They’re Not Making Oil Like They Used To: Stealth Peak Oil?

The good news is that U.S. oil production has recovered to pre-pandemic levels. The bad news is that only 60% of it is really oil.

U.S. oil production exceeded the 2020 pre-pandemic peak of 20.3 mmb/d in October and November of 2022 (Figure 1). Unfortunately, less than 60% of U.S. “oil” production is really oil. The rest is non-petroleum and comes from natural gas, corn & refinery gain.

Figure 1. Less than 60% of U.S. “oil” production is really oil. The rest is non-petroleum and comes from natural gas, plants & refinery gain. Source: EIA STEO & Labyrinth Consulting Services, Inc.

Natural gas liquids (NGLs) accounted for 6 mmb/d (30%) of U.S. “oil” production in 2022 (Figure 2). The NGL portion of total output increased 5% in 1H 2020 as crude production fell. Tight oil increased from 9% to 61% of crude + condensate from 2010 to 2022 and while it is oil, it has lower energy content than conventional oil.

Figure 2. Natural gas liquids accounted for 6 mmb/d (30%) of U.S. oil production in 2022. Source: EIA STEO & Labyrinth Consulting Services, Inc.

Natural Gas Liquids (NGLs)

NGLs do not come from crude oil but are produced from natural gas. They are gases in the subsurface but are separated as liquids at surface temperatures and pressures at natural gas processing plants (Figure 3). NGLs include ethane, propane, butane and pentane.

Ethane constitutes the largest share (~55%) of NGL production. It is used almost exclusively to produce ethylene, which is then turned into plastic bags, anti-freeze and detergent. Propane accounts for 31% of NGLs but only 3% of it is used as a transport fuel; its main use is home and water heating. Butane makes up another 16% of NGLs and its main uses are fuel for cigarette lighters and portable stoves, and as a propellant in aerosols.


Figure 3. What are natural gas liquids? Source: EIA, AFP tech and Labyrinth Consulting Services, Inc.

Refinery Gain

Refinery gain accounted for more than 1 mmb/d of U.S. oil production in 2022.

Gain occurs during refining because petroleum products coming out of a refinery are less dense than the crude oil going in. The volume of refined products is therefore greater than the volume of crude oil intakes. That volume difference is called refinery gain.

For example, the average density of crude oil is 846 kg/m3 (Figure 4).  Gasoline accounts for about 45% of each barrel of refined U.S. crude oil and its density is 744 kg/m3. That means that approximately 1.14 barrels of gasoline are produced from each barrel of oil. That is refinery gain.


Figure 4. What is refinery gain? Source: EIA & Labyrinth Consulting Services, Inc.

Fuel Ethanol

Fuel ethanol accounted for more than 1 mmb/d of  U.S. oil production in 2022. It is denatured alcohol made by fermenting the sugar in the starches of grains like corn (Figure 5). It is blended with gasoline to extend the use of that fuel.

Figure 5. What is fuel ethanol? Source: Resource Media and Labyrinth Consulting Services, Inc.

Tight Oil

Tight oil accounted for more than 7 mmb/d of U.S. oil production in 2022. Less than 5 mmb/d of conventional oil was produced in 2022. Unlike, natural gas liquids, refinery gain and fuel ethanol, tight oil is petroleum.

It has, however, a lower density and corresponding lower energy content than conventional oil. Permian tight oil, for example, has about 93% of the energy content (5.5 mmBtu/barrel) as the standard conventional oil required by U.S. refineries (5.9 mmBtu/barrel) (Figure 6).

Figure 6. Permian tight oil has 7% lower energy content than the average oil intake for U.S. refineries. Bakken & Eagle Ford tight oil have 4% lower energy content. Source: Enverus, EIA & Labyrinth Consulting Services, Inc.

Some may argue that 7% is not that much when it comes to a fuel as potent as oil but it is the difference between an “A” and a “B” in school. Put differently, imagine if world oil crude & condensate supply fell by 7%. That’s half of what Saudi Arabia produces. It’s five times more than Libya produces yet whenever its production falls because of civil conflict, world oil price is profoundly affected.

More importantly, tight oil does not contain the middle distillate compounds necessary for diesel production. Figure 6 shows the density (API and specific gravity) of the key conventional grades of oil, and for the Bakken, Permian and Eagle Ford tight oils. Tight oil is fine for making kerosene, jet fuel and gasoline. It cannot, however. be used for producing diesel without blending it with heavier oils, and diesel is the main cash product and workhorse of the modern global economy.

The U.S. can never be oil-independent because it will always need to import heavier oil to make diesel.

Figure 7. Many U.S. oils lack the heavy compounds needed to make diesel but are good for making kerosene, jet and gasoline. Source: Global Security, Enverus, EIA & Labyrinth Consulting Services, Inc.


Stealth Peak Oil

If any of this sounds strangely familiar, it is because it was anticipated 20 years ago by Peak Oil.

Peak Oil was a flawed concept because of its preoccupation with predicting a date for world oil production to peak. Many of its key precepts, however, were sound namely, that price would rise as oil quality decreased and decline rates increased. I have shown the pronounced decrease in oil quality at least in the United States.

Figure 8 shows that the total production base decline rate for the U.S. is about 37% per year. The data is for all wells drilled from 2000 through 2022 in the largest American producing regions that account for 88% of total crude oil and condensate output.

Figure 8. The red queen effect. The U.S. oil production base declines at 37% per year. Source: Enverus & Labyrinth Consulting Services, Inc. This is the aggregate decline for all wells, not just the rate for any given vintaged year shown in the figure.

The Peak Oil assumption that oil prices would increase chiefly because of the depletion of known reserves was simplistic. First of all, reserves are a moving target. The values published by BP or EIA are resources, not reserves. Reserves are a volume at a price, not a fixed quantity. When the price of oil increases, proved reserves increase, and vice versa.

Second, new reserves are always found despite the direst expectations to the contrary. The world has been 10 years away from producing all of its reserves since I began my career in the oil business 45 years ago.

More importantly, price formation is more complex than an inverse correlation with reserves especially in an increasingly financialized world. Markets care more about supply more than about reserves. Supply is current production plus inventories and spare capacity (the amount of production that can be converted into supply in about 90 days). The tens of billions of barrels of resources or potential reserves in plays unlikely to become supply in the near-term (like Orinoco tar sands) do not impress oil markets.

Supply urgency is what moves oil prices higher and markets have felt urgency for most of the last two decades. As Colin Campbell and Jean Laherrere insightfully observed 25 years ago,

“The world is not running out of oil—at least not yet. What our society does face, and soon, is the end of the abundant and cheap oil on which all industrial nations depend.”

WTI prices averaged $90 per barrel in 2022 dollars since 2003 (Figure 9). That is more than twice the average price during the preceding two decades. That $90 average includes the low-price periods from 2014 through 2017 (price collapse from shale output), and 2020 through mid-2021 (Covid) in which prices fell to their lowest levels in modern history.

Figure 9. WTI prices averaged $90 per barrel for two decades despite lower prices from 2015 through 2017 and during the Covid pandemic. Source: EIA, U.S. Dept. of Labor Statistics & Labyrinth Consulting Services, Inc.

Total world liquids production has recovered to 99% of 2018 average level but crude oil plus condensate has not and remains more than 4 mmb/d below late 2018 levels (Figure 10). The world is not running out of oil but investors and credit markets are unwilling to underwrite the drilling necessary to increase oil output.

Figure 10. Total world liquids production has recovered to 99% of 2018 average level but crude oil plus condensate remains more than 4 mmb/d below late 2018 levels. Source: EIA, Cansim, Enverus & Labyrinth Consulting Services, Inc

As David Fickling recently stated,

“Ultimately, it will be central banks that will read crude its last rites.”

The likelihood of a secular return to lower oil prices is as unlikely as is a technological breakthrough that results in enough new oil supply to modify pricing. That is because the “lower-for-longer” pricing after 2014 because of shale plays was a relatively insignificant anomaly in the larger price scheme in Figure 9.

The consequences of higher energy prices following Russia’s invasion of Ukraine have made the world more energy-aware. For some, they have increased the resolve to make a transition away from oil and other fossil fuels. At the same time, they have shown just how dependent we are on fossil energy, and will continue to be for decades to come.

Peak oil was a model that made reasonable if simplistic assumptions based on M. King Hubbert’s method of forecasting future oil production trends from proved reserves.

What I am describing is not a model. Oil quality has decreased, production decline rates have increased, and long-term secular prices are higher. Those are facts, not theory.

The sooner we stop expecting a miracle of technology or a quick transition to renewable energy, the better we will be able to cope with a more difficult energy future.


  • The underlying message here is that oil and fossil fuels will one day cease to be there.Will be depleted.Countries must start urgently to develop other sustainable sources of energy.Oil will not be there in say 200 years to come?

    • Art Berman


      Energy substitution is a lovely idea but first, you have to have a substitute that in fact, substitutes for the energy density and content, price and availability of fossil fuels.

      Renewable energy is not that substitute.



  • juan arias

    Dear Mr Berman, do you think the world is about to take rationing mesures for oil, electricity, even food, due to energy shortage. We are 8.000 milion people, China and India are consuming a lot of energy, specially coal and many of natural resources are declining.


    Juan. Barcelona (Spain)

  • Fidra

    Very interesting, Mr. Berman. I’m curious, if figure 10 is expressed as energy content instead of volume (BTUs instead of barrels), what does that look like in terms of peak and possible decline? Thank you for helping us all see the truth of our situation.

  • […] In der jüngsten Folge seiner “Great Simplification” unterhält sich Nate Hagens mit dem Ölgeologen Art Berman, dem man wahrlich nicht vorwerfen kann, er wisse nicht, wovon er spricht.  Die beiden pontifizieren lange über die US-Ölproduktion, die sich “nach alter Definition” bei weitem (noch) nicht auf das Vor-Covid-Niveau erholt hat – was wohl nur für altgediente Peakoilistas zugänglich ist. Aber ab ca. Minute 20 erläutert Berman, was die “All Liquids”, mit denen jetzt überall herum gefuchtelt wird, alles beinhalten (außer “C & C”): Raffineriegewinne, NGLs, und Corn ethanol zum Beispiel. Für die, die besser Englisch lesen als hören, erklärt Berman das in seinem Blog.  […]

  • juan arias

    Dear Mr Berman, in this situation, do you think all governaments around of the world will have to take rationing mesures?. Thanks you

  • […] mit einem selbst gemachten Chart, das in etwa Bermans “Figure 2″ – siehe hier –  […]

  • […] of whether a peak was reached in November 2018, might yet prove to have been premature.  As Art Berman has shown, following lockdown, at 100 million barrels per day, world oil production is almost back […]

  • John Day

    Art, great presentation. Have you seen the publication by LaHererre et al. (LaHerre, Hall, and Bentley). I will send it to you just in case. It shows that total reserves of conventional oil are only about half of what BP and IEA project.

    • Art Berman


      Yes, I’ve seen Jean’s & Charlie’s latest and forwarded my criticisms to them.

      I don’t think that Hubbert linearization is a credible method because the it relies on the ratio of production divided by cumulative production. As the denominator grows, the non-linear relationship between the two variables becomes linear permitting the forecast but it’s an artifact IMO.

      Hubbert Linearization is a reasonable first-look method but hardly a sound basis for URR forecasting.



  • Fairly complicated and very important development very well explained.

  • Rfnrybizweenie

    1. I haven’t seen that much total liquids stuff. Read more about growth of C&C, and acknowledgment that it is below pre Covid. But if something bothered you, uh OK.

    2. I heartily agree that propane is not oil. (And avoid “total liquids”.)

    3. I don’t agree with some of the comments about tight oil quality though. For one thing, the energy content is mostly of import if you burn it. And we don’t. And for that matter, I am always against “BOE”, which is energy based, but over prices the value of nat gas. So, if you are going to value oils by energy content, do you also value gas that way? Or do you (rightfully) push back on BOE silliness.

    4. Your comments about diesel are more on track. It’s the products that matter. However, you also have to look at sulfur. shale oil tends to be sweet. Heavy oil tends to be sour. If you just look at market value, you can see that WTI-Houston (light sweet) and WCS-Houston (heavy sour) trade with WTI worth more. Yes, if oils get too light, there’s a discount. But it’s a parabola…and 40-45 is optimal in the market. Maybe less diesel, but also low sulfur, and less resids (less work to crack). I mean if just being heavy was optimal, tar sands would be worth the most!

    5. Also 48 is not an average value for Permian C&C. It’s low 40s. You did the right thing, pushing back on silly news articles about total liquids, but have the wrong take on Permian API (almost all is piped and almost all is piped in slugs of 40-45 API grade contracts…it’s actually hard to find a lighter cargo, even though there is some incentive to separate, not mix.) This is also the case if you trade at Rotterdam. I haven’t seen a single cargo of 45+ West Texas. (Eagle Ford you do see commonly at 47.)

    Dated but relevant:

    6. Also, EIA reports API in a 5 degree histogram for lower 48. (AK is omitted but its about 30, so this is actually conservative.) (click and download the eighth Excel down, on the right, NOT “table 5”)

    If you pull that, you’ll see that current L48 is 73% under 45 API and 91% under 50 API. For comparison, in NOV19 (the peak of US C&C), L48 was 74% under 45API and 90% under 50 API. So, it’s really not a dramatically different mix now, versus then.

  • Rfnrybizweenie

    You should also take a look at this article (chart 3). The idea that LTO is less valuable than medium sour (like Mars) is just wrong. Despite all the handclutching about diesel, LLS sells at a premium to Mars. Yes, the premium has shrunk, but it has not reversed.

    Note: the article is from 2019, but the story remains the same today (and LTO is actually not yet at late 19 levels of production yet). And you can look up the prices for Mars and WTI-Houston on CME for free (or trading sites like Platts or Argus if you have a paid account). Despite having “more energy”, Mars gets a small discount versus WTI. That’s because it has more resids and more sulfur.

Leave a Reply

This website uses cookies and asks your personal data to enhance your browsing experience.