natural gas output for the US.

         

         1. Legend reversed in chart above, sorry. Corrected chart below.

            

         2. Hi Dennis I recently sent you a email on the subject of lars and his net export math. I don’t wanna seem demanding but do you think you could get back to me sometime after Thanksgiving?

            1. Jacob,

               See your email, sorry I missed yours.

   2. Nony,

      Excellent point. I looked at the natural gas data a little more closely and if we ignore the STEO forecast for 2024 and 2025 we find the dry natural gas output grew at an average annual rate of 4.2% from 2005 to 2023 while total dry gas consumption grew at an average annual rate of 2.3% over the same period.

      I wonder if there will be a bit of a conundrum in the future over the right natural gas price. It has to be high enough so producers can make money, but low enough that LNG exports can be profitable. It will be a difficult needle to thread. I agree there is a lot of potential natural gas resource, the question is how much of it is an economically recoverable resource. I think it will be less than many optimists believe.

      1. Carbon prices will probably be the main driver of natural gas prices for the consumer. In Europe and East Asia, carbon emissions are likely to get more expensive in the near future. This makes the conundrum even more problematic.

2. My new post:

   US crude exports 2014 – August 2024
   19 Nov 2024
   https://crudeoilpeak.info/us-crude-oil-exports-2014-to-august-2024

   This article includes a link to a video by
   Sal Mercogliano, a maritime historian at Campbell University and author of the youtube channel “What’s going on with shipping”. It’s worth watching. He has 5 recommendations for the new administration to improve US shipping

   1. 1. Nice post. I like how a lot of what you do is just process the overall data and visualize it, versus only showing data that supports a particular point.

      2. I remember arguments with Big Orrin in 2017-2018. He said that nobody would use US crude. That didn’t prove right.

      3. The idea that nobody wants light sweet is kind of silly. Sure there has been some light on light competition between US and Africa. But light sweet still gets a price premium versus medium sour (and even more versus heavy sour) in all regional markets. Even the US, which is closest to the problem. But also Europe…and definitely Asia. The market has handled this just fine.

      I mean refiners live in this world of looking for the cheapest cargo and trading off value versus price. So a $2.00 price premium shrinking to $1.00 means something to them. But producers live in the (traded) WTI world. And daily swings of a dollar or more are totally normal. So they really, really, really don’t care where the crude ends up. Just what they get at Cushing (or Houston or Rotterdam). And the long, slow, partial shift of medium sour to light crude is not an issue for the producers. They are much more affected by price gyrations from OPEC, recessions, wars, war jitters, oversupply, fires in Alberta, hurricanes in the GOM, etc.

      The whole “too much light sweet” is such an example of a little bit of knowledge being dangerous. I remember Rusty Braziel saying that if the market really preffered medium sour, he’d build a plant to dump sulfur and asphalt (which are incredibly cheap) into light sweet oil.

      4. Imported WTI (the actual West Texas Intermediate oil, not the financial surrogate from Cushing) is now the key source of supply for setting Brent price. Like it’s an official part of the crude that goes into Brent price nowadays, not just a traded alternative.

      5. The Canadian ATL refineries can process medium sour (e.g. Persian Gulf just fine. If they are doing US light grades it’s because of economics. They were mostly doing Basrah, prior to the US shale boom. They are pretty nimble and shift sources based on economics and access the global markets via sea. They have the facilities to handle desulfurization/hydrogenation and medium to heavy API (visbreakers or cokers).

3. Curious about the electricity generating GW capacity and electrical generation (TWh x 1000) by source. “Other sources” has over 60GW capacity right now and will be nearing the capacity of nuclear by mid-2025, but shows no production of TWh in the adjacent bar graph. Why?

   1. Joe Clarkson,

      Capacity is a less useful number as some types of generation operate at 95% capacity and others at 25%. For that reason I focus on generation rather than capacity, I ignore the left side of that chart.

      1. OK, I agree that capacity is not as important as generation, but am still mystified that the “other sources” capacity, all 60-75 GW of it, has no generation at all.

4. Conventional dry natural gas output in the US declined at an average annual rate of 5.84% per year from Jan 2010 to Dec 2019, pretty much all increase in US natural gas output in the future will be from shale gas. The chart below looks at the natural log of shale gas output in the US from Jan 2007 to August 2024, the slope of the output curve indicates the annual rate of growth. From Jan 2021 to Dec 2022 the rate of growth was about 7% per year, but the most recent 24 months (Sept 2022 to August 2024) there was a slight annual decrease of 0.27%. Higher natural gas prices might fix this problem, but it is unclear what long term price is needed to lead to a long term investment in more LNG export capacity and what long term US demand will look like, the bet on LNG export capacity looks like a risky proposition as low natural gas prices may lead to a lack of natural gas supply.

   

   1. The report linked below from April 2024 expects an oversupply of LNG for World Markets, this points to lower levels of US LNG exports due to lower LNG prices on World markets and lack of profitability of further LNG export capacity expansion.

      https://ieefa.org/resources/global-lng-outlook-2024-2028

      North American LNG export capacity is expected to increase by 13 BCF/d from 2024 to 2028 with 9.7 BCF/d from US, and 2.5 BCF/d from Canada, and 0.8 BCF/d from Mexico. See link below for more detail.

      https://www.eia.gov/todayinenergy/detail.php?id=62984

      Article below gives some insight on near term LNG prospects for US, it will be interesting to see what happens to Henry Hub prices over the next 4 to 5 months.

      https://boereport.com/2024/11/20/us-lng-exports-primed-to-jump-as-price-arb-to-europe-opens-wide-maguire/

      1. Natural gas price futures have been heading higher of late, $3.44/MCF at Henry Hub as I write and $14.45/MCF at TTF (Netherlands). The spot price for Henry Hub natural gas averaged $1.79 per million BTU (1.037 MCF = 1 million BTU) for week ending November 15.

         

5. Scan here now and then.

   Item one, the World Statistical Report quotes US natural gas reserves at about 12.7 trillion cubic meters. This is well into the shale era and it would be nearly impossible for that number to not include an estimate of shale gas.

   Production, or consumption, is just under 1 trillion cubic meters per year. So yes, that’s 12 years. The folks at BP or that new outfit that they’ve outsourced to do have a caveat on all of their reserves numbers saying that they are in the process of reevaluating methodology, but those numbers are not a new departure from previous presumptions, and indeed have increased since fracking. So 12.

   As regards Trump’s impact on production, there is a fundamental problem with pretty much all assessments of what is coming in the new administration. If folks do not live in red areas they don’t generally understand what just happened. The campaign was not about issues. It was about push back against the coastal contempt held towards fly-over country, it was about anger concerning the persecution and prosecutions, and it was about warmongering.

   Somehow the Democratic campaign did not survey the populace as to how compelling their issues were in forming attitudes. Right up until the last days they remained confident that white women were motivated primarily by abortion. This was never true, and it was measurably never true. It was a very badly researched campaign. The dangers of believing one’s own propaganda.

   Probably the most powerful reality for the new administration’s oil policies is the degree to which Elon Musk has embraced Trump. Musk lost his aura of left-wing hero when he bought Twitter and erased its censorship, and that put Tesla at great risk. Were he still a left-wing hero with no embracing of Trump, absolutely no reluctance would exist in the new administration to wipe out all EVs. With Musk now a major player in the administration, measures will be taken to save Tesla from policies that erase government pressure in favor of EVs. Something clever will take place that provides Tesla with an overwhelming advantage over all of the competition because efforts will be made to wipe out that competition via policy. Anyone who is expecting the new Chinese EVs to have a path inward to the US can rethink their thinking.

   What was unthinkable a year or two ago when Hybrids became ascendant, perhaps now is thinkable. A Tesla hybrid.

   Perhaps the most intriguing possibility is ANWR exploratory drilling. One would expect a lot of it, and maybe outright government subsidized. When it shows nothing there, the impact on not just administration thinking, but everyone’s thinking about oil is going to change.

   1. Watcher – While I might not disagree with most of this, you left out the most compelling part. Lalala was simply not electable (no primary) and vast majority of Americans did not want to hear her cackle laugh for the next 4 years. Add to that the fact the administration tried to keep senile joe running when it was crystal clear he had no business. Also, there is a longstanding tradition of switching back and forth between the parties, now to complete the cycle the stock market will take a 30+ percent haircut for next couple years…

      1. Vast majority? Did you look at the results?

         Kamala Harris won more than 74 million votes, or 48.3% of the vote. This was not a blowout for Trump, it just appears that way because of the electoral college, the way we shade states (red or blue).

         It was a close election. Trump has now won 2 of 3 elections, one popular vote, and has never broken past 50% of the American population.

         I mean, it’s just not true to say that Trump is even close to 55 or 60 percent of the people.

         1. My biggest point is that the election had very little to do with Trump, it could have been a better or worse republican running against an unelectable Dem, same outcome. Can you imagine if the R candidate was moderate and more centrist? It would have been an absolute landslide. Dems found someone liked even less than Hillary, I think this is a watershed moment that may take a very long time for the party to recover from (if at all). Dems are licking their wound and asking, why are we so dumb? For the record, I’m neither R/D…can’t wait to see 3rd/4th/5th parties come into existence…

   2. Watcher —
      Good to see you back. I miss not seeing your frequent posts.

      1. Doug, the most interesting thing in the world of oil in recent years is the failure of Ukraine oil consumption to grow. Tanks, personnel carriers, mobile missile carriers . . . scurrying about to and fro . . . and in terms of consumption, utterly overwhelmed by population fleeing to Europe.

         239K bpd pre Covid. 204K 2023.

         Propaganda issues no doubt, but that has to be millions gone to live elsewhere.

6. 1. Shale gas (and gas overall) has a long history of slowdowns when price goes low. Look at 2012, 2015, 2020, etc. 2020 was the recession of course. But earlier slowdowns, were really weather related. But in all cases, it was a demand/price crash slowing growth. NOT GEOLOGY. Not “using up the resource”. Not “Peak Gas”. Demand slumped because of a warm winter (which is hard to predict) and then inventory got full and price dropped and rigs dropped. In some cases even shutins of producing wells.

   You can see a previous example of worrying about a temp slowdown here:

   https://cassandralegacy.blogspot.com/2013/07/the-shale-gas-revolution-is-it-already_7.html

   Or here (on this very blog):

   https://peakoilbarrel.com/collapse-of-shale-gas-production-has-begun/

   But in each case, production recovered after the glut got worked down.

   Now…if prices get very high (and future strip also high) and production slows down…then we actually done run out of shale! But that’s not happening and I doubt it happens for decades. There’s a lot of the Haynesville left to drill and other gassy plays like the Eagle Ford or Ohio Utica or even the Woodford in OK and TX. (Huge amount of Marcellus left also, but it is sort of capped by the successful efforts to limit egress pipelines.)

   2. On LNG: there’s a big scrunch factor when building an LNG terminal. It’s a lot of money and you have to commit years ahead of time. And you are competing versus Qataris and new projects offshore in Africa (which have very cheap sources of gas, cheaper than Henry Hub…but building the facility can be more expensive!).

   It’s a very different market from oil, where the external market is much larger than the internal market. So only so many projects (around the world) can go forward. In oil, you drill it and dump it on the market. In LNG, you need to line up customers ahead of time and win the bid rounds versus other projects. (In US gas, you can drill and dump gas on the HH market…but that’s not LNG export, that’s just local production.)

   https://www.youtube.com/watch?v=_SYcFAvc7HQ

   (Don’t need to watch whole video…just see discussion from 3:00 to 10:00 on difference of oil and LNG).

   LNG costs money (wastes gas, essentially) to do the liquefaction and keep cold during transit. It’s massively inefficient versus pipelines of course, but is the reasonable solution for places like Japan or Hawaii. But in a non-war world, Europe would get gas from Russia by pipe…it makes economic and even environmental sense versus the waste of liquefaction.

   3. But it also doesn’t help when the administration in power does things like this, either:

   https://www.reuters.com/business/energy/biden-pauses-approval-new-lng-export-projects-win-climate-activists-2024-01-26/

   It’s not just the actual projects held up, it’s also knowing the capriciousness of the federal government can stop you from shipping. Who wants to build a 40 year asset, if there’s a possibility that President AOC shuts it down in 2032. It’s a risk premium that has to be added onto the economics.

   P.s. Use of a log scale makes the recent increase look more plateu-ed than it is in the real world.

   See here for a linear chart (and all gas, which is what matters economically…when you buy it you don’t put a little gold sticker on all the shale molecules):

   https://www.eia.gov/dnav/ng/hist/n9070us2A.htm

   1. Nony,

      I agree prices will influence output, yes a natural log chart shows how the exponential rate of growth has changed over time. Here is the Shale gas on a linear scale, I have not claimed there is no more shale gas to produce, only that recently production growth has slowed compared to the recent past (2017-2020). Do you see a big increase over the past 25 months? I don’t see it.

      The president did not stop LNG exporters from exporting, he held up the approval process for new LNG facilities. The biggest risk for LNG exporters will be overcapacity and potentially falling natural gas prices due to slowing demand for natural gas.

      Shale gas output increased by a factor of 5 from 2008 to 2017 and by about 1.66 from 2017 to 2024, that’s the point of the log chart.

      Jan 2021 to Dec 2022 the OLS linear annual average increase in shale output was 5.5 BCF/d. From Sept 2022 to August 2024 (using 24 months to reduce seasonality) the OLS linear average annual decrease was 0.2 BCF/d for shale gas output.

      

      1. Nony,

         Another factor to consider is that a large piece of shale gas output comes from the Permian Basin, as that basin starts to run out of tier one and tier 2 drilling locations the Permian is likely to peak and then decline, probably decline begins by 2029 and perhaps as soon as 2027 (my guess is 2028), shale gas output from the Permian may start to decline around the same time.

         You suggest there is plenty of shale gas left in Marcellus, Utica, Haynesville, and other shale gas focused basins. I agree there are technically recoverable resources, but I wonder if much of it will be profitable to produce. The chart below divides shale gas production into tight oil focused basins (Permian, Bakken, Eagle Ford, and Niobrara) and all other basins (which i call shale gas focused). The chart covers the past 36 months (Sept 2021 to August 2024). This includes 2022 when natural gas prices were relatively high and over the entire period natural gas prices were not very high (monthly average Henry Hub Spot price was $3.99/MMBTU).

         

   2. Nony,

      Here is a chart with total dry natural gas and shale gas, all of the increase in natural gas output since 2007 has been from shale gas, that is the point, conventional output has been decreasing since 2003. From 2007 to 2024 conventional NG output fell by about 24 BCF/d. Shale gas is where the action is.

      

   3. US non-shale natural gas gross withdrawals. From 2009 to 2023 the average annual rate of decrease was about 5.4%.

      

   4. Nony,
      Regarding current underdeveloped sources of natty …
      Western Haynesville continues to shine as Comstock reports on an ongoing basis.
      The App Basin’s Upper Devonian formations were estimated to hold ~100 Trillion cubic feet economically recoverable resource by Gregory Wrighstone a decade back. The ~60 then-developed unconventional wells were estimated to have EURs of 1 to 3 Bcf over their lifetime. Consequent production from these – and an additional handful – Burkett, Genesee, Geneseo, et al wells have resulted in numerous units producing in excess of 4/5(+) Bcf cums already.
      Should Pacific Coast Mexico LNG plants expand/be built, the Permian ass gas could become a highly profitable LNG product for the Asian market. (This could unleash the staggeringly huge potential of the now-dormant Mancos B).
      Rusty Hutson’s Diversified Energy continues to vastly improve operating procedures for their tens of thousands of stripper wells. These approaches could have a big trickle down effect on smaller operators’ viability should they implement these hardware/technical advances.
      Coalbed methane is still there, but unattractive at these low prices.
      Heck, methane is PRODUCED – and recovered – every frickin’ day at RNG plants throughout the country (Pennsylvania just opened another one recently).
      With all your sharp, economically sensitive analysis, Nony, I offer you this to ponder …
      When it becomes widely accepted that low cost natty will be the norm far off into the future (from a strictly supply focused perspective), the entrepreneurial spirits will arise. While you dismiss natty (LNG/CNG) for wid(er) spread use in transportation, I believe you will be found incorrect. The price spread between liquids (diesel and petrol/gasoline) versus natgas alternatives is way too large for the market to continue to ignore.
      So, yes, we live in a world with ‘permanent’ low cost natgas in North America. This realization is apt to manifest in further applications of this wondrous resource … power generation and transportation being two of the most prominent.

      1. Coffeeguyzz,

         If natural gas prices remain low, not much of it will be produced. Many of the high resource estimates come from a period when natural gas prices were higher in the US (back in 2008 the annual average spot price was $8.86/MMBTU nominal for HH spot price or $12.90. MMBTU in 2024 $). At $2.50/ MMBTU we won’t see a lot of production growth. The Haynesville has has a good run from 2017 to 2023, not clear if it will recover with low HH prices.

         

         1. Marcellus, Haynesville and Utica from Jan 2018 to Aug 2024, not much growth since Jan 2023. About 5% growth per year (2.1 BCF/d annual rate) over the period on the chart.

            Data from https://www.eia.gov/naturalgas/data.php

            

            1. Hi Dennis:
               About that chart showing not much growth since January 2023, I heartily agree, its because the futures, and most other gas prices went from good to subeconomic within about a month.

               According to the Moore Research Center chart natural gas dropped from a robust 6 in December 22, to mostly 2’s in January 23. There hasn’t been a sustained rally over 3 since. When the natural gas producers can hedge out somewhere in the upper three’s gas production will start rising. Producers are restraining gas production now because of lousy prices.

               https://www.mrci.com/pdf/ng.pdf

               1. DCLonghorn,

                  I agree low natural gas prices caused output to fall. Futures prices have risen recently, but it will take a few months (4 maybe?) to see how the market reacts. Supposedly the LNG exporters need about a $11 spread between HH and TTF to export profitably to Europe. So this would imply TTF at about 15, if HH is at 4, note also that the EIA expects that as natural gas price increases US demand for natural gas will fall, so the natural gas market may become more dependent on international prices and long term contract prices for LNG exporters. In addition a lot of new LNG capacity is expected over time which may drive international prices lower as over capacity in LNG drives prices lower.

                  We may be approaching a situation where only associated gas from tight oil producers is profitable over the medium to long term.

7. I wrote a long post but the SpamAdmin deleted it . 🙁 Probably for too many hyperlinks…i.e. too many sources!

   Oh well…serves me right. Won’t bother with that…stick to unsourced chat style comments. And not waste my work.

   1. Nony,

      Keep the links to under 5, then if you need more just do another comment.

      1. Also, he can create his own blog, write something there, and then link to it.

      2. Nony,

         Alternatively you could save any comment with more than 3 links to a word processor in case it gets caught by the spam filter, I think it catches anything with more than 4 hyperlinks, but it might be 3. I restored your comment.

8. Dennis – Your shock models adapted to monthly world production. Since production is below 83 mb/d, do you believe something less than 2,500 Gb is reasonable?

   

   1. Kengeo,

      The chart below shows the parabola you have fit to the June 2020 to July 2024 data, as is clear this is not a good model for the general World C plus C data set. The World C plus C URR is not likely to be as low as 2500 Gb and it may well be that 2900 is also too low, it may be 3100 Gb or perhaps higher, it will depend in part on demand for liquid fuel.

      Does the parabolic model you have created look reasonable in your view? It is only wrong by 80 Mb/d for 2015 and it continues to negative values before 2015 and after 2032. Maybe go back to the drawing board.

      

      1. My best guess would likely be around ~2000 Gb, so roughly dropping ~4 mb/d annually for next 15 years.
         2025-77 mbpd (~150 Gb)
         2030-57 mbpd (~130 Gb)
         2035-37 mbpd (~100 Gb)
         2040-17 mbpd (~80 Gb)
         2040 to 2060 (long tail, 130 Gb)

         URR for conventional is 2,000 Gb, other unconventional deposits will never amount to more than ~10-15 mbpd, they will likely be the long tail after ~2050.

         URR for economic oil is likely only 1750 Gb or so, this is what brought the 2004 peak oil event and when the term unconventional oil became a buzzword.

         1. Also, I think a much much steeper drop from 80 to 60 mbpd could be possible, something as drastic as 10-15 mbpd over the course of 4-6 months, many different scenarios could have this result.
            All it would take is a roughly 25-30% reduction from US, Russia, and Saudi Arabia…
            Price could drive production down or other factors…

            Long-term, gradual drop in production would actually be a good thing…a 2 mbpd annual decrease would likely be about as close to perfect as possible…

            1. Kengeo,

               Anything is possible, the planet could collide with a large asteroid, or we could have a nuclear holocaust. The biggest annual decrease historically was 6132 kb/d in 2020 over the period from 1973 to 2023. So you are thinking about 5 times more than that, I disagree strongly that is likely except in the case of a planetwide catastrophe such as WW3 or a something at the level of the Permian era asteroid event. The likelihood is small, but remains possible.

         2. Kengeo,

            Here is your model in chart form with 4% decline after 2039 and where I assume average output in 2024 will be 82 Mb/d with output decreasing by 4000 kb/d each year for 15 years.

            I do not think this is close to being correct, just putting it out there for people to see.

            

         3. Kengeo,

            A decline of 4 mb/d annually for the next 15 years would be a disaster, but it wouldn’t surprise me very much. I don’t expect it to happen though.

            This although I remember the regularly written comments by ROCKMAN on theoildrum website. About how agressive secondary and tertiary crude oil recovery techniques deplete oilfields very rapidly. Good before Peakoil happens….not so good thereafter.

            I don’t pay too much attention to the hypothetical crude oilproduction graphes.
            Geopolitics, etc. will make those smooth graphes not very relevant. For a guess they are ok

9. We agree, both 1750 Gb and 3500 Gb are both very unlikely , somewhere in between (2650 Gb) plus or minus 15-20% seems more plausible.
   Given current production of 1450 Gb, and 1P reserves of 450 Gb, seems a low estimate would be 1900 Gb
   Using Rystads latest estimate for 2P reserves would yield nearly 2200 Gb for a reasonable estimate of URR. This is 20% less than 2650 Gb. Average of 1P and 2P URR is 2050 Gb. This is very close to my best guess if 2 Tb. This is implies that 2/3 of conventional oil has been produced, leaving only 0.5 Tb left.

   1. Kengeo,

      Keep in mind that the definition of proved reserves is that there is a 90% probability that reserves will be larger than the proved reserve estimate at present. Reality is that it is probably more like a 99% probability when contingent resources, reserve growth and discoveries are taken into account. The 2P reserves would supposedly have a 50% probability that reserves would be larger, if we ignore contingent resources, reserve growth and discoveries, reality is probably on the order of a 75% probability that URR will be larger than 2P reserves plus cumulative production (about 2240 Gb). I would put the 1 sigma range (roughly a 70% probability the URR will be in this range) as 2700 to 3100 Gb with 2900 Gb being my best guess (with 50/50 odds of URR being higher or lower than this) maybe a 50% probability of URR being in the 2800 to 3000 Gb range, with 25% probability it is lower than 2800 Gb and a 25% probability it will be higher than 3000 Gb. Conventional will be around 2750 Gb +/- 250 Gb in my view.

      1. Fascinating misunderstanding of statistics and reserves Dennis. A reasonable range is between 2 – 2.4 Tb, best estimate being 2.2 Tb.
         Your math:
         90% -> 99%
         50% -> 75%
         So let’s correct your numbers:
         3 Tb / 174% =1.724 TB
         1.724 Tb x 140% =2.4 Tb

         Reminder on 2P definition: Midpoint scenario of recoverable resources, often considered the most realistic estimate for planning and forecasting.

         1P = 1.9 Tb

         For 75% likelihood-2 Tb, we can estimate as follows:
         99% – 1.8 Tb
         95% – 1.85 Tb
         90% – 1.9 Tb (1P) – (10% URR is lower)
         80 – 1.95 Tb
         75 – 2.0 Tb
         70 – 2.05 Tb (30% URR is lower)
         60 – 2.1 Tb
         50 – 2.2 Tb (2P) Best Estimate
         40 – 2.3 Tb
         30 – 2.4 Tb (70% URR is lower)
         20 – 2.5 Tb
         10 – 2.6 Tb
         5 – 2.9 Tb (95% likelihood URR is lower)
         1 – 3 Tb (99% likelihood URR is lower)

         And please don’t forget that Laherrere believes 2P shouldn’t be taken seriously as it’s overly optimistic and likely plain wrong.

         1. Kengeo,

            Laherrere’s estimates have increased over time as 2P estimates are the best estimate at the present time under an assumption that no contingent resources become reserves in the future and that there are no further discoveries, I understand statistics quite well. You also misinterpret Laherrere’s analysis in my view.

            My analysis takes into account the potential for future reserve growth and discovery, an analysis that ignores the fact that there are about 500 Gb of discovered contingent resources and that there is likely to be technological development in the future that lowers costs, making even more resources viable along with ongoing discoveries.

            USGS estimated conventional resources to be about 3000 Gb in 2000 and unconventional resources might be 500 Gb, so I would put the upper limit at about 3500 Gb (99% probability URR will be lower). I would put the minimum at about 2500 Gb (when we take all resources into account and technological change, discovery and reserve growth) where there is a 99% probability URR will be higher than 2500 G. Note also that despite people thinking my estimates are too optimistic, historically I have been too conservative.

            Your reasonable range is not very reasonable simply because you fail to understand reserves and resources. Keep in mind that 2PC is about 1200 Gb (this does not even require any discovery or reserve growth, this is the engineering best estimate of discovered resources). Add 1500 GB of cumulative production and we have 2700 Gb for a best guess URR under an assumption of no new discovery, no reserve growth and no further technological development(none of these are reasonable in my opinion).

            Note also that for the Rystad estimates from 2017 to 2023 the 2PC estimate remained the same while about 30 Gb were produced annually over that period. The combination of reserve growth and new discoveries over that period was on average about 30 Gb per year. You seem to think this will end starting in 2024, I disagree. Perhaps the average level will be less over the next 10 years, say 20 Gb per year on average. That gets us to a URR of 2900 Gb as the reserve growth plus discovery occurs, or lets say it is 10 Gb per year on average over the next 20 years, we still get to 2900 Gb.

            Another thing to consider is we know URR can’t be less than 1500 Gb as we have already produced that amount, that is the floor. There is a lot of room above 2900 Gb and much more than 2900 Gb could potentially be produced, I am not so optimistic that I think it cannot happen, but I am hopeful that my guess of 2900 Gb is roughly correct. I would prefer that your guesses were correct from an environmental perspective, but I am just not that much of an optimist.

   2. Dennis –
      To summarize, a conservative estimate of remaining oil uses 1P plus 100 Gb for a total of 2 Tb – this uses pessimistic assumptions about reserve growth, undiscovered resources, and technological advancements.

      A Best Estimate uses 2P of 2.2 Tb

      A moderately optimistic estimate is 2.4 Tb (75% confidence level meaning URR is likely lower, but not by a lot). However, this requires significant future contributions from reserve growth, discoveries, and technological advancements. While not the highest possible estimate, it reflects a scenario with relatively favorable conditions for oil recovery.

      1. From

         https://www.spe.org/en/industry/petroleum-resources-classification-system-definitions/

         Contingent Resources

         Contingent Resources are those quantities of petroleum which are estimated, on a given date, to be potentially recoverable from known accumulations, but which are not currently considered to be commercially recoverable.

         It is recognized that some ambiguity may exist between the definitions of contingent resources and unproved reserves. This is a reflection of variations in current industry practice. It is recommended that if the degree of commitment is not such that the accumulation is expected to be developed and placed on production within a reasonable timeframe, the estimated recoverable volumes for the accumulation be classified as contingent resources.

         Contingent Resources may include, for example, accumulations for which there is currently no viable market, or where commercial recovery is dependent on the development of new technology, or where evaluation of the accumulation is still at an early stage.

         Reasonable time frame is typically 5 years.

         Determining whether a contingent resource can be classified as reserves depends on it being classified as “commercial”.

         From 2018 PRMS (pages 6-7): https://www.spe.org/en/industry/reserves/

         2.1.2.1 Discovered recoverable quantities (Contingent Resources) may be considered commercially mature, and thus attain Reserves classification, if the entity claiming commerciality has demonstrated a firm intention to proceed with development. This means the entity has satisfied the internal decision criteria (typically rate of return at or above the weighted average cost-of-capital or the hurdle rate). Commerciality is achieved with the entity’s commitment to the project and all of the following criteria:

         A. Evidence of a technically mature, feasible development plan.

         B. Evidence of financial appropriations either being in place or having a high likelihood of being secured to implement the project.

         C. Evidence to support a reasonable time-frame for development.

         D. A reasonable assessment that the development projects will have positive economics and meet defined investment and operating criteria. This assessment is performed on the estimated entitlement forecast quantities and associated cash flow on which the investment decision is made (see Section 3.1.1, Net Cash-Flow Evaluation).

         E. A reasonable expectation that there will be a market for forecast sales quantities of the production required to justify development. There should also be similar confidence that all produced streams (e.g., oil, gas, water, CO2) can be sold, stored, re-injected, or otherwise appropriately disposed.

         F. Evidence that the necessary production and transportation facilities are available or can be made available.

         G. Evidence that legal, contractual, environmental, regulatory, and government approvals are in place or will be forthcoming, together with resolving any social and economic concerns.

         …

         2.1.2.3 To be included in the Reserves class, a project must be sufficiently defined to establish both its technical and commercial viability as noted in Section 2.1.2.1. There must be a reasonable expectation that all required internal and external approvals will be forthcoming and evidence of firm intention to proceed with development within a reasonable time-frame. A reasonable time-frame for the initiation of development depends on the specific circumstances and varies according to the scope of the project. While five years is recommended as a benchmark, a longer time-frame could be applied where justifiable; for example, development of economic projects that take longer than five years to be developed or are deferred to meet contractual or strategic objectives. In all cases, the justification for classification as Reserves should be clearly documented.

         1. Dennis –

            Wishful thinking can only take a person so far.

            Here’s the fatal flaw in your argument:

            Peak oil (conventional) is now 8 to 20 years in the rear view mirror. For Conventional URR of 1.6 Tb we would have expect a peak on or near year 2000 at 68 mbpd, actual was very close to that value.

            For 1.8 Tb, peak in 2007 at 72 mbpd would be expected and was observed.

            Now for the next value of 2.2 Tb, a peak of 78 mbpd should have happened but did not (was 70) – >10% too low.

            For 2.4 Tb, 2025 peak expectation is 90, but actual is looking like ~64 mbpd. 40% too low!!!

            You can continue this (forecast).
            2.6 Tb – 2030 at 95 mbpd, actual will be ~52 without significant changes. This is 80% too low.
            2.7 Tb – 2033 at 98 mbpd, if 46 then this will be 113% too low…
            By time we make it to you high estimate of 3.4 Tb, peak year is 2050 and conventional production is 115 mbpd, if 23, then this is 500% too low…

            You need to focus on the oil that matters…focus on the data points that matter.

            My guess is you will make up some interesting rationale so you can hold on to your overestimates…

            

            1. Kengeo,

               Peak conventional was 2016 at cumulative output of about 1273 Gb. The rise of unconventional output made conventional producers such as OPEC restrict their output. My URR estimate is for conventional and unconventional output. I generally do not assume that peak is required to occur at 50% of URR, it seldom does. Why do you think a polynomial fit to data tells you anything, it doesn’t.

               We will see how it plays out, I hope you are right, but think you are not.

            2. Kengeo,

               Where are you getting these peak values? Not clear you are correct on those. For a conventional Hubbert model fit to EIA data with URR of 2500 Gb I get a peak in 2016 at about 73.9 Mb/d, just slightly above the actual peak of 73 Mb/d. Note however that in the real world output does not follow a logistic curve. So as far as what data matters, that is never very clear and any model will be only a rough approximation.

               1. Kengeo,

                  Hubbert curve for World C plus C based on HL for 2000 to 2019 EIA data, URR about 3000 Gb. Peak is 2024 at 81.7 Mb/d.

                  It is likely 2024 World C plus C output will be about 82 Mb/d, so this is roughly correct for 2024.

                  

               2. Hubbert Model (3000 Gb URR) and EIA data for World C plus C.

                  

               3. Longer term Hubbert Model (3 Tb) and data 1900-2100.

                  

               4. Slightly higher URR estimate of 3.2 Tb leads to peak of 82.5 Mb/d in 2027 for Hubbert Curve, compared with EIA data through 2023 and Shock Model forecast through 2029.

                  This is how a Hubbert curve is done properly.

                  

               5. Dennis – For that to occur (3 Tb), conventional production would need to be 25 mbpd greater than it currently is…
                  Can also verify by looking at OPEC and Russia. Together they account for 39 mbpd, falling by 2.5% annually. Conventional oil production decline will continue its march down, losing 1.5 mbpd annually.
                  US shale will continue along its plateau for several more years, entering 15% annual decline by 2030 at the latest. 2030 production will be at most 75 mbpd.

               6. Kengeo,

                  The Hubbert curve is for conventional and unconventional production, currently unconventional output is about 12.7 Mb/d (2023 tight oil and oil sands output), total C plus C was about 82 Mb/d in 2023, conventional about 69 Mb/d. You might be forgetting about extra heavy oil output from Canada and Venezuela (API gravity of 10 degrees or less) which was about 4 Mb/d in 2023, this is added to tight oil output of about 8.7 Mb/d in 2023.

         2. Kengeo,

            Various Hubbert curves, note that cumulative production at the end of 2023 was about 1500 Gb, currently OPEC plus is limiting their collective production by about 2 Mb/d and has been limiting production for most of the period from 2019 to 2023. There is also increasing output from Argentina, Brazil, Canada, Guyana, and Norway expected over the next 5 years at least.

            

            1. Dennis
               Not sure the Hubbert concept really applies to the Canadian oil sands, where product transport capacity and water limitations play a big role in how the resource is extracted, while geologic factors are minimal.

               1. Old chemist,

                  Resources that are limited will tend to deplete over time, I generally model these separately (unconventional and conventional resources), but was attempting to simplify.

                  For all resources there are many factors at play including product transport capacity and water is certainly an issue for any oil play as sometimes there is water drive involved in secondary recovery and there is almost always produced water to deal with.

                  The Hubbert concept has been applied to coal by Laherrere and Rutledge, not sure why it could not be applied to oil sands or any limited resource.

                  In this case the resource I am considering is crude plus condensate in all forms (conventional and unconventional).

         3. Hubbert curves, slight variation

            

            1. For chart above URR, Peak output and peak year

               

      2. We could use your low estimate (2000 Gb) for the bottom (with at least a 99% probability that URR would be higher),

         I would still use 2900 Gb as my best guess and 3800 Gb as the high end estimate with a 99% probability that URR would be lower. My 70% confidence interval would be 2700 to 3100 Gb and 95% confidence interval of 2300 Gb to 3500 Gb.

   3. Kengeo,

      Cumulative production through the end of 2023 was 1500 Gb, so maybe 1950 Gb for a very low end estimate, I would say a 99% probability the URR will be higher given the reality that contingent resources exist ( about 500 Gb) and reserve growth and discovery are very likely.

   4. Always like these URR discussions, as much for what they leave out, as they include..

      The underlying assumption is that current size and complexity of the background system continues to operate ‘normally’, to gain access to these reserves. The ‘normality’ we are assuming is a system of growth in the background, including size, energy use and complexity.

      The assumption is false, as we lose ‘normality’ when overall energy extraction starts to decline, leaving less energy for every activity on average, including maintenance of existing complexity.

      A lot of the remaining theoretical URR will remain in the ground as we lose the complexity to gain access to it, so it’s a misnomer in the first place. URR can and will shrink.

      1. Hideaway,

         If things play out as you assume they will, you may be correct and URR would be smaller, difficult to anticipate technological progress leading to greater efficiency in energy use. You assume energy use will contract, I also make that assumption, the difference is that I see a lot of inefficiency that can be removed from the system, my guess is that you do not.

         1. Dennis you continue to do the hand wave of technology will save us, while ignoring physical laws of the universe, in this case the law of diminishing returns. All the ‘easy’ as in ‘cheap’ technological improvements were made decades ago, now every improvement is massively expensive and has smaller returns.

            All that technology improvements do is allow greater energy inputs, to produce the technology that allows a bit more of whatever resource to be extracted. More technology is greater complexity of the entire system and is a common theme throughout the physical universe as complex systems grow.

            The hand wave of lots of inefficiencies, without explanation of exactly what it would take to rid ourselves of these inefficiencies, as in the full accounting of energy inputs to produce whatever solution, is actually nonsense, as they would have already happened if a cheap enough solution was available.

            When civilization is on the energy downslope we lose size and complexity so the ability to gain access to the remaining oil will be greatly compromised as it relies on all the modern complex technology to remain in place.

            1. Energy per unit of real GDP for World continues to decrease and rate of population growth is slowing. Those are facts that you ignore.

               1. Dennis we are still adding around 70,000,000 people per year to world population, which last time I looked meant still growing, so increased size complexity laws still apply, it’s a fact and I don’t ignore it, you do…

                  Real GDP is a man made construct to be whatever people want it to be. By leaving out asset price inflation from official inflation numbers, we never get ‘real’ inflation to adjust GDP to get ‘real GDP’. Any competent understanding of economics would know this, yet it evades just about every economist, because they want things to look better than they really are for the median person/family in the developed world where the median person/family is materially worse off over the last 5-6 decades.

                  You totally avoided the law of diminishing returns and the size complexity laws and went off in a tangent. These are physical laws that are inescapable. Thinking that some man made construct (real GDP), to get around reality, is an explanation is foolish.

                  If size or energy use, start to decline, then the complexity crashes in a chaotic way, as per universal laws!!
                  Go and do a bit of research on size complexity laws that have been discovered in various different sciences, all independently of each other, instead of trying to be an economist that thinks physical laws of the universe don’t apply to us!!

               2. Hideaway,

                  Many well respected demographers (Wolfgang Lutz and others) expect World population is likely to peak before 2060. After that we will see fewer people and when population decline reaches 1.5% which the real GDP per capita growth rate, then real GDP growth reaches zero, if population decline rates increase to more than 1.5% then real GDP for the World decreases, also as the World becomes wealthier real GDP per capita growth rates may also approach zero or even fall.

               3. HIDEAWAY –

                  Any competent understanding of economics would know this, yet it evades just about every economist

                  I’m not crazy. I’m the only sane one. The rest of the world is crazy.

               4. Hideaway,

                  I mentioned the law of diminishing returns, which you seem not to understand.

                  See

                  https://en.wikipedia.org/wiki/Diminishing_returns

                  The law states that the marginal returns of increasing one factor of production while all others are held fixed (ceteris paribus) will see eventually decreasing returns. So what does a producer do? They increase other factors of production as needed so the output is optimized. If a factory is at optimum capacity and more output is needed, another factory will be built, also sometimes process improvements are devised that can increase capacity.

               5. Dennis, for once could you look at something in scientific terms instead of economic terms?
                  https://content.wolfram.com/sites/13/2018/12/21-3-2.pdf

                  https://ecologyandsociety.org/vol7/iss3/art3/

                  “Systems grow relatively rapidly at first because they have not yet acquired any impediments to growth. Indeed, we may suppose that the rapid growth rates of immature systems best represent the universal urgency toward equilibration. However, we then notice a decline in the growth rates of older systems. There seems to be a law of diminishing returns for growth, and in ecological systems this is revealed as a limitation on diversification. I have proposed (Salthe 1993) that the limit to continued growth of any kind is imposed by information overload, because it is being loaded into a finite locale that is no longer growing.”

                  You can also look at any mining operation on the planet, they need to increase energy and material use to gain the same quantity of metal or mineral as in the past, it’s a law of diminishing returns due to lower ore grades on average, we used all the easy to get stuff first, same with oil, gas and coal, we have falling EROEI which is another form of the law of diminishing returns.

                  How is it the real world escapes your perception? Are you so overtrained in economics you can’t see the forest for the trees??

               6. Dear Alimbiquated, how about you educate yourself in these fields for a few years before making comments about it’s only me..
                  Perhaps you could look up the work of Prof William Rees, or Prof T Garrett, or Prof T Murphy, or Prof S. Michaux, or Prof J Tainter, etc, then get back to me about how all these people are incorrect…..

               7. Hideaway,

                  The law of diminishing returns is a concept in economics, often misapplied by others. As I said you do not seem to understand the concept. Mines get depleted over time obviously, that is not the same as the law of diminishing returns. EROEI will fall as resources get depleted also not the law of diminishing returns.

            2. Hideaway consider that as peoples/countries find oil products or nat gas less affordable or accessible they will use what supplies they do have more carefully. Progressively so.
               What does that translate to in action?
               The uses that are most optional will get gradually weeded out by pricing mechanisms or government rules.
               There will be increasing incentive to make do with less- better engines, better insulation.
               And a higher fraction of the combustion will be directed toward production of non-fossil energy generating mechanisms that will in effect act as fossil fuel age extenders- hamburger helper if you prefer. And this can go a very long way in stabilizing energy availability as the population peaks and then enters the decline phase.

               I know that you believe that the utilities of the world greatly overestimate the net energy output from all of these mechanisms like hydro, nuclear, solar, wind, geothermal. I see it differently, with the world utilities having a front row seat on this interplay. Their viability depends on understanding it and making correct long term decisions to navigate the scenario.
               Just watch what they are doing…what mechanisms they are deploying. That will show you exactly what has the best net energy output/capital ex. Its the proof in the pudding.

               [of course governments also impose some direction on these industry decisions for various reasons such as national energy security, domestic job/industry support, pandering to voting constituents, or environmental damage control]

               1. Hickory, going for more ‘essential’ uses of resources instead of all uses of them defies how we’ve developed our industries and economies..

                  For example consider high end computer chips, which take great complexity to produce at scale. they are used mostly in discretionary gaming consoles and say special sensors in horizontal oil rigs. If you limit their use and ban gaming consoles, the demand for these chips crash. Why wouldn’t the company making them go bust due to lack of sales and production of the chips needed for horizontal oil rigs become unavailable, so ending the complex operation of drilling horizontally.

                  It’s just an example of how the unravelling of complexity by rules or competition will work in a chaotic way when either size or energy use decline work their magic of universal physical laws on the complexity we’ve created.

                  A clear look at history of civilizations finds they all collapsed because of an internal energy depletion, while trying to maintain their complex hierarchy. Whether it was climate change, increased salinity, or invasion by outside forces, it all comes back to less food, animals, trees and slaves, the energy sources of those civilizations, diminishing, leading to their collapse.
                  Our current civilization is magnitudes larger than any prior one and relies upon 6 continent supply chains to maintain the complexity while we are still growing in size and complexity.

                  You concentrated upon what ‘utilities’ would do, yet we have an entire complex system of civilization. What use are utilities that rely upon complex supply chains if there is no food reaching cities due to other complex systems breaking down as in fuel and fertilizer to farms relying upon complex machinery??

                  It’s an entire deal, all at once when energy or size starts to decline, chaotic feedback loops effect every internal complex system leading to collapse.

                  The perfect example of size complexity laws in action are in stars, where the larger the star the more spectacular the unwinding of the complexity internally. Small stars don’t go supernova, large ones do.

               2. I do agree with you about much of that. Collapse conditions to level of disorder and decline that we see in places like Haiti have a strong likelihood of becoming much more widespread over the next 3 decades. No one should be smug enough to think that their region is immune to loss of civil society or prosperity since profound disorder and poverty can appear quickly, triggered by a pretty long list of factors.

               3. Hideaway,

                  Society is not based on physical laws alone, it is understood using knowledge such as sociology, psychology, and economics. Society is not a biological organism or a software engineering project, making assumptions about software complexity or biological complexity and how these might apply to social systems is quite likely to lead one astray.

                  Nothing in social science is ever clear cut, no repeatable lab experiments are possible. Any understanding of how society works once understood by many will lead to changes in behavior as actors use this knowledge to game the system and will lead to a new system where the old rules no longer apply.

                  As soon as you understand a social system, the system changes and requires going back to the drawing board.

                  The law of diminishing returns applies to a single input where everything else remains fixed, nothing ever remains fixed and the law never applies in general.

               4. Dennis, I have never read a more complete lack of understanding of how our systems work, than this reply of yours…

                  Giant ant colonies, are social structures and follow the same size complexity power laws, and biological systems as well as urban centres all follow size power laws, perhaps you should look into a lot of work by prof Geoffrey West on what he calls scaling power laws of cities and corporations.

                  Basically it doesn’t matter if there are social systems involved or not, the size complexity rules apply and because we require the high complexity to do so much of material, energy and food gathering of the modern world, when the size or energy inputs fall the system collapses fairly quickly.

                  Basically the bigger the system the harder the fall. The largest stars collapse the fastest, the largest civilization the world has ever endured will also collapse much faster than prior small ones. Social interactions have no influence over physical laws.

                  Your argument about social systems is akin to working around the laws of thermodynamics because you don’t want them to apply to you.

               5. Hideaway,

                  Not really familiar with complexity analysis, my guess is that there are a variety of opinions within the field.

                  A sample work at link below.

                  https://link.springer.com/article/10.1007/s10699-023-09917-w

                  I get a little lost in the mathematics at definition 6 on page 1164 of the pdf, I have not learned beyond basic multivariable calculus, differential equations and statistics, I am not familiar with set theory, topology and such that advanced mathematicians would be familiar with.

               6. Dennis it’s not analysis of complex systems, it’s size and complexity physical laws you need to look up, which has been discovered in several separate areas of science independently of each other.

                  Which is why I specifically mentioned Prof Geoffrey West’s work, or perhaps Georgiev with size complexity laws in stars.

                  It’s not about the complexity itself, it’s the physical laws that govern all growing complex systems, hence the relationship and importance of size.

               7. Hideaway,

                  The two are connected and many of the people that look at this stuff are physicists and mathematicians (which the author of the article is). If you are going to try to extend the analysis to the World you need to go beyond simple physical laws, you need to look at the entire picture which includes both physical and social science and philosophy. Your focus is too narrow.

                  Also rules of thumb like the bigger they are the harder they fall while intuitive, intuition often leads us astray, like the thinking that heavier objects fall faster than light objects, which seems true in practice but is wrong in general (say in a vacuum which applies to most of the universe).

                  Also keep in mind civilizations are social structures, obviously physical laws are important, social interactions are also important in understanding the rise and fall of civilizations. Civilization is not fixed it changes over time.

               8. Dennis the physical laws of the universe are those that have allowed us to create this civilization, and will be the same ones that end it. All the social, philosophical stuff are just man made constructs, which are subsets of what the physical laws of the universe allows.

                  You discuss these concepts like an economist who thinks money solves everything and we can replace every physical aspect of our existence without limits, in a finite world.

                  Everything we need cannot be replaced by a substitute, as per economist thinking.

                  If you believe substitution to be true, then please explain what humans can replace oxygen with, to breathe. I believe that no social or philosophical thinking will overcome this physical law…

               9. Hideaway,

                  I am not arguing that there is a substitute for everything, but in some cases one form of energy can be substituted for another, aluminum can in many cases be used as a substitute for copper as a conductor, wind, solar and batteries can replace coal and natural gas used for producing electricity, and heat pumps can replace natural gas or oil boilers for heating buildings and water. I have university degrees in both physics and economics, I am well aware of physical laws and am familiar with some social science as well.

                  Why is it that systems become more complex do you think? If there were no advantage the larger more complex systems would not be able to compete with less complex systems. If complex systems lose any advantage that complexity confers the systems will adjust to a more efficient size and complexity, you assume for some reason this cannot occur, I disagree. Social systems are adaptable and constantly changing, there are not fixed social laws as there are physical laws, social laws and norms are devised by humans and are adapted as needed.

                  Yes physical laws are fixed, though note that out understanding of physical laws changes over time and technological and engineering progress allow advances that in centuries past would have been laughed at as the imaginings of a dreamer. Jet flight and smart phones come to mind and even the ease and speed of land transportation realized today would have been unthinkable 200 years ago.

               10. Dennis systems don’t become more complex unless they grow, hence the relationships between size and complexity of systems. The bigger the system the more complex it becomes..
                   The growing complexity helps solve problems withing the growing system, and as such allows the system to become more efficient. As I’ve previously explained look up the work of Prof Geoffrey West to learn more about this. There are power laws throughout nature where a system double the size needs only a 75% increase in energy and materials, or thereabouts. Human settlements correspond to the 85% power law for every doubling in size of the physical settlement. It’s more efficient.

                   An Elephant is much more complex than a mouse with many more branches of internal networks, and especially brain neuron size and function, but as it’s a million times the size of a mouse it only uses 750,000 times the resources (food, water) of a million mice because of the efficiency gains in it’s huge complex size.

               11. Hideaway

                   I’ll apologize in advance for not yet taking the time to dig up your G. West reference material. I will do so, but would like to ask a clarifying question first if I could.

                   You note that … “Human settlements correspond to the 85% power law for every doubling in size of the physical settlement. It’s more efficient.”

                   I read this as indicating that a doubling of population size and/or built environment somehow only requires 85% more incremental resources/energy. Is that your intent here?

               12. Hideaway,

                   You noted that … “there are power laws throughout nature where a system double the size needs only a 75% increase in energy and materials, or thereabouts. Human settlements correspond to the 85% power law for every doubling in size of the physical settlement. It’s more efficient.” You cite Geoffrey West as your source for this.

                   In short, I’m skeptical at best.

                   It appears likely that the underlying primary source for this is:

                   Bettencourt, Lobo, Helbing, and Geoffrey B. West (2007) Growth, innovation, scaling, and the pace of life in cities. PNAS, vol. 104, no. 17, 7301–7306

                   The authors note the well know power law relationship of many physiologic characteristics with body mass. That is well established. However, they then suggest that there is a quantifiable metaphor with cities and population. Not “size of the physical settlement” as you note. Population.

                   They do not provide good support for their data sources, either in the paper or in the linked source material. They claim that their conclusions are “being shared by all cities belonging to the same urban system and sustained across different nations and times.” However, their timeframe appears to be centered very tightly around the year 2000, with the majority of data coming from the US with a smaller leavening of data from China and Germany.

                   They have a serious boundary value problem to start with. I know you are familiar with the work of Rees, as you have cited him here. I would suggest that his work on ecological footprints is much more fundamental and representative of reality.

                   It is also hard to see how a data set for what is essentially a single point in the human experience is supposed to provide a valid basis for such a conclusion across all cities, populations and times.

                   They go on the show a number of power law exponent values representing this supposed power law relationship with population. The two in the 85% range you note are for ‘length of electrical cables’ (0.87) and ‘road surface’ (0.83). These two look like they are only for some ill-defined data set from Germany for 2002, and their approach to the analysis is flawed from the start.

                   Nope, not buying it from what they show.

                    

               13. Hi T Hill,
                   Firstly, when I was discussing size of human settlements, it was always population size, nothing else and there has always been a saving in materials and energy, ergo efficiency by having a larger group and specialization, plus in all the internal networks, just like any organism.

                   However I also agree that in modern times no town or city is in isolation by itself in the modern world, which is a weakness in their arguments, but no reason to throw a huge amount of research out the window. I believe the relationships they have discovered would have been far better as separate entities many hundreds or even thousands of years ago before we humans set up networks of trade.

                   In the modern world, we have 6 continent supply chains with huge quantities of goods being transported across the planet and many networks that make any town or city reliant upon others for parts of their existence.

                   As you would no doubt understand a quick look at a paper to “buy it” or not will never give a full understanding of the overall situation. There are also many scientific papers discussing the same phenomenon of size complexity laws in other social animals like ant colonies, or wasp colonies, plus research into storm complexity related to size, and again separately star size complexity physical laws. Just because you don’t like the implications of such a law of nature, doesn’t mean it should be rejected.

               14. Hideaway

                   You note that “… there has always been a saving in materials and energy, ergo efficiency by having a larger group and specialization, ergo efficiency…”. This is a common assumption. Proof is harder to come by, and in fact research shows mixed results on this point. One thing that West and his colleagues do get right is that cities tend to concentrate wealth. Wealth tends to increase consumption (to a point), resulting in the use of more materials and energy. Your assumption is particularly in question if we use a physics based definition of efficiency as the ratio of the output energy to the input energy and consider more than a single narrow aspect of cities.

                   I’m probably as susceptible to confirmation bias as anyone, but when I said I didn’t “buy” it perhaps I was being too informal. I noted acceptance of the biological power laws. I don’t accept the attempt by West & co. to develop similar relationships for the built environment associated with cities. To be more specific, one fatal flaw in their work is their boundary conditions and the use of Metropolitan Statistical Areas as city boundaries. For a thought experiment, imagine drawing an impenetrable wall at the MSA boundary. For their conclusions about ‘length of electrical lines’ or ‘road surface’ to be valid, the city must be able to function without these features outside of that wall.

                   West and co. do not present a “huge amount of research”. I’m throwing out their conclusions because of underlying problems with it and other research that points in different directions. It’s not about liking the results or not.

            3. And thus begins and extensive back and forth between Dennis and Hideaway.
               I summarize it as: Dennis: Efficiency! Hideway: Not!
               I have been listening to a number of podcasts about AI and what is in store from us, ranging from Meh to We’re all gonna die. So I had chatGPT open at the time, and ever curious I asked chatGPT whether AI would produce more oil for us.

               What followed was an extended and fairly nuanced answer, which indeed captured in its first two sentences both of these positions. “AI itself cannot directly produce more oil because oil production relies on physical extraction from geological formations. [Hideaway] However, AI can significantly enhance the efficiency and yield of oil production processes through various applications: [Dennis]”

               It then went on to outline the ways in which AI will affect oil production.

               1. Duanex,

                  The oil resource is limited, on that I agree. The disagreement (mostly with Kengeo) is the correct number. Rystad estimated about 1500 Gb of remaining resources as of Dec 31, 2023 (published in July 2024), cumulative production at the end of 2023 was about 1500 Gb, this leads diectly to my 3000 Gb estimate which happens to be what the USGS estimated for conventional oil URR back in 2000, unconventional is likely to be around 200 Gb minimum (this includes both tight oil and extra heavy oil which has API gravity of 10 degrees or less) which gets us to 3200 Gb based on the USGS estimate. My guess is 2900 to 3300 Gb (to account for some potential reserve growth as technology and knowledge improves over time).

   5. Several Hubbert curves, the curves between 2.5 and 3.5 Tb seem to fit best for data from 1990 to 2023.

      

      1. Hi Denis and others! The more I see historical world production curve, the more I think Hubert curve can only be applied until 1970-1975. After that period, the growth is linear in an attempt to still increase production while fighting against natural decline. The production being flat for about 10 years now, we may foresee a deep decline in the coming years.
         The production curve will not be symmetrical as in Hubert curve as technology helps to kind of anticipate today a future production. A bit more like negative binomial distribution.

         1. Hi Chris,

            If we use data from 1896 to 1975 to fit a Hubbert curve we get a URR of 12 Tb, if we do a fit on 1896 to 1980 the URR falls to 3 Tb. So it does not work out as you expect.

            

            1. Dennis clearly one of the stupidest graphs you have ever drawn. If you had extended the peak and roll over from the actual prior to 1970’s graph and matched it with known discoveries, the peak would have been around 30 years earlier and much, much lower.

               We know statistically that the inflexion point on the way up is around 1 standard deviation in time away from the peak, and the distance to 2 standard deviations to the left can be calculated from existing data by the percentage of the total production under the curve.

               The simple reality is that we were undergoing a normal distribution curve prior to 1973, and have had a linear increase ever since.

               This huge change came because the US with unbridled capitalism allowed the fast depletion of the valuable resource, while OPEC decided to gain a greater share of their resource by hiking prices. The world suddenly woke up to limited resources, improved efficiency, and made substitutions for oil use.

               The natural curve we were on, changed to only a linear increase from that point and has been on this linear increase ever since, until we hit physical limits where we can’t increase future production at all. I agree it’s most likely sometime in the near future.

               We have already crossed the point where the natural increase in the prior to 1973 gaussian curve would have peaked and rolled over, probably by around 20-25 years, and it would be on the downslope now, just as conventional C+C actually is!!

               The point we are approaching is where the from that natural curve would reach 1 standard deviation on the right hand side of the curve when the acceleration to the downside increases.

               The linear increases in oil production since around 1973, greatly reduced the natural rate of oil use in the late ’70’s through to around 2000’ish, which would put peak further into the future and of course made all peak oil predictions of the time look silly.

               The chart Kengeo put up at 11/26/24 at 11.34 am is far closer to what you should have drawn here, in reply to Chris. I do not take Chris or anyone else here to be a fool, but your chart above is treating him as one which is ridiculous.

               This above chart clearly shows you don’t want to use the actual numbers from before that 1972-3 period as the basis for any curve, because it would give results you don’t like…

               1. Hideaway,

                  Just fit the logistic to the data as I said. The curve used is not a Gaussian, it is a logistic.

                  Chris said we should use the data to 1970 to 1975, that is what I did, used data from 1896 to 1975, the curve charted is the result. I agree it is ridiculous, that was the point which I though could be left unstated for the intelligent readers here. The problem with such an analysis is there are many different results that are possible. Yes growth was exponential for many years and switched to linear growth after the oil shocks of the 70s and 80s.

               2. A logistic curve has fat tails, the early tails on the left of your logistic curves do not fit the reality of what happened prior to the mid ’70’s, not remotely close. A gaussian fit is far, far, better, as it matches the data…

                  Since the mid ’70’s we have had a linear increase in C+C, so it doesn’t fit any curve at all!!

                  Taking the existing curve up to the mid ’70’s into the future using all known parameters of normal distribution curves and the discovery of oil up to that point (which is well over 80% of all discoveries), gives us a picture of what would have happened without the sudden change to a world realising oil was limited.

                  The really nonsense bit is that most people in the peak oil world try to make the linear aspect of oil growth conform to a curve, when it clearly doesn’t, while ignoring the actual data that did confirm to a normal distribution curve..

                  As I stated earlier, Kengeo’s curve that hits peak in the year 2000, looks far more accurate than most, with current high production of C+C dragging future use into the present…

               3. Hideaway,

                  The logistic fits fine over the 1896 to 1973 period. Here is data vs logistic.

                  

               4. Logistic on 1985 to 2023 data. B is data J is Logistic. Data from 1973 to 1985 is not linear, R squared over that period for OLS linear fit is about 0.3. For comparison 1985-2019 has an R squared of 0.98. Shorter periods might lead to lower R squared so comparing 1973 to 1992 to 2000 to 2019, the first period has R squared of 0.68 and the later period (both 20 years) has an R squared of 0.94. Also if we compare the 13 year periods 1973-1985 and 2007-2019, R^2 was 0.31 for the early period and 0.94 for the later period.

                  

               5. Hideaway,

                  The shock model accounts for discoveries explicitly, note that time from discovery to first output varies, picking 30 years is not a bad guess, but the timing has changed over time, from peak discoveries and peak output for conventional oil was about 53 years (1963 and 2016). I use about 39 years for average time from discovered resource to producing reserve.

10. https://www.msn.com/en-us/money/markets/austria-says-russias-gazprom-will-cut-off-natural-gas-supply-this-weekend/ar-AA1uagZR

    Russia cuts off natural gas exports to Austria

11. https://www.reuters.com/business/energy/argentinas-vaca-muerta-gas-export-plan-is-pipe-dream-2024-11-21/

    Vaca Muertas gas export plan a pipe dream

12. I think there’s a lack of understanding of how LNG economics works. When the HH to destination price diff spikes from $10 to $20, it doesn’t drive near term exports to increase. Because the facilities are running 24-7 365 Christmas and Easter already. And it takes 5+ years to get a new facility done.

    Similarly when price diffs decrease from $20 to $10, it doesn’t decrease the amount of export. Because running the facilities still makes sense.

    A huge amount of the cost of LNG is fixed cost (sunk cost). So once it’s built you run it. Even at prices below what is required to pay off the capital. Because the cash profit is still positive.

    These are rough numbers (old and rounding and to make the point and the costs are likely slightly high, but directionally correct).

    If you figure HH at $3, it costs about $10 to liquefy and transport. Of the $10, you’re looking at about $6 of capital cost, $2 of variable cost (opex), $1 to transport, and $1 of profit. So, at $13 in the destination point, you’re sitting pretty. If they drop from $26 in Europe to $13 in Europe, you still run.

    In fact, you still run until prices drop to $6. This is because the capital cost is sunk. You have to pay the bonds even if you turn off the facility. Only the cash operating costs for the facility (mostly energy) and the shipping cost are non-sunk. (avoidable).

    An LNG plant is sort of like a producing well. Once the well is drilled, prices have to drop very low before you shut it in. Or if you want a closer example, think of oil sands megaprojects. They are also mostly upfront capital. Even if the project ends up being negative NPV (a bad investment), it still makes sense to run the facility until prices get extremely low (and the cash profit is a cash loss).

    1. Nony,

       The price differentials drive demand, it may be that there is not always adequate demand for LNG facilities to run near full capacity. Obviously new facilities take time to build. Generally the facility might run at zero profit, but once profits fall to less than zero, it makes no sense to continue running. It is not clear that HH prices will continue at $3/ MMBTU. The long term average (20 years) from November 2004 to October 2024 for HH spot prices is $4.31/ MMBTU. If we look at long term average of TTF to HH from 2005 to 2023 (data from EI Stat Rev of World Energy) the average spread was $4.90, if they need a long term spread of $10/ MMBTU for full cycle profits, it seems a very risky investment indeed to expect that over 40 years. Even if we only look at 2021 and 2023 (ignoring 2022 due to war premium that year) the spread is about $11/ MMBTU, enough for a profit, but probably not sustainable.

       With likely future World overcapacity in LNG export facilities this seems a very risky investment indeed.

    2. Nony,

       Here are the average export prices for LNG from the US, if your numbers are roughly correct this does not look like a good long term business.

       

13. Of course the Biden administration did not restrict exports from existing facilities. I never claimed the converse. We were discussing new facilities and I mentioned the risk premium for them. And if you read the entire article, you’d see there are elements of the left wing of the Democratic party that want to stop even existing facility exports. (And there is some legal power for the government to do so.)

    It’s sort of similar to fracking. Yes, Biden didn’t shut down all fracking. And yes, Harris changed her stance from 2019 to 2024 on fracking when she needed to politically. But the industry sees that and makes some inherent calculation of risk premium. It’s not that much different from overseas investments where there is a risk of expropriation.

    And that doesn’t mean projects don’t go forward. Just that investors demand more return to compensate for the risk. Which affects projects on the margin.

    And it’s a rational reason to ask for a risk premium. There have been several examples where billions of dollars were spent by the industry and then lost because of government decisions. One example is the billions spent on NYS lease bonuses in the Marcellus, before the frack ban was put in place. Another is the Constitution Pipeline. Another is Keystone.

    1. Nony,

       There is always risk in any investment. In my view the bigger risk for LNG export terminals is that supply may not be available at prices that allow a profit. Worldwide demand for natural gas may wane as cheaper forms of energy supply become available that do less damage to the environment. A climate that enables food to be grown on the planet is thought to be a plus by some.

       1. Dennis

          One of the big unknowns with regard to electricity demand is the push into AI. I keep reading articles that claim wind and solar will not be able to keep up with AI demand. NG will be required to fill the demand for electricity and the chemical industry. These articles are basically opinions and don’t offer facts.

          1. Ovi,

             Data centers will be built in places with the lowest electricity costs, currently in the US that is Texas which has a lot of wind power and has been expanding solar power quite rapidly. Whether supply grows as fast as demand remains to be seen.

             I have also seen those reports. The data for annual net electricity generation from 1990 to 2023 from the EIA is below, using link below.

             https://www.eia.gov/electricity/data.php#generation

             The average annual rate of increase from 2005 to 2023 was about 4.4 TWh per year (OLS Linear regression). Roughly a tenth of a percent per year on average over the past 18 years (based on an exponential trendline using OLS).

             

             1. Dennis

                In Canada, many crypto miners are based in Quebec because they have cheap hydro power, Quebec is also home for a big Aluminum smelter because it uses a lot of electric power

                1. Ovi we could see the same in high wind and solar areas of the US like midwest(wind) or southwest(solar), Texas has excellent wind and solar resources so is ideal, probably why they have very low wholesale power rates.

          2. The data centers for cloud and AI will pay very high rates for electricity that is 24/7/365.
             These data operations are/will be handling functions such as all financial transactions, security and military operations (internal and external), all medical information systems, mechanical automation operations of industry and commerce, air traffic control, autonomous land transport, education,media and communications, and don’t forget the big new emerging tool of collective thought control.

             Nat gas will play a leading role in the US electrical system, with healthy contributions from solar, wind, hydro, nuclear and some coal. Perhaps neo-Geothermal as well. Maybe the outlook will appear a little different in the 2040’s. No guess made on that.

             1. At the World level from 2011 to 2023 electricity output grew at about 2.4% per year with non-fossil fuel electricity output growing at about 4.4% per year over that same period. Fossil fuel may continue to play some role as backup, mostly in natural gas peaker plants, but nuclear, solar, wind and batteries will gradually replace fossil fuel energy over the next 2 to 3 decades.

             2. If we assume the growth rates for total electricity anf non-fossil fuel electricity follow recent growth rates in the future we get this. Looks like 5 decades with these conservative assumptions, my guess is that non-fossil fuel electricity growth may eventually reach 7% per year. If correct, fossil fuel electricity generation is replaced with non-fossil fuel by 2047 (23 years in the future).

                

                1. What source of data did you use for current elect generation…actual generation or capacity? Thank you

                2. I used electricity generation from the Energy institute, I converted TWh to EJ. Fossil fuel generation is the sum of oil, natural gas and coal electricity generation, and non-fossil is total minus fossil. I ignore capacity, I don’t find that a useful measure.

                   1000 TWh= 3.6 EJ (1 TWh= 0.0036 EJ) or 1 EJ=277.78 TWh.

14. Ovi – I’d say it’s a pretty well known issue already (examples below):

    1.Ashburn , Virginia, USA: Known as “Data Center Alley,” Ashburn hosts a substantial concentration of data centers, consuming approximately 25.59% of Virginia’s total electricity in 2023. This immense demand has raised concerns about grid capacity and the environmental impact of such concentrated energy usage. 

    2. Dallas-Fort Worth, Texas, USA: The region has experienced a significant increase in data center development, with over 150 facilities established by major tech companies. This expansion has contributed to a 13 TWh rise in commercial power demand between 2019 and 2023, leading to uncertainties in future wholesale power prices and grid stability. 

    3. Atlanta, Georgia, USA: Atlanta is emerging as a major data center hub, prompting local utilities to plan new gas plants to meet the growing energy needs. This development has sparked debates over the environmental implications and the strain on the city’s power infrastructure. 

    4. Dublin, Ireland: Dublin’s rapid data center growth has led to projections that these facilities could account for up to one-third of Ireland’s electricity consumption by 2026. This surge has raised concerns about the national grid’s capacity to handle such demand and the potential impact on energy prices and sustainability goals. 

    5. Amsterdam, Netherlands: The Amsterdam area has faced grid capacity issues due to the rapid expansion of data centers, resulting in delays and restrictions on new projects. Authorities have implemented stricter regulations to manage the impact on the power grid and environment.

    6. Singapore: Singapore imposed a moratorium on new data centers in 2019 due to power constraints. Although the ban was lifted in 2022, the country continues to face challenges in balancing data center growth with sustainable energy consumption.

    7. Northern Virginia, USA: The region’s dominance in data center capacity has led to concerns about whether residents will have to foot the bill for future power lines, highlighting the strain on local infrastructure.

    1. What isn’t news, is that when any meaningful development that requires power is proposed,, the world is being destroyed army will be out in force to try and stop it. Except in Red China.

15. Kangeo

    Thanks

    Good stuff. Have you been following this issue for a while to have this info at your finger tips?

    1. Nope, just know people in that industry who are dealing with the headaches…
       #5 being one of the more recent issues

    2. Ovi- We all have that info at our immediate finger tips.
       I suggest this portal for search
       https://www.perplexity.ai/

       Just ask a specific question…the relevance of the output is all about giving a good prompt.
       It gives a great summary, as well the link to its most relevant source material.

       but there are of course other similar tools.

       1. Electricity demand

          https://www.perplexity.ai/search/electricity-demand-world-futur-YTTs9bZaTLa2UtR3GGMAyw

16. The Rig Report for the Week Ending November 22.

    
– US Hz oil rigs increased by 3 to 435. They are down 24 rigs from April 19 and are up 8 relative to their recent lowest count of 427 on July 24th. Three of the rig increase were in Texas.
    
– In New Mexico, Permian rigs increased by 1 to 94 while the Texas Permian was unchanged at 190.
    – In New Mexico, Eddy added 2 rigs while Lea dropped 1

    – In Texas, Midland and Martin were unchanged at 26 and 27 respectively. The rig count for these two counties has not changed for the last 4 weeks.

    – Eagle Ford held steady at 40.

    – NG Hz rigs dropped 2 to 84.

    

17. Frac Spread Report for the Week Ending November 22

    The frac spread count decreased by 1 to 221. It is also down 60 from one year ago and down by 51 spreads since March 8.

    

18. A quick survey for those of you here who know better than I:

    T or F. “The quicker we extract oil now, the steeper the decline on the other side of peak.”

    I’m guessing true for wells & fields but false for the world.

    (In the inverse way that I never thought I’d live to witness the effects of climate change, I never expected the oil extraction rate to roar on for so long.)

    1. For the world, it could go either way since global production is aggregate of all fields. If majority of oil fields are overproduced and reserve growth does not happen, then global production will match that of the fields which best represent it, which would be steep declines.

       Example: Overproduction in fields like Mexico’s Cantarell or U.S. shale wells, where high initial output leads to steep declines.

       Looking at it, for offshore we might assume 22 mbpd of production will decline 15-20% annually. Losing 3-5 mbpd each year without additions/growth.

       Based on these factors, oil price will likely increase 16% annually for next 5 years. Production will try to increase in response to price, but a loss of at least 2.5% each year will likely ensue.

       1. You’re ignoring what is going on in Europe and China. European banks play a big role in redistributing dollars through out the global economy. As do Japanese banks.

          Without the circulation of dollars there will be no recovery for either Europe or China. The rising dollar is tells you dollars aren’t circulating. It equals global recession. Oil prices will be doing well to stay above $50 over next 5 years. Because there is nothing on the horizon that is going to fix Europe or China.

          Matter of fact tariffs are the horizon.

          1. HHH, maybe time for you to predict WTI at $25 again?

             1. Right around the corner, any day now.

    2. In a more extreme scenario, production loss might occur very rapidly with 10 mbpd for 2 successive years. An example could be losing Saudi Arabia and Russia at the same time, or losing a large number of small producers.
       Global exports peaked from 2016-2019 and have not returned to those levels, while unlikely if they make it back to those levels, it wouldn’t be until 2026/2027 at the earliest…which in spite of a 2.5% decline rate, seems improbable…

       Russia for example is declining 2% annually since 2019, or more recently about 3.5%. Russia likely has used almost 70% of their overall reserves.
       Applying similar for the global conventional:
       50% URR in 2013, 1.5% annual decline
       60% URR in 2020, 2.5% annual decline
       70% URR in 2025, 3% annual decline
       75% URR in 2029, 5% annual decline
       80% URR in 2034, 7% annual decline

       1. Kengeo

          While I think your concerns are very real, there is an argument to be made that technology has made the break even price lower or are in the process of making them lower in many cases for both oil and gas. But not in all cases – in the end geology would set the limitations.

          What I am talking about is the ability to utilise advanced software. A part of AI is to run a scenario a million times if necessary with too many inputs the way it was before to make any conclusions what so ever. Not true anymore. The ability to interpret data make seismic data, reservoir data and not at least to judge where exactly to drill much more valuable. A lot of questions arise if these assumptions are taken to heart. We are extending a plateau phase for oil probably (since this is the topic). There will eventually be an immense pressure for virgin or not good enough developed assets to apply better technology.

          Too low prices is not good as well…taking energy for granted (toolkits for good low energy lifestyles need to be developed further).

    3. Mike B,

       For a given URR if peak occurs at a higher proportion of URR (say 60% of URR), then decline will be steeper than if peak output occurs at 40% of URR. So I would say true for the World. On problem is that we don’t really know what URR will be, my guess is between 2.8 and 3.2 Tb with a best guess of around 3 Tb for World C plus C.

19. Oil production in Azerbaijan is expected to reach 29.2 million tons in 2024
    25 November/ 14:25

    Baku. In 2024, oil production in Azerbaijan is expected to reach 29.2 million tons. This was announced by Economy Minister Mikayil Jabbarov.

    In 2023, 30,189.6 thousand tons of oil were produced in Azerbaijan. Thus, in 2024, the figure may decrease by 3.3%, Interfax writes.

    Jabbarov recalled that the peak of production was recorded in 2010 and amounted to 50.8 million tons.

    “In 2014, oil production in Azerbaijan amounted to 42.1 million tons, in 2018 – 38.8 million tons, in 2021 – 31.6 million tons,” he said.

    The Minister also noted that the average price of Brent crude oil this year is expected to be around $80 per barrel, compared to $99 per barrel in 2014

20. When attempting to work out how much oil is left, it always bemuses me about why people are trying to use a gaussian curve to data that clearly no longer conforms to a normal distribution. It’s statistical nonsense..

    From 1920 to the early ’70’s oil production was clearly following a normal distribution curve, which Hubbert observed and made his calculations upon. Since the mid ’70’s we’ve had 50 years of linear growth at a relatively constant rate, making a totally different shape to the production graph. Trying to ‘fit a curve’ to something that is no longer a curve, gets more ridiculous as every year/decade passes..

    If you really want any accuracy with a normal distribution curve, then only use the first 50 odd years when the data conformed to the gaussian model, then project froward using known parameters of changes in angle at certain statistical probability levels to give final known dates if not actual numbers under the curve. I did this myself a few months ago, with peak of a normal distribution being well behind us and the acceleration downward, past any plateau starting around 2026-29.

    Of course the linear increase in oil production since the mid ’70’s has meant we’ve brought forward future production into the present, in the latter half of that period, while reducing production growth in the early part of the linear increase.

    Make no mistake, that when production does start to fall, the acceleration to the downside will be rapid, and it’s easy to tell because all the newer more technical forms of oil production all have steeper depletion profiles, from fracking to deep sea wells.

    Is there any chance of people everywhere, not just here, stopping the ridiculous curve fitting to data that hasn’t followed a normal distribution for over 50 years??

    1. Hideaway,

       It is not a gaussian fit, it is a logistic function which describes the cumulative output see

       https://en.wikipedia.org/wiki/Logistic_function

       https://en.wikipedia.org/wiki/Hubbert_curve

       also there is this

       https://sepwww.stanford.edu/sep/jon/hubbert.pdf

       1. Hideaway,

          I use the Oil Shock Model developed by Paul Pukite, see this post for a simplified explanation

          https://oilpeakclimate.blogspot.com/2015/02/the-oil-shock-model-with-dispersive.html

          The online reference is no longer available, but at your local university library you may be able to find this book, where the model is discussed in detail.

          https://agupubs.onlinelibrary.wiley.com/doi/book/10.1002/9781119434351

          1. Dennis –
             Why in the world would you model Tar Sand and Shale along with conventional?!?

             Keep it simple
             1500 Gb produced
             200 Gb OPEC remaining
             100 Gb non-OPEC remaining
             100 Gb unconventional remaining
             100 Gb discoveries growth remaining (mostly unc.)
             URR (P90) 2,000 Gb

             Hubbert peak excluding unconventional:
             2005-75 MB/d (actual was 81.8)

             With unconventional:
             2007-82 MB/d (actual was 82.5)

             We know that conventional oil peaked from 2008 to 2018 timeframe with an average production for those 10 years of 80.7 MB/d. Peak was 2015 at 82.7 MB/d.
             This implies a URR of 1.9 Tb.

             The best fit for a Hubbert Peak is 1.8 Tb.

             Between 2016 and 2022 URR peaked at a value slightly greater than 1.7 Tb. If 2P is used URR is ~2 Tb.

             For 2.6 Tb a peak of 105 MB/d is required
             For 2.75 Tb 111 MB/d
             For 3 Tb 121 MB/d

             Based on the observed peak of 82.7 MB/d we know that conventional URR must be ~1.8 Tb, if we add another 0.15 Tb for unconventional we get a URR of 1.95 Tb. This matches the 2P reserves of ~450 Gb.

             1. Kengeo,

                You would need to show your work, your peak values are nonsense. You would need to show your work, you seem to be picking random numbers. The peak in 2018 was 82.9 Mb/d see

                https://www.eia.gov/international/data/world/petroleum-and-other-liquids/annual-petroleum-and-other-liquids-production?pd=5&p=0000000000000000000000000000000000vg&u=0&f=A&v=mapbubble&a=-&i=none&vo=value&&t=C&g=00000000000000000000000000000000000000000000000001&l=249-ruvvvvvfvtvnvv1vrvvvvfvvvvvvfvvvou20evvvvvvvvvvnvvvs0008&s=94694400000&e=1704067200000

                Cumulative C plus C output was 1358 Gb at the end of 2018, implying a peak of at least 2716 Gb, but note that peaks are often temporary and there was a cut back in production by OPEC in 2019 followed by a pandemic as you may remember.

                Let’s consider a previous peak in 1980 at 59.6 Mb/d which was not surpassed until 1989 (59.8 Mb/d), cumulative output was 425 Gb at the end of 1980, implying a URR of 850 Gb, obviously that was wrong then (with the benefit of historical output).

                Note that I don’t think the peak necessarily occurs at 50% of URR, assumptions of symmetry are likely flawed. URR estimates based on 2P reserves will always be underestimates, this is the reason Laherrere has underestimated URR on many occasions. The Rystad 2PCX estimate of about 1536 Gb at the end of 2023 is probably a reasonable place to start, add cumulative production of 1503 Gb and we get 3039 Gb for a rough guess for URR, but note that over the past 6 years that reserves (as measured by 2PC resources) have increased by about 30 Gb per year (about 180 Gb over 6 years), will we see further growth in the future? Perhaps URR increases over time to 3200 or even 3500 Gb, much will depend on future demand for oil resources.

                1. Dennis –

                   It’s simple, you can use BP statistical review data.
                   I calculate unconventional oil and see that it peaked between 2005 and 2017. If match peak and current production we get a URR of 1.7 Tb. We we use Rystad 1P reserves for 2023 plus production we also get 1.7 Tb. The unconventional oil likely has another 100 Gb of 1P, so if you want to be generous you can call URR 1.8 Tb. This matches the data, I think you have myriad excuses why data shouldn’t match.
                   This is pretty much my same argument from several years ago, the difference is that it’s now backed up with data from 2-3 additional years of production. You will note that US growth over 12 months is zero, yet another reason to expect only decline moving forward. Between 2023 and 2024 world production have likely fallen by 3%, this will accelerate moving forward. Even if you wave your arms about discoveries and growth, that might only get you another 200 Gb, now you are at 2 Tb….but definitely not 3 Tb. You know better than to evaluate conventional and unconventional oil together, not sure why you do it, my guess is obfuscation, let’s you continue along the same nonlinear thought process that gets you to 3 Tb. Time will tell.

                2. Kengeo,

                   When I do the analysis on conventional and unconventional, you are not convinced either, I have done that before, you didn’t get that either.

                   It is not obfuscation, I am simplifying the analysis

                   here is a link to my work

                   https://docs.google.com/spreadsheets/d/1Hr3GaUesVWKQmlW5u95Y5z4powLMRovU/edit?usp=drive_link&ouid=105320434049434900507&rtpof=true&sd=true

                3. Kengeo,

                   Yes time will tell, I think 2800 to 3200 Gb is the right range with a best guess around 3000 Gb. You have a bunch of different guesses, but seem to be focused on 2000 Gb, the World reached 1000 Gb in 2005 and Deffreyes claimed the peak had arrived, almost 20 years later and we still have higher output than the 2005 “peak” with an additional 500 Gb produced. We may be near the middle of a peak plateau which started in 2016 and may end around 2030 with output wandering between 81 and 83 Mb/d over that period.
                   That scenario matches the Rystad 2PCX URR of around 3000 Gb. Assuming no future discoveries, reserve growth, or contingent resources becoming reserves could lead to underestimates.

             2. Kengeo,

                Or we could look at Rystad estimate of 1200 Gb for 2PC resources plus 1500 Gb of cumulative production which gets us to 2700 Gb (with no discoveries or reserve growth) now assume 10 Gb per year on average of discoveries plus reserve growth (3 times less than the average over the past 6 years) over the next 30 years and we have increased the URR to 3000 Gb.

                Alternatively we could use the Rystad 2PCX estimate (which includes disoceries and reserve growth expected in the future (this is 1500 Gb) and add the cumulative production of 1500 Gb and we also arrive at 3000 Gb. This estimates is consistent with a peak of 82 Mb/d, your estimate of a 121 Mb/d peak being needed for 3000 Gb is ridiculous.

                I think I see where our approach is different. You seem to be using all the data to do a curve fit from 1900 to 2023. Typically the Hubbert analysis is trying to predict the future rather than the past, so we look for a curve that matches the past 10 to 20 years, for my analysis I try to match 2004 to 2023 (20 years), this results in a URR of 2990 Gb. If we try to match the entire data set from 1900 to 2023, we do indeed get a URR of roughly 1870 Gb with a peak in 2003 at 77 Mb/d, in 2023 this Hubbert curve has World C plus C output at 54 Mb/d with cumulative output at 1454 Gb, so not very helpful predicting future output (28 Mb/d too low or about 34% too low).

                In general I don’t think the Hubbert curve is all that useful (or the related Hubbert linearization.)

          2. Dennis how is it you and everyone else is so blind to the 2 halves of the oil production graph over the last 100 odd years??

             Prior to the early ’70’s it was obviously a gaussian fit, with shallow tails on the very left. A logistic function has fat tails out to 4 standard deviations. Have a good look at what you drew up thread, the first half of the graph..

             https://peakoilbarrel.com/wp-content/uploads/2024/11/783358-1.png

             Your logistic function clearly just crosses a gaussian normal distribution curve, then the linear trend since mid ’70’s just happens to be on the logistic function chosen at that point, pretty much a straight line since. It’s definitely not just you that does it, it’s throughout the peak oil world and it’s poor use of statistics.
             The first half of the graph is the unfettered growth phase that tells us the most, not the last 50 years where growth was lower than a pure market would allow, while the last half has been dragging future use into the present. The nice gentle slope on the way up of the logistic function, simply didn’t happen, so why expect the downslope to be nice an gentle?

             

             1. Hideaway,

                I agree the logistic function does not fit the production curve very well, nor do I expect the Hubbert curve will be followed on the downslope. The change in the curve after 1974 wa due to a change in OPEC behavior and generally higher oil prices than before the Arab Oil embargo.
                In chart below dashed line is Oil shock Model scenario compared with a Hubbert Curve fit to 2005 to 2023 World C plus C data, both scenarios have a URR of about 3 Tb.

                

                1. Units for chart above are millions of barrels per day for vertical axis, horizontal axis is year (CE).

    2. Hideaway,

       Would you please share the work you report on forecasting future production? It would be interesting to hear your assumptions and see your predicted decline rates.

       1. Hi T Hill, I did the work with my daughter a couple of months ago, she is a university lecturer of statistics to post grad students, so I’ll have to go back over all the data we used so might take a bit of time to put it back together, as it’s a busy time of year for her.

          I didn’t save it at the time as it really just shows we reach an inflection point in a few years where the decline in production of C+C accelerates downwards 2026-2029 period with an error margin of something like 10%, which is nothing different to most ‘guesses’, except the rate of decline is likely much higher because we are dragging so much future C+C use into the present.
          I’ll put all the numbers up hopefully in the next few weeks sometime.

          It’s just all the stupid curve fitting we read here and elsewhere trying to fit a curve to a linear increase we’ve had for 50 years, while there was a very accurate exponential rise in the prior 50 years, just triggered a response earlier. The graph of oil production over the last 100 years is clearly in 2 halves that don’t match period.

          1. Hideaway,

             Yes the growth in World C plus C output from 1938 to 1972 is exponential at about an annual rate of 7.2%. From 1900 to 1929 growth was also very rapid at an average annual rate of 7.9%. The Great Depression (1930-1938) has slower growth of about 5.5%. Growth from 1982 to 2019 was about 1.2% per year on average, but I agree a linear fit is better with an average rate of increase of 812 kb/d over that 37 year period. From 1973 to 1980 the exponential annual rate of increase was 3.2% on average.
             From 1900 to 1938 the average rate of increase was about 7% per year.

             If we take the 100 years fro 1920 to 2019 and divide in 50 year periods, 1920 to 1969 had average growth of 6.08% per year and 1970 to 2019 had average growth of 1.1% per year or (for a linear regression) 713 kb/d. Note the R squared is slightly higher for the exponential fit than the linear fit for the 1970-2019 period.

21. Hideaway,

    The linear increase was about 37 years 1982 to 2019. The exponential increase was about 70 years 1903 to 1972. Higher oil prices after 1972 led to slower demand growth for oil.

22. It’s necessary to plot a front end curve for the early data, and a back end curve for 2000+ data, both curves with best fit match a URR of 1.7 Tb (unconventional) which matches total production plus 1P estimates.

    

    1. Kengeo,

       No idea what this is, what are the units for the vertical axis?

       1. GTons, can multiply by 20.2 to get mb/d.

          1. Thanks. So about a 71 Mb/d peak in 2000, not clear where you get the conventional estimate. My estimate for 2000 is about 67 Mb/d or 3.31 Gt.

23. Backend curve

    This is on account of the delay in production growth which occurred between 1980 and 1995.

    

24. If URR is increased to 2 Tb, a curve fits the data on average:

    

25. Another scenario where production decline occurs rapidly over the next 5 years.

    

    1. Dennis – Is world production about 1-2 mb/d below your best guess as of 2023?
       Here’s from an old post:
       Best Guess Yearly C+C Production
       …….. Denniis ……. Ron
       2020 ….. 76.0 ….. 76.0
       2021 ….. 77.0 ….. 77.0
       2022 ….. 78.7 ….. 80.0
       2023 ….. 80.6 ….. 79.0
       2024 ….. 82.3 ….. 78.0
       2025 ….. 83.5 ….. 77.0
       2026 ….. 84.3 ….. 76.0
       2027 ….. 84.8 ….. 75.0
       2028 ….. 85.0….. 74.0

       I think 2025 is the year we will see production drop below 80…
       Average of both your and Ron’s guess for 2025 is 80.

       1. Kengeo,

          My current guess is lower with peak around 83.2 Mb/d in 2028 or 2029, actual output in 2023 was 81.9 Mb/d so that guess was too low and currently I think output will again be about 81.9 Mb/d in 2024. My current guess for 2025 is 82.5 Mb/d.

26. “Kazakhstan expects to produce 88.4 million tons of oil this year, down from an initial target of more than 90 million tons, reflecting maintenance of large oil fields and Kazakhstan’s commitment to OPEC+ production cuts. That translates to about 1.82 million barrels per day, based on a ratio of 7.5 barrels per ton. Kazakh exports account for more than 1% of global supplies. Starting in 2026, the country plans to produce more than 100 million tons of oil per year.”

    https://en.usm.media/kazakhstan-will-reduce-oil-transit-through-russia-by-80/

27. The Hubbert curve below uses a piecewise fit to different periods (1900 to 2019) to fit the data best, URR is 3.3 Tb and compared with Shock Model of roughly 3.1 Tb. Probably future output will be between these two scenarios, but possibly less than Shock Model if transition to electric transport proceeds quickly (I am not optimistic enough to believe this will occur). Peak for Hubbert model is 84.8 Mb/d in 2029 and for Shock Model is 83.1 Mb/d in 2029.

    

    1. Dennis – It’s clear that you have an extreme misunderstanding of Hubbert Curves. You should review the fundamental concepts including overall width (2 x time to go from 10% of peak to 100% peak, about 100-120 years). Peak width, around 30 years. In this case we calculate the area under the curve as 2 Tb. The 3 Tb URR you are chasing would need a significantly longer duration 150-200 years or a much higher peak production, or combination of both. It is impressive how you contrive the production curves to fit your model, you’ve graduated from cherry picking to completely misrepresenting the entire theory.

       1. Kengeo,

          Read a little more on the theory, I have it right try 1900-1980 and 2004-2019 and splice the early result up to mid 60s and second result from mid 60s to 2200, the second curve may need to be shifted 5 years to the left to get cumulative production to match (due to over production relative to these curves from 1965 to 1982), the result will be a hubbert model with a peak around 85 Mb/d with URR of about 3170 Gb. This is consistent with USGS estimates, Rystad estimates and even estimates by Laherrere (in 2022). I think this may be a bit high, my guess is around 3000 Gb.

          Usually the models would go from start of production in 1870 to 2200 or so (roughly 300 years), the area under the curve is cumulative production which is easily added up in a spreadsheet.

          Link to my Hubbert model

          https://docs.google.com/spreadsheets/d/186lVXg6mnaPqGlXAtQ1O4PxqOi9Kmkiv/edit?usp=sharing&ouid=105320434049434900507&rtpof=true&sd=true

       2. Chart with model, vertical axis is Mb/d

          

28. Dennis – Even if we inflate the reserves to such high levels as you propose, say 3.3 Tb. Two things are apparent.
    1. Conventional peaked ~15 years ago.
    2. Unconventional peaks in 2025.
    Subsequent production is lower by around 1-2 mbpd annually.
    Even an unconventional amount of 1 Tb does little to the overall Hubbert curve, it’s a case of too little too late. Like getting to the party 50-60 years after it started.
    I believe you are continuing to grossly overestimate the volume of crude oil that is readily available to the economic system. Fortunately electric vehicles are beginning to be produced at fairly high rates.

    Ultimately, here’s what this means:

    Conventional: Production from conventional fields will likely decline steadily by 2030, contributing around 40–50 mb/d (down from ~70 mb/d in 2023).

    Unconventional Oil:these resources could contribute 25–30 mb/d in 2030 if investment continues, but may fall to 15–20 mb/d in a low-investment scenario.

    Do you see investment increasing to be able to produce all of the unconventional reserves?

    1. Kengeo,

       I have tended to underestimate in the past, I hope that you are correct and the world transitions to electric transport faster than I think will occur. There is likely to be continued demand for water transport and air transport even if land transport demand fell to zero, probably about 30 Mb/d of demand will remain for the rest of the century. The 3000 Gb number seems pretty sound, might be a bit lower (100 Gb maybe) in a very optimistic transition scenario. I agree 3.3 Tb is probably too high, but think anything under 2.9 Tb is too low. Conventional peaked in 2016 which is 8 years ago, it is being held back by OPEC to balance the oil market and has been on a plateau. I expect unconventional oil to peak in 2027.

       1. Dennis – You have been beating that sad drum for too long, it’s contributing to your tendency to grossly overestimate URR. Maybe take a fresh look after the new year?

          1. Kengeo,

             A very recent estimate that I produced also underestimated output in 2023, but may have been too optimistic after 2025, it seems demand has been lower than I predicted. Time will tell, I think we will continue to disagree, I don’t expect to change my thinking until 2026 or 2027. Maybe 2025 if we see oil prices crash due to lack of demand or if prices spike and we see output remain flat. I doubt we see either of these happen.
             What strikes me is that you seem to be stuck in 2005 when people thought World URR would be 2000 Gb, even if we just use 2P reserves we get at least 2200 Gb, if we add contingent resources that have been discovered and use the engineering best estimate we add another 500 Gb and get 2700 Gb, this is the minimum reasonable estimate. Then consider that 180 Gb of reserves have been added in the past 6 years and 2900 Gb is not really much of a stretch, what happens over the next 30 years, does discovery and reserve growth stop? Seems unlikely to me.

             1. Dennis, …”Then consider that 180 Gb of reserves have been added in the past 6 years”…

                Could you give a reference for this please. All new discoveries I’ve read about are way lower than this. Are you including NG in this, or just C+C which the conversation has been about?

                1. Hideaway,

                   look at chart below, I believe it is self explanatory.

                   

                2. I am considering C plus C only, not including any NGL or natural gas.

                   The 2PC resource has remained about the same from 2018 to 2024 while about 180 Gb of oil have been produced, this suggests over that time that recoverable resources have increased at about 30 Gb per year on average over that 6 year period.

                   Chart in previous comment is a screen shot from link below

                   https://www.rystadenergy.com/news/global-recoverable-oil-barrels-demand-electrification

                   Note that at the start of 2018 (Jan 1) Rystad was estimating about 3500 Gb of C plus C resources (using 2PCX estimate plus cumulative C plus C production as of Jan 1 2018 of 1327 Gb). Their estimate for URR has decreased over the past 6 years by about 500 Gb an average of about 83 Gb per year.

    2. Kengeo,

       Conventional resources are unlikely to decrease as much as you expect (a decrease of about 2 Mb/d annually) OPEC has spare capacity of 2 Mb/d and conventional output is increasing in Guyana, Brazil, and Norway, I expect conventional will remain roughly on a plateau until 2029 and unconventional output will likely increase in US, Canada and Argentina.

29. NOVI Ted Cross has this Thanks for Fracing!
    https://www.linkedin.com/posts/ted-cross-tx_this-thanksgiving-week-im-giving-thanks-activity-7267178672879484928-uWI4?utm_source=combined_share_message&utm_medium=member_desktop_web

    1. BOE GOR was at 5:3 in 2017~2019, when Appalachian and Haynesville still increase obviously. and now these 2 dry gas fields have stalled, the GOR increase in mainly due to oil basins GOR increase

       

       1. I guess Ted Cross excels as a marketer since I’ve never heard of someone being so audacious on this day of thanksgiving to suggest that thanks to fracking the USA alone produces more than 20 million barrels of oil.

          He should also thank nuclear power in that case because that can be converted to an oil equivalent.

          And he should thank solar cells too.

          And he should thank rivers and gravity too.

          1. Paul,

             He is looking at C plus C plus NGL, in Aug 2024 NGL output was 7030 kb/d and C plus C output was 13400 kb/d. The NGL has lower energy content, lets guess 75% of a barrel of crude which would be about 18670 kboe/d, but Cross was relying on the average person thinking that all barrels are the same.

             1. Dennis,
                He specifically called out fracking-specific sources, which is 8 million barrels/day of crude. I think 2/3 of total NG is also from fracking, so that means he has 13 million barrels equivalent barrels in fracked-only NG for the 21 million number he claims.

                This is all in the context of fracking-only. Apparently the Trump administration wants to up this by 3 million barrels/day. It bet it will all be done by this equivalent book-keeping since no way can they make this up by crude only.

                1. Paul,

                   My mistake, he is talking about barrels of oil equivalent for natural gas and oil from shale gas and tight oil. In August 2024 shale gas output was 13.8 Mboe/d and tight oil output was 8.6 Mb/d so for combined fracking output from shale resources was 22.4 Mboe/d in August 2024.

                2. Paul

                   The original report I saw with Besent’s projected production increase was 3 million barrels per year and yes equivalent so most of it will come from NGLs. It was a double take but then I realized it was basically only 10 kb/d increase. I thought clever since it will catch the headlines. All the new reporting I see is 3 Mb/d, which we know is out of the question.

                   Searched for original release but cannot find it.

                3. Ovi,

                   Note that the barrels of oil equivalent trick includes natural gas, NGL and C plus C. For the US this is close to 38 Mboe/d (as of August 2024), so a 3 Mboe/d increase is about an 8% increase, perhaps over 4 years (2% per year) this might be accomplished, but only with higher prices.

                   Quote by Bessent:

                   Three million more oil barrels equivalent a day from U.S. energy production. That would be my 3-3-3.

                4. Ovi and Paul,

                   For marketed natural gas and C+C converted to boe the average US increase from 2011 to 2024 has been about a 1.22 Mboe/d annual increase in US petroleum and natural gas over that period and about a 5% annual increase on average. So a 3 Mboe/d annual increase is unlikely, but 3 Mboe/d over 3 years might be possible.

30. Dennis why did you leave out a dozen years in the middle?

    How about you show the straight line linear fit from 1973 to 2023? Anyone can pick and chose their starting points and finishing points to prove anything they like. Hence the expression lies, damm lies and statistics..

    If you actually drew the graph with those 2 halves you’ve drawn, just put together, it will look totally ridiculous, which it is..

    The reality is we had unfettered growth in oil production up until the world as a whole realised it was a limited valuable resource we were using up too quickly, with a group (OPEC) able to control price and quell demand as the world tried to adjust to reality.

    Since around 1973 the fairly constant growth rate has never solved the problem of more energy needed by the growing system of civilization, despite use of other energy especially coal and gas replacing the exponential rise in oil production.

    We have had 2 separate halves to the oil age, prior to around 1973 with exponential growth and since then with mostly linear growth, because civilization itself is addicted to oil. Trying to fit a curve to 2 separate halves is nonsensical.

    The first half with unconstrained growth is what gives us the reality of the situation. The managed second half with stupid additions to reserves all at once by OPEC countries, which most likely don’t exist, should never be counted.

    As I’ve clearly stated several times now, Kengeo’s graph showing the natural peak around the year 2000 appear to be most accurate, and if we extended actual production numbers to present, we would see current production well outside the curve, meaning we’re dragging future use into the present

    1. Hideaway,

       I think you are wrong. Look at the Rystad estimates, they reduce the OPEC reserves, I agree the increase in oil prices from 1973 to 1980 changed behavior. The two different models can be connected by actual data over the transition period from 1973 to 1984 which is difficult to model. Doesn’t look ridiculous, it is what happened. In 2024 US$ the crude oil price increased from about $24/b in 1972 to $129/b in 1980 (annual average price) about a fivefold increase in price. So you are absolutely correct that demand for oil was reduced by the higher oil prices after 1972. I showed the the “linear” part of World production is matched fairly well by a logistic fit to 1985 to 2023 data. I am not wedded at all to the logistic, it is a special case of the more general Oil Shock model.

       So the “natural peak” is around 2000, I prefer to deal with the World as it is rather than how I wish it to be. The actual peak (so far) was 2018, I expect we might see a new peak by 2027 or so. I find it interesting that a fit on 1896 to 1980 results in a URR of about 3000 Gb with a peak in 2006 at 146 Mb/d. Rather than production being pulled forward, it has been pushed into the future.

       I would love for you and Kengeo to be correct as it would result in less damage to the environment, I try to look at things as they are rather than how I would like them to be. World C plus C URR is more likely to be 3000 Gb than the 2000 Gb that Kengeo wishes it will be, it might be as low as 2900 Gb in a very optimistic EV transition scenario perhaps with autonomous vehicles helping to speed the transition. I doubt that is a realistic scenario, but have been wrong many times in the past, I expect that will continue in the future.

       If we divide 1973 to 2019 into thirds we get an r squared of about 0.47 for 1973-1988, about 0.92 for 1989-2004, and 0.90 for 2005-2019, the earlier period is not as linear as the second two periods. In addition, just looking at the fit the years from 1973 to 1985 do not look very good relative to the OLS trend.

       1. Dennis, your logistic curve that crosses the pre 1973 period once and is way too high in the fat early tail, and way too low from where the exponential growth in the ’60’s races above it, does not follow the reality of the actual data.

          Just because you can make the straighter part of the logistic curve fit one place on the graph, when we know growth has been constrained early and is now linear, certainly does not mean the logistic curve is a better fit. Just look at the future fat tail, it does not come close to correspond with the past actual data prior to 1973.

          I agree with you that from the mid ’70’s we pushed current oil use into the future, but the natural curve of production would have peaked then started to fall and the linear increase past around the year 2000 has gone to the opposite extreme and is now dragging future use into the present, which is why we currently rely upon so much highly complex oil to keep our civilization running like shale oil, tar sands and deep sea extraction, that was never needed prior to around the year 2000, even though it’s not counted in conventional C+C..

31. Interesting piece

    https://oilprice.com/Energy/Crude-Oil/Exxon-Pours-Cold-Water-On-Trumps-Drill-Baby-Drill-Plans.html

32. Oil faces uphill fight as demand woes, oversupply challenge OPEC efforts: Reuters poll

    (Reuters) – Oil prices could stall in 2025 as economic weakness in China clouds the demand picture and ample global supplies outweigh support from an expected delay to a planned OPEC+ output hike, a Reuters monthly poll showed on Friday.

    The survey of 41 economists and analysts predicted that Brent crude would average $74.53 per barrel in 2025, down from a forecast of $76.61 in October.

    That is the seventh straight downward revision in the 2025 consensus for the global benchmark, which has averaged $80 per barrel so far in 2024.

    U.S. crude is projected to average $70.69 per barrel in 2025, below last month’s expectation of $72.73.

    Sentiment among oil traders “has turned very negative due to concerns about the global economy, especially about China’s economy and demand growth, and concerns about OPEC+ being able to align supply with demand,” said Stratas Advisors President John Paisie.

    Earlier this month OPEC lowered its forecast for global oil demand growth this year and next, highlighting weakness in China, India and other regions.

    Oil demand in top consumer China is expected to increase modestly due to recent stimulus measures, but structural economic challenges and the rise of new energy vehicles may restrict growth, analysts say.

    Global oil demand was seen growing by 1 million-1.5 million barrels per day in 2025, the poll showed.

    The International Energy Agency, meanwhile, expects global oil supply to exceed demand in 2025 even if cuts remain in place from OPEC+, which includes the Organization of the Petroleum Exporting Countries and allies such as Russia.

    More to this article. Click on above blue headline to get it.

    1. Hi Ron,

       Thanks for checking in, hope all is well.

       1. I am doing fine Dennis. I haven’t been posting much lately because I am engaged in another project that I hope to complete before I kick out. I actually think I have at least a couple of years left bur you never know when one gets my age. However, I feel like I am in perfect health right now.

          I actually check the list at least once a day. I check the list from the comments file instead of the actual posts. That way I scan them to see if there is anything interesting as all the posts come in in sequent at time of posting. And that also allows me to check the pending file at the same time.

    2. Ron

       Quote from your comment: “Oil demand in top consumer China is expected to increase modestly due to recent stimulus measures, but structural economic challenges and the rise of new energy vehicles may restrict growth, analysts say.”

       Below is a link on the impact of EVs in China. China knew it had to get away from oil and they started 10 years ago. Unfortunately the new Guy in the US doesn’t get and has misled everyone one else in Michigan and the rest of the US.

       “The more rapid-than-expected uptake of EVs has shifted views among oil forecasters at energy majors, banks and academics in recent months. Unlike in the US and Europe – where peaks in consumption were followed by long plateaus — the drop in demand in the world’s top crude importer is expected to be more pronounced. Brokerage CITIC Futures Co. sees Chinese gasoline consumption dropping by 4% to 5% a year through 2030.

       “The future is coming faster in China,” said Ciaran Healy, an oil analyst at the International Energy Agency in Paris. “What we’re seeing now is the medium-term expectations coming ahead of schedule, and that has implications for the shape of Chinese and global demand growth through the rest of the decade.”

       For a global oil market, which has come to rely on China as its main growth driver for most of this century, that will erode a major pillar of consumption. The country accounts for almost a fifth of worldwide oil demand, and gasoline makes up about a quarter of that. The prospect of a sharp drop from transport is also coming on top of tepid industrial consumption due to slowing economic growth.

       The growing popularity of electric trucks, as well as those that run on liquefied natural gas, is also weighing on demand for diesel. Chinese consumption of the fuel peaked in 2019 and will drop by 3% to 5% a year through 2030, UBS Securities Co. said in a note this month.”

       https://www.bnnbloomberg.ca/investing/commodities/2024/11/28/chinas-ev-boom-threatens-to-push-gasoline-demand-off-a-cliff/

       1. Thanks Ovi,

          Chart from the piece you linked

          

       2. Thanks Ovi/Ron.
          This is the early part of a big trend. Maybe even in time to avoid the harshest early affect of peaking oil.
          In the US and Canada we generally don’t see this trend (electrification of transport) as clearly as most countries, who have to import much or all of their petrol products. China also has the added incentive of insecurity…with their sea borne energy cargo imports having to transit choke points controlled by an adversary.

          “Brent Crude oil prices are expected to average $74.53 a barrel next year, the monthly Reuters poll of dozens of analysts showed on Friday, as the experts downgraded their price outlook for the seventh consecutive month.”

       3. As China’s 60 million unsold housing units weigh in on diesel demand as well. When you won’t be needing any more housing. EVER!! As many Chinese already own multiple homes and there is still 60 million unsold units. Their population also already peaked in numbers and is in contraction.

          Demand out of China is going to be lower than they are projecting.

          The Saudi’s will eventually open up whatever is left of the spigot because they need the revenue. Oil prices be dammed.
          It will not surprise me if they abandon the oil cuts over the next 9-12 months. Maybe as early as January.

          Recently there was a big deal made about the Chinese issuing a dollar bond in Saudi Arabia. It was a $2 billion bonds issuance. Well the Saudi’s owe about $300 billion in USD denominated debts. And every month they are going deeper into debt because they need $100 oil to balance their budget.

          They need dollar revenue to service that debt. Regardless of what oil prices are they need more dollars period. They will be forced to either open up the oil spigots. Or sell reserve assets in order to get dollars. Remember Saudi’s have a pegged currency and absolutely need to hold onto those dollar reserves.

          So my guess is they be opening the spigots.

       4. Thanks for the post Ovi. Yes, we have an idiot for a president. He thinks he will be able to increase US oil production by 3 million bpd just by opening up all federal land for drilling. He is in for a rude awakening.

          I agree, China has seen the light concerning EVs. Trump thinks oil will last forever.

          1. “He thinks he will be able to increase US oil production by 3 million bpd just by opening up all federal land for drilling. He is in for a rude awakening.”

             There is no such thing as a rude awakening for Trump. No matter what the result he thinks he got it right.

             There will be a rude awakening for the citizens and the countries that are currently getting USA exports.

             1. Also, there will be a rude awakening for US citizens that think, like Trump, that tariffs are a tax on exporters. They are a tax on importers, a tax that will be passed on to the consumers. Everything imported will go up. That means just about everything because even cars that are manufactured in the USA are dependent upon imports for most of the parts.

                1. Ron,

                   Also other nations will retaliate with there own tariffs to match those of the US (except perhaps on energy), this will also reduce employment in the US, hurting US workers, so we get a double whammy of higher inflation and lower GDP. This will not be a good time for the middle class, but the wealthy will continue to get wealthier.

33. September Production Drops

    Production dropped by 157 kb/d to 13,204 kb/d. August revised down by 40 kb/d to 13,361 kb/d. GOM down by 215 kb/d.

    

    1. Ovi,

       Interesting how close the STEO estimate was for Sept 2024 (13210 kb/d), they expect an increase in October of 240 kb/d, though this may be revised in the December STEO.

       1. Dennis

          Yes very close. Blew me away. Somehow the STEO has some very good sources of info and personnel. I think GOM was down due to hurricanes.

          Hurricane Helene made landfall as a Category-4 storm on the Florida Gulf Coast on September 26. The storm caused catastrophic flooding across the southern Appalachians, widespread wind damage from the Gulf Coast to the North Carolina mountains and storm surge flooding along portions of western Florida.

    2. Presumably the GOM drop was from hurricanes and will come back.

       Utah hit a new high at 197. Ohio also hit a new record and continues to leave LA in the dust.

       MT inched past KS and is even past LA this month (as is KS)! But I suspect LA had some state waters production affected by hurricanes. And they will get that back.

       —

       Natty is also slightly down. Even more slightly down if you take AK out of the equation, which you should. But that’s not Art Berman shale running out. That’s prices being miserable last fall from oversupply and weather.
       (EIA says they averaged $2.28 in SEP24.)

34. Another thing about Chinese demand. The government for months has been talking stimulus. Doing stimulus. Cutting rates.

    They been screaming for months that there is a bond bubble there and they’d be issuing lots of new bonds which would send yields soaring higher. Because all the stimulus would equal growth and inflation.

    Yeah bond yields are still at all time lows. The market doesn’t believe their stimulus has any effect whatsoever.

    You can tie falling demand for oil and oil products to the overall falling demand that the bond market just looks at these stimulus measures and laughs.

35. The Rig Report for the Week Ending Wednesday November 27

    
– US Hz oil rigs decreased by 1 to 434. They are down 25 rigs from April 19th and are up 7 relative to their recent lowest count of 427 on July 24th.
    – The Texas rig count was unchanged at 235
    
– In both New Mexico and Texas, Permian rigs unchanged at 94 and 190 respectively.
    – In New Mexico, Eddy and Lea were unchanged at 48 and 46 respectively
    – In Texas, Midland and Martin were unchanged at 26 and 27 respectively. The rig count for these two counties has not changed for the last 5 weeks.
    
– Eagle Ford held steady at 40.

    – NG Hz rigs added 1 to 85.

    

36. Frac Spread Report for the Week Ending November 29

    The frac spread count is at its lowest level since May 2021. In May 2021 the frac spread count was rising since it was coming out of the Covid shut down.

    The frac spread count decreased by 6 to 215. It is also down 61 from one year ago and down by 57 spreads since March 8.

    

37. An update on OPEC November production has been posted.

    https://peakoilbarrel.com/opec-update-november-2024/

    A new Open Thread Non-Petroleum has been posted.

    https://peakoilbarrel.com/open-thread-non-petroleum-nov-30-2024/

Comments are closed.