By Scott Lapierre, Petrophysicist/Reservoir Characterization Director/founder Shale Specialists LLCEdited for Oil and Gas 360


Introduction

On the Nature and Character of the Widespread Oil Production Shortfalls Reported by the Wall Street Journal- oilandgas360On January 2, 2019, the Wall Street Journal (WSJ) published Fracking’s Secret Problem—Oil Wells Aren’t Producing as Much as Forecast[1] ; the first in a series of articles[2,3,4] critical of the Great American Oil Shale Revolution. The authors employed a third-party energy consulting firm to re-forecast “16,000 wells operated by 29 of the biggest producers in oil basins in Texas and North Dakota”, and then compared the updated forecasts to original corporate projections originally published to justify drilling campaigns and lure in investors. The WSJ authors claim that “two-thirds of projections made by the fracking companies between 2014 and 2017 in America’s four hottest drilling regions appear to have been overly optimistic”[1]. The authors further singled out Pioneer Natural Resources (PXD) and Parsley Energy (PE) as the WSJ’s newly generated projections pointed to 25% less recovery than what “their owners projected to investors”.

On December 31, 2019 The Wall Street Journal published a follow-up piece entitled As Shale Wells Age, Gap Between Forecasts and Performance Grows which reinforces the notion that US shale oil production forecasts are too optimistic. Stringent new demands from investors to generate positive free cashflow (+FCF) and return on capital employed (ROCE) require that today’s capital be offset with revenues from tomorrow’s production. Therefore, overoptimistic forecasts place capital at additional risk only offset by increasing commodity prices.

Since the WSJ articles were published, the industry has undergone dramatic change as evidenced by the announcements of reorganizations and layoffs in an effort to boost profits by reducing costs[5,8,9,10]. Also worth noting, for the first time in the 10-year history of the shale oil revolution, capital budgets are being reduced not as a direct response to structural changes in commodity prices but rather in response to investors who no longer seem willing to tolerate value destruction from unprofitable development Together, the widespread layoffs[8,10], executive retirements[5] and capital budget reductions[9] suggest there is a yet-to-be-identified, or yet-to-be-addressed, fundamental obstacle hindering the commercial production of oil from frac’d shale reservoirs.

The present article explores public data from key operators in major basins for evidence of the shortfalls reported by the WSJ to independently confirm their existence and determine their materiality and uniformity. The present article further attempts to characterize the common nature of the shortfalls and provide a context for comparison with original projections. While covering such a topic necessitates delving into technical aspects of reservoir engineering, an attempt is made to keep the discussion friendly to the layperson by keying in on just two types of production data plots. A typical rate-versus-time plot is described and used to introduce the rate-versus-cumulative- oil-produced plot format. The introduced plot format will provide for very quick, visual comprehension of the nature, materiality and universality of the reported deviations. Becoming comfortable with the new plot format will greatly assist the reader with visually identifying the problem and quickly approximating any impending reductions to ultimate oil recoveries.

Typical Production Plot Formats

Production data is commonly displayed in Investor Relation slide decks via one of the two formats shown in Figure 1. Figure 1-A shows a typical rate-versus-time plot where [barrels of oil equivalent per day (BOEPD)] are plotted against [time]. Figure 1-B shows a plot where [cumulative oil equivalent produced (BOE)] is plotted against [time on production]. Corporate projections for ultimate recovery by the end of a well’s life are overlain as ‘type curves’ and provide an important scale and context for translating early-time production data into a magnitude of estimated ultimate recovery (EUR). Without the assistance of an overlain type curve these plot formats do not innately convey a sense of ultimate recoveries. Production is typically presented normalized for lateral length due to the reality of developing leases supporting various wellbore length capacities. Normalization for lateral length is not a subject of controversy among industry experts.

On the Nature and Character of the Widespread Oil Production Shortfalls - Figure 1 -oilandgas360

Figure 1: Common formats for displaying production relative to forecasts. (A) is an example of barrels of oil equivalent daily rate (BOEPD) plotted against time (days). (B) is an example of cumulative oil production in barrels of oil equivalent (BOE) plotted against days on production. Presenting only the first and most prolific months of production is assisted by multiple type curves being overlain to represent various projections of ultimate recovery. Typical well life, determined by assuming a terminal, constant- rate exponential decline of 3%-14% coupled with a 1-2 barrels per day minimum economic rate suggest 20-50 year well life. Source: Pioneer Natural Resources and Parsley Energy IR presentations.

Typically, engineers plot early-time oil and gas production rates versus time on a semi-log plot (Figure 2). Quite commonly, a forecasting method of decline curve analysis (DCA) first introduced by the geologist J. Arps in 1944[11] is used. Accordingly, a hyperbolic math function is ‘fit’ through ear ly-time data by selecting equation parameters that optimize an empirical regression. Figure 2 illustrates the rate-versus-time plot and forecast generated by fitting a hyperbolic decline function through the early-time data.

On the Nature and Character of the Widespread Oil Production Shortfalls - Figure 2 - oilandgas360

Figure 2: (Illustrative) Typical plot of oil production plotted against time illustrating a forecast generated with a hyperbolic decline function whose parameters have been selected to optimize the function’s fit through early-time data

 

The hyperbolic decline function has specific characteristics that make it useful as a curve to fit through production data. Namely, the hyperbolic math function can reproduce natural decline profiles characterized by initial annual decline rates that lessen over time. For example, in steeply declining shale wells, the first year of production may end with an exit rate 60% to 95% of the initial oil rate. Fortunately for longevity of production, each additional year enjoys a less severe annual decline rate. Nonetheless, ultimately the life-prolonging lessening of decline severity terminates and a final, minimum, terminal constant-rate decline (D min) sets in. The appearance of a brutally constant, minimum, decline rate (Dmin) marks the beginning of the end of the economic life of a well. Figure 3 illustrates the consecutive lessening of annual decline rate characterized by the hyperbolic decline function. A minimum decline rate (Dmin) is invoked to reflect the transition to constant-rate (exponential) decline to accommodate the natural, internal reservoir mechanisms that develop at the end of a well/reservoir’s life. D min values of 3% to 14% have been in widespread use since the beginning of the shale oil revolution based largely on observations of legacy vertical production in places like the Midland Basin[12].

On the Nature and Character of the Widespread Oil Production Shortfalls - Figure 3- oilandgas360

Figure 3: (Illustrative) Hyperbolic decline function fit through early-time production data as per Arps DCA illustrating year-on-
year lessening of annual decline rate with an assumed constant-rate, minimum decline equal to 7% being reached in the distant years

 

The nature of the production shortfalls identified by the Wall Street Journal appear on a rate-versus-time plot as a gentle shallowing of production over time relative to an original forecast. Figure 4 illustrates the subtle appearance of a typical shortfall as seen on a rate-versus-time plot.

On the Nature and Character of the Widespread Oil Production Shortfalls - Figure 4 - oilandgas360

Figure 4: (Illustrative) Typical nature of the production shortfalls identified by the Wall Street Journal as viewed on a rate-versus- time plot

 

Unfortunately, since the x-axis represents time, projecting the observed production data and corporate forecast forward does not immediately convey the magnitude of the shortfall nor its effects on ultimate recovery. A plot format in which the x-axis represents the cumulative oil production more vividly conveys both the magnitude and the nature of the reported shortfalls. With a properly scaled x-axis on a rate-versus-cumulative oil production plot, the shallowing of production that was ambiguous on the rate-versus-time plot, will appear as a pronounced, marked, downward curvature from a straight line angled toward the bottom right of the plot (Figure 5).

In the rate-versus-cumulative oil produced plot space, a straight line angled toward the bottom right of the graph approximates the period of gradual lessening of annual decline characteristic of the hyperbolic decline function (with ab exp of ~1). Furthermore, a straight-line angled toward the bottom right of the graph will reliably approximate the forecast methodology widely used to generate published corporate forecasts. Consequently, rate-versus-cumulative oil produced plots provide a quick reference to ultimate recovery as continuation of the initial near-linear portion of the trend “points” to an approximation of the EUR for a traditional forecast that assumes a 3% to 14% D min and a b exp equal to 1. Downward curvature from the descending near-linear trend represents that transition to a minimum, constant-rate terminal, exponential decline (or a reduction in b exp to below a value of 1). Such downward curvature fundamentally marks the beginning of the end for commercial well production. Once a downward curvature is established, a projection of that downward curvature reliably “points” to a new approximation of ultimate recovery for that well. Any downward curvature observed in the semi-log rate-cumulative oil plot represents a marked deviation from the type curve philosophy used to make many of the corporate projections supplied to investors over the last decade. As will be shown in the next section, a significant portion of the reported production shortfalls can be attributed to premature transitions to constant-rate terminal decline. The “ Dmin Problem”, to which it shall henceforth be referred, stems from widespread use of incorrect assumptions for values of D min (3%-14%).

On the Nature and Character of the Widespread Oil Production Shortfalls Reported by the Wall Street Journal - Figure 5 - oilandgas360

Figure 5: (Illustrative) Oil rate plotted against cumulative oil produced more vividly reveals the magnitude and nature of a production shortfalls as a premature transition to constant-rate terminal exponential decline

While the rate-cumulative oil production plot dramatically highlights visibility of the Dmin problem; adding the ratio of gas produced per barrel of oil (GOR) reveals a reliable and potentially important precursor to the Dmin problem  runaway GOR.

Figure 6-A and B illustrate how an exponential rise in GOR trends precede the Dmin Problem. Figure 6-A illustrates the seemingly benign, early stages of rising GOR that has been described by several CEO’s to provide only positive uplift through increased NGL revenues[13,17]. Figure 6-B illustrates the typical evolution of a rising GOR trend wherein previous arguments of a benign nature appear to fall apart.

On the Nature and Character of the Widespread Oil Production Shortfalls - Figure 6 - oilandgas360

Figure 6: Illustration of exponential rise in GOR. Initially, rising GOR appears benign and presents some marginal uplift. However, as GOR trends escalate into an exponential increase they become decidedly correlated with the Dmin Problem (B)

The “ GOR Problem”

Since the beginning of the Great American Oil Shale Revolution, corporate forecasts provided by shale companies have nearly unanimously assumed GOR (and % oil) would be flat forever. It is worth noting that Percent Oil and GOR are often used interchangeably wherein Percent Oil is roughly the opposite of GOR. Only after the Permian Stumble of Q2 of 2017, when several key producers gassed-up unexpectedly with a few missing production guidance, did the public receive the first acknowledgements by CEO’s of rising GOR. At the time, CEO’s claimed the un -forecast rises in GOR were an anticipated phenomenon[13] despite corporate forecasts incorporating such being exceptionally rare. Rising GOR’s had furthermore been dismissed as benign, having no negative impact on oil production. As will be demonstrated in the following section, widespread rising GOR trends are demonstrably quite universally correlated with premature transitions to constant rate, terminal declines across the majority of US shale oil production.

On the Nature and Character of the Widespread Oil Production Shortfalls - Figure 7 - oilandgas360

Figure 7: PXD earnings release slides from 2014 and 2018 revealing assumptions of constant GOR and % Oil over the entire life of production

 

Data Review of Wells Drilled to-Date in the American Oil Shale Revolution

Leveraging an oil rate (& GOR)-versus-cumulative oil produced plot enables exceedingly quick assessment of production data in the context of implied ultimate recoveries. Shale Profile Data Analytics services were used to generate a series of plots that allow quick interrogation of oil production and GOR profiles for key basins/plays within the United States (Midland Basin, Bakken, DJ-Niobrara, and Eagle Ford). Play-wide data is presented along with three individual well examples from each basin/play. Near-linear trends, approximating common DCA practices and parameters are overlain to represent what has traditionally been claimed as typical EUR’s for typical wells in these plays. Taking advantage of the implications of downward curvature on a rate-versus-cumulative oil produced plot enables comparison of the implied production shortfalls referenced by the WSJ relative with traditional forecasting techniques. Three examples are provided for each major basin/play. The reader is expected to notice that dramatic – yet regular – deviations from traditional projections are widely manifest (near-linear oil rate trend transitioning to downward curvature that indicates substantially less ultimate oil recovery). Furthermore, the reader is expected to appreciate the implied magnitude of the widespread shortfalls in production relative to published projections provided by industry over the last decade. The data shown verifies the observations made by the Wall Street Journal and further suggests that oil production from shales has been overestimated by approximately 50% compared to the 10% to 63% cited by the WSJ for the Midland Basin and the Eagle Ford.

On the Nature and Character of the Widespread Oil Production Shortfalls Figure 8 - oilangas360

Figure 8: Shale Profile analytics data of 8,473 horizontal wells from the Midland Basin demonstrating consistent occurrence of 10- to 20-fold increases in GOR and abundant downward curvature in oil rates revealing premature transition to constant-rate minimum terminal decline. Note: GOR in upper right plot scaled from 0 scf/BO to 20,000 scf/BO is equivalent to 100% oil to 23% oil

 

On the Nature and Character of the Widespread Oil Production Shortfalls - Figure 9 - oilandgas360

Figure 9: Shale Profile analytics data for typical horizontal well from the Midland Basin (Example #1) plotted against all Midland Basin Hz wells illustrating prevalence of the Dmin Problem and its correlation with the GOR Problem. Note: GOR in upper right plot scaled from 0 scf/BO to 20,000 scf/BO is equivalent to 100% oil to 23% oil

 

On the Nature and Character of the Widespread Oil Production Shortfalls - Figure 10 - oilandgas360

Figure 10: Shale Profile analytics data for typical horizontal well from the Midland Basin (Example #2) plotted against all Midland Basin Hz wells Illustrating prevalence of the Dmin Problem and its correlation with the GOR Problem. Note: GOR in upper right plot scaled from 0 scf/BO to 10,000 scf/BO is equivalent to 100% oil to 37% oil

 

On the Nature and Character of the Widespread Oil Production Shortfalls - Figure 11 - oilandgas360

Figure 11: Shale Profile analytics data for typical horizontal well from the Midland Basin (Example #3) plotted against all Midland Basin Hz wells illustrating prevalence of the Dmin Problem and its correlation with the GOR Problem. Note: GOR in upper right plot scaled from 0 scf/BO to 10,000 scf/BO is equivalent to 100% oil to 37% oil

 

On the Nature and Character of the Widespread Oil Production Shortfalls - Figure 12 - oilandgas360F

Figure 12: Shale Profile analytics data of 15,914 horizontal wells from the Williston Basin demonstrating consistent occurrence of 10-fold increases in GOR and abundant downward curvature revealing premature transition to constant-rate minimum terminal decline

 

On the Nature and Character of the Widespread Oil Production Shortfalls -Figure 13

Figure 13: Shale Profile analytics data for typical horizontal well from the Williston Basin (Example #1) plotted against all Bakken Hz wells illustrating prevalence of the Dmin Problem and its correlation with the GOR Problem. Note: GOR in upper right plot scaled from 0 scf/BO to 10,000 scf/BO is equivalent to 100% oil to 37% oil

 

On the Nature and Character of the Widespread Oil Production Shortfalls -Figure 14 -oilandgas360

Figure 14: Shale Profile analytics data for typical horizontal well from the Williston Basin (Example #2) plotted against all Bakken Hz wells illustrating prevalence of the Dmin Problem and its correlation with the GOR Problem. Note: GOR in upper right plot scaled from 0 scf/BO to 10,000 scf/BO is equivalent to 100% oil to 37% oil

 

On the Nature and Character of the Widespread Oil Production Shortfalls -Figure 15 -oilandgas360

Figure 15: Shale Profile analytics data for typical horizontal well from the Williston Basin (Example #3) plotted against all Bakken Hz wells illustrating prevalence of the Dmin Problem and its correlation with the GOR Problem. Note: GOR in upper right plot scaled from 0 scf/BO to 10,000 scf/BO is equivalent to 100% oil to 37% oil

 

On the Nature and Character of the Widespread Oil Production Shortfalls -Figure 16 -oilandgas360

Figure 16: Shale Profile analytics data of 7,472 horizontal wells from the DJ Basin demonstrating consistent occurrence of 10- to 20-fold increases in GOR and abundant downward curvature revealing premature transition to constant-rate minimum terminal decline. Note: GOR in upper right plot scaled from 0 scf/BO to 20,000 scf/BO is equivalent to 100% oil to 23% oil

 

On the Nature and Character of the Widespread Oil Production Shortfalls -Figure 17 -oilandgas360

Figure 17: Shale Profile analytics data for typical horizontal well from the DJ Basin (Example #1) plotted against all DJ Basin Niobrara Hz wells illustrating prevalence of the Dmin Problem and its correlation with the GOR Problem. Note: GOR in upper right plot scaled from 0 scf/BO to 20,000 scf/BO is equivalent to 100% oil to 23% oil

 

On the Nature and Character of the Widespread Oil Production Shortfalls -Figure 18 -oilandgas360

Figure 18: Shale Profile analytics data for typical horizontal well from the DJ Basin (Example #2 plotted against all DJ Basin Niobrara Hz wells illustrating prevalence of the Dmin Problem and its correlation with the GOR Problem. Note: GOR in upper right plot scaled from 0 scf/BO to 20,000 scf/BO is equivalent to 100% oil to 23% oil

 

On the Nature and Character of the Widespread Oil Production Shortfalls - Figure 19 - oilandgas360

Figure 19: Shale Profile analytics data for typical horizontal well from the DJ Basin (Example #3) plotted against all DJ Basin Niobrara Hz wells illustrating prevalence of the Dmin Problem and its correlation with the GOR Problem. Note: GOR in upper right plot scaled from 0 scf/BO to 20,000 scf/BO is equivalent to 100% oil to 23% oil

 

On the Nature and Character of the Widespread Oil Production Shortfalls - Figure 20 - oilandgas360

Figure 20: Shale Profile analytics data of 22,356 horizontal wells from the Eagle Ford play demonstrating consistent occurrence of 10-fold increases in GOR and abundant downward curvature revealing premature transition to constant-rate minimum terminal decline

 

On the Nature and Character of the Widespread Oil Production Shortfalls - Figure 21 - oilandgas360

Figure 21: Shale Profile analytics data for typical horizontal well from the Eagle Ford play (Example #1) plotted against all Eagle Ford Hz wells illustrating prevalence of the Dmin Problem and its correlation with the GOR Problem. Note: GOR in upper right plot scaled from 0 scf/BO to 10,000 scf/BO is equivalent to 100% oil to 37% oil

 

On the Nature and Character of the Widespread Oil Production Shortfalls - Figure 22 - oilandgas360

Figure 22: Shale Profile analytics data for typical horizontal well from the Eagle Ford play (Example #2) plotted against all Eagle Ford Hz wells illustrating prevalence of the Dmin Problem and its correlation with the GOR Problem. Note: GOR in upper right plot scaled from 0 scf/BO to 10,000 scf/BO is equivalent to 100% oil to 37% oil

 

On the Nature and Character of the Widespread Oil Production Shortfalls - Figure 23 - oilandgas360

Figure 23: Shale Profile analytics data for typical horizontal well from the Eagle Ford play (Example #3) plotted against all Eagle Ford Hz wells illustrating prevalence of the Dmin Problem and its correlation with the GOR Problem. Note: GOR in upper right plot scaled from 0 scf/BO to 10,000 scf/BO is equivalent to 100% oil to 37% oil

 

Universal and Widespread Trends

Further exploitation of the informationally rich Rate versus Cumulative Oil Produced plots available in Shale Profile’s data analytics services allows one to ascertain the production status of the vast majority of American shale oil production with just a few figures. According to the EIA at the time of this writing, US shale production is responsible for 60% of total US oil production.

On the Nature and Character of the Widespread Oil Production Shortfalls - Figure 24 - oilandgas360

Figure 24: EIA.gov data showing increasing contribution of tight oil to US total production

Additionally, according to ShaleProfile’s analytics data, US shale production is dominated by four basins/plays: Permian Basin, Eagleford, Bakken, and DJ Basin.

 

On the Nature and Character of the Widespread Oil Production Shortfalls - Figure 25 - oilandgas360

Figure 25: ShaleProfile data illustrating the four major basins comprising 72% of US shale oil production (Delaware Basin excluded)

 

Combining Figures 8 through 23 allows one to quickly perceive the impending widespread production shortfalls affecting 72% of US shale oil production (43% of total US oil production). The GOR and Dmin Problems are arguably universally manifest across the vast majority of US shale oil production thus affecting 43% of total US oil production. Note: the 72% & 43% figures were calculated with the omission of the Delaware Basin which was not a part of the current study.

On the Nature and Character of the Widespread Oil Production Shortfalls - Figure 26 - oilandgas360

Figure 26: Composite plot of Figures 8-23 illustrating widespread manifestation of the GOR and Dmin Problems across 72% of US shale oil production (43% of total US oil production) Note: Delaware Basin production is excluded in these statistics

 

Conclusions

Review of current public data reveals that there is ample evidence to confirm the existence of the production shortfalls reported by the Wall Street Journal. Furthermore, given the use of conservative decline curve assumptions employed by the WSJ, the shortfalls presented in this work appear far more significant in magnitude than what was previously reported. Using rate-versus-cumulative oil produced plots makes it possible to efficiently approximate the implied shortfalls relative to existing forecast techniques and their corresponding type curves. It has been further shown that a common nature exists in the observed shortfalls. The Dmin Problem, a premature transition to constant rate decline (relative to a 3%-15% assumption), is a widespread occurrence. Additionally, the GOR Problem, an exponential increase in GOR, appears to correlate with – and often forewarn of – the imminent onset of the Dmin Problem. It is further concluded that since the Wall Street Journal compared (1) new “re -forecasts” which assumed ~30-year well life with single-digit terminal declines (Dmin), with (2) original corporate projections, actual shortfalls as identified in the present work appear substantially more severe when compared with the method employed by the WSJ. Based on the data and figures presented in this work, when compared to existing forecast methodologies, widespread production shortfalls in excess of 50% appear imminent across a material fraction of total US shale oil production.

References

  1. Olson, , et al, “Fracking’s Secret Problem —Oil Wells Aren ’t Producing as Much as Forecast, The Wall Street Journal, January 2, 2019
  2. Olson, and Elliott, R., “Frackers Face Harsh Reality as Wall Street Backs Away”, The Wall Street Journal, February 24, 2019
  3. Matthews, , et al, “Shale Companies, Adding Ever More Wells, Threaten Future of U.S. Oil Boom” , The Wall Street Journal, March 3, 2019
  4. Elliott, , “Frackers, Chasing Fast Oil Output, Are on a Treadmill”, The Wall Street Journal, April 8, 2019 5. Elliott, R., “ CEO of Shale Driller Pioneer Natural Resources Retires Abruptly ”, The Wall Street Journal ,

February 21, 2019

  1. Scheyder, , “ Investors push U.S. shale firms to separate executive pay from drilling ”, Reuters, September 29, 2017
  2. Scheyeder, , “ Shale firms pump up dividends as industry focus on returns grows ”, Reuters, March 25, 2018
  1. O’Donnell, , “Pioneer Natural Resources looks to cut costs with ‘generous’ buyout offer ”, The Dallas Morning News, April 5, 2019
  2. Eaton, , “ Pioneer Natural to cut spending, slow output growth in 2019 ”, Reuters, February 14, 2019
  3. Hiller, , U.S. shale producers turn to jobs cuts as investor pressures mount ” Reuters, April 9, 2019
  4. Arps, J., “Analysis of Decline Curves”, Petroleum Technology, Chapter 11 pages 228-247, September 1944 Wood Mackenzie, “Everything is accelerating in the Permian, including decline rates”, Insight, July 10, 2018 13. Pioneer Natural Resources Q1 2017 earnings release conference call transcripts, Seeking Alpha , May 5, 2018,

pages 12-17

  1. Pioneer Natural Resources Q2 2017 earnings release conference call transcripts, Seeking Alpha , August 2, 2018, Q&A page 20
  2. Lee, Texas A&M University; “Death by Bubble Point: Fact or Fantasy?” 2018 Ryder Scott Reserves Conference; September 13, 2018
  3. Fulford, S.; Apache Corp.; “ MULTIPHASE PRODUCTION FORECASTING “BUBBLE POINT DEATH?”
  4. Sheffield, ; Pioneer Natural Resources; first quarter, 2019 conference call transcript; May 7, 2019

 


Legal Notice