Peak Shale Production?

Dear Subscribers,

I would like to share with you an academic paper written by my colleague Prof. Douglas B. Reynolds. Conclusions are posted below. Below is the link to the whole paper. Comments are welcome. I will forward them to Prof. Reynolds. I will post his replies below this post.

Source: Halliburton.

U.S. shale oil production and trend estimation: Forecasting a Hubbert model

CONCLUSION AND POLICY IMPLICATIONS

Given the challenges of estimating a Hubbert Curve trend before the peak occurs, even assuming appropriate characteristics are in place, and noting that a straight-line estimate for the U.S. shale oil trend is almost as accurate as a simple Quadratic Hubbert Curve with data ending in 2019, there is no discernible evidence that U.S. shale oil has or will peak soon. Therefore, the simple trend tests are inconclusive. But since the pre-2016 Hubbert Curve estimate is very close to the full data set Hubbert Curve, then a URR of Lower 48 shale oil projects to 40 to 50 Billion barrels with a peak date in a business-as-usual economy with normal demand at about 8 to 9 mbd in 2022. Even with a dummy variable to mimic the 2015 change in price, there is not much change in URR or peak production or peak date.

During all of 2019 the drilling rig count declined even as oil prices modestly improved, and where the trend of drilling follows a 25-week lag in price. The 2019 decline in rigs mimics the 1956 to 1958 decline in conventional oil rigs. This is the exact point at which a Hubbert trend for the U.S. conventional oil was estimated by Hubbert, giving substantive evidence of a near term peak in production. That is, a roughly 50% increase in the cumulative production amount from the drilling rig signal point denotes where peak oil will be. The plateau of oil rig count numbers starting in June 1 of 2018 suggests that a Hubbert peak for U.S. Lower 48 shale oil is imminent.

Finally, the 2021 and early 2022 data show a strong lack of an information effect, which would normally not happen if the production peak were far off, and where substantial cumulations of production were already in hand to enhance the information effect. The price and Hughes rig count rise would easily increase production if a dominant information effect were occurring. Since production is flat or declining with the rise in price and rigs, then there is a very strong depletion effect dominating the information effect, and that peak production is past or quite near.

One fundamental difference between the U.S. Lower 48 conventional oil Hubbert Curve and the U.S. Shale oil Hubbert curve is the smaller but much quicker shale oil trend. Where as the conventional oil middle two thirds of the pre-1970 production increase was at 4% per year, the shale oil middle two thirds of the pre 2019 increase was at 25% per year. One might believe that the speed of the rise of the shale oil curve is simply due to technology being so much better. More pertinently, shale oil is the source rock for conventional oil, and once conventional oil reserves are pinpointed, then their source rock is discernible. Therefore, there is a faster information and depletion effect for shale oil, which is the real danger in the scarcity and growth analysis. See Figure 5 for a comparison.

Figure 5 shows that the 1970s U.S. Lower 48 oil production decline occurred at a rate of 250 thousand barrels per day reduction in output per year, enough to cause stagflation worldwide, where BP (2021) statistics show OPEC kept its output dependable during nearly all of the 1970s. Thus, the U.S. caused the oil price induced stagflation of the 1970s not OPEC. In contrast, the U.S. Lower 48 shale oil decline for the mid to late 2020s projects to decline at 570 thousand barrels per day in output per year, a daunting downsizing adaptation. Since Reynolds and Umekwe (2019) show non-U.S. shale oil will be difficult to expand, then the world's energy transition away from petroleum will be more challenging than the 1970s, or even the Soviet Union's and Eastern Europe's fall in the 1990s due to its oil production decline that reduced GDP.

According to Wiorkowski, “The generalized Richards (1959) function and the Weibull curve yield drastically different estimates of ultimate recovery (445.6 bbl and 235 bbl, respectively), yet both curves seem to fit the data well as measured by the magnitude of the residual variance.” Therefore, even as of 1976, Wiorkowski found a divergent mathematical model to show two and a half times as much URR as Hubbert, meaning it can be tough to make predictions even after the peak in oil. However, the Wiorkowski construct was not based on economic theory and even now Lower 48 conventional oil looks to hit 190 billion barrels and not Wiorkowski's 445.6 billion mark. Given that whale oil and Lower 48 conventional oil both follow closely a quadratic Hubbert Curve to an above 90% R squared level, then that gives the curve some evidence of its theoretical construct being true when there are free institutions. The U.S. Lower 48 shale oil is likely to follow a similar pattern and may have peaked already since the U.S. has free markets and since natural gas is a complement in production value for the shale-resource and since the shale oil industry may have consolidated after the COVID market lull.

As Smil (1991) said, “since (energy) consumption has been based largely on the supply of high-power density (i.e., concentrated energy) fossil fuels, the future transition to arrangements dominated by low power density biospheric fluxes (solar, wind and hydropower) will require major socio-economic transformation.” Nuclear, coal and alternative hydrocarbons may help, but world economies are still not ready for this transition shown in Orlov (2008) and explained in Reynolds and Kolodziej (2008) and Reynolds (2016). The business-as-usual market progression has changed due to the pandemic, but maybe this has helped the world adapt to what will become much higher petroleum prices.

As Wiorkowski admitted, “… if any of the estimates are correct, such policy decisions must be made in the near future.” Now may be that time. Simmons (2005) stresses, “It is urgent that we begin planning now for the non-conventional energy future.” And Catton wonders, “nature must in the not-too-distant future institute bankruptcy proceedings against industrial civilization … (if it does not get) the opportunity to begin the drawdown.” As Galbraith (1958) suggests much of the demand that exists for goods and services is artificial, and Bertrand Russell urges cultivating leisure activities that are not resource intensive. Or as Hall et al. (1992) say, “It appears that it would be wise to at least plan for the contingency that omnipotent technology will not be the savior of infinite economic growth.” Nevertheless, an energy and economic crisis is imminent and the world needs to prepare for it and recognize the importance of OPEC in conserving this resource for future generations along the lines of Hotelling (1931) or Solow (1974).