GUEST COLUMN
`
`1984 and Beyond:
`The Advent of Horizontal Wells
`
`Vik Rao, Senior Vice President and Chief Technology Officer, Halliburton
`
`The single most important productivity improvement in
`the history of the petroleum business may have been the
`implementation of horizontal wells. The engineering and
`economic challenges its early innovators faced were steep,
`but rapid advances between 1984 and 1994 progressively
`broke down the challenges. A Shell executive once confided
`to me that, in the early days of that period, one needed per-
`mission to plan horizontal wells, but by the late 1990s, one
`needed permission not to plan one. That is the hallmark of a
`“disruptive technology”—at first it is viewed with suspicion
`and elicits risk avoidance, but after industry acceptance, the
`technology becomes the norm and deviations from it are
`viewed with disapproval by the very people who questioned
`the technology in the first place.
`In the late 1970s, Teleco perfected the technique to measure
`well position and direction while drilling. Then it and others
`added important lithology-marker technology in the form of
`natural gamma and resistivity measurement. The early days of
`measurement while drilling (MWD) were marked by low reli-
`ability, but the industry persevered because of the cost savings
`in not having to stop to make openhole position measure-
`ments. Positioning in 3D space was now available on the fly.
`
`The First Reports
`Horizontal wells were still a curiosity. Then, in the early
`1980s, reports started trickling in of directional drillers try-
`ing something really different. They were making radical
`angular changes using a nonrotating drillstring, with a motor
`for propulsion and a bent sub for angle build. But instead of
`following convention, which called for pulling the string and
`
`Vik Rao is senior vice president and chief
`technology officer for Halliburton. He
`previously held executive management
`positions in research and development,
`product launch, reservoir studies, and
`sales and marketing. Rao joined the com-
`pany in 1974 as a senior research engi-
`neer and serves as a director on the boards
`of Fiberspar and Prime Photonics. He also
`serves on the advisory boards of KaDa Research, PointCross, the
`University of Houston School of Engineering, the University of
`Texas at Austin Petroleum Engineering Department, and Zebra
`Imaging. Rao is chairman of the SPE R&D Advisory Committee,
`the author of more than 20 publications, and has been awarded
`more than 15 patents. He earned a BS degree in engineering from
`the Indian Institute of Technology in Madras, India, and MS and
`PhD degrees in engineering from Stanford University.
`
`drilling the new section without the bent sub and motor, they
`drilled ahead with the assembly, this time rotating the string
`and providing motive power by the rotary and the motor. The
`bent sub in a rotary mode held angle, and the steerable system
`was born.
`
`Groundwork for Advancement
`I still remember reading the first such report—I thought the
`authors were nuts! Bent sub flopping around: What would that
`do to the hole shape, and what about stressing the string? Well,
`as it turned out, these were tractable issues and one more brick
`was in the wall to enable efficient angled drilling. Note that,
`once again, the advance was to eliminate a rig-time hog. The
`significance was that the early horizontal wells cost roughly
`2.7 times as much as conventional wells, and while well pro-
`ductivity was higher, reduction in well cost was an important
`objective in those days of decision silos that separated drilling
`and reservoir actions. There are some who believe, and I can be
`counted among them, that horizontal wells were a trigger for
`sustained integrated decision making, although clearly the shift
`to asset units, which occurred during the same time period, was
`a significant driver. Decisions about wells were made now not
`by functional units, but by asset teams made up of representa-
`tives from the functional units. These events, together with the
`key advent of formation evaluation while drilling (FEWD), laid
`the groundwork for this significant advance.
` In October 1985, two young Shell petrophysicists, Andy
`Greif and Craig Koopersmith, published a paper titled
`“Petrophysical Evaluation of Thinly Bedded Reservoirs in
`High Angle/Displacement Development Wells with the NL
`Recorded Lithology Logging System” in a relatively obscure
`forum (The Tenth Formation Evaluation Symposium, Canadian
`Well Logging Society, October 1985). The impact, however, was
`far from obscure. An MWD tool had made resistivity measure-
`ment of a quality that eliminated openhole wireline logs on the
`final 12 wells of the 24-well program on the Cougar Platform
`in the Gulf of Mexico. In this particular case, wireline logs of
`the time were incapable of detecting and evaluating the thinly
`bedded turbidite deposits. So, not only were they an effective
`substitute, but they were better. The previously passed-over B
`sand was now a prolific producer. The specialized application
`drove these young men to make the effort to seriously consider
`the new technology and, ultimately, to take the risk to elimi-
`nate the conventional crutch. Today, elimination of wireline
`logs in favor of FEWD is common.
`The success of the electromagnetic wave resistivity sen-
`sor spawned concerted activity in the industry, and the first
`quantitative porosity sensor appeared in 1987. By the end of
`the decade, the NL Industries (Halliburton today) offering
`was augmented by Schlumberger, and the FEWD industry
`
`118
`
`1 of 3
`
`JPT (cid:127) OCTOBER 2007
`Ex. 2092
`IPR2016-01517
`
`

`

`U.S. Horizontal Wells Per Year (Cumulative)
`
`12,000
`
`10,000
`
`8,000
`
`6,000
`
`4,000
`
`2,000
`
`0
`
`3,000
`
`2,500
`
`2,000
`
`1,500
`
`1,000
`
`500
`
`0
`
`U.S. Average Rig Count
`
`1984
`
`1986
`
`1988
`
`1990
`
`1992
`
`1994
`
`1996
`
`1998
`
`2000
`
`U.S. Average Rig Count
`
`U.S. Horizontal Wells Per Year (Cumulative)
`
`Fig. 1—The explosive growth of horizontal wells.
`
`was now in full stride. A footnote to this episode is that in
`general, the petrophysical community became progressively
`comfortable with accepting a lower-quality log and forming
`new judgments regarding fitness for purpose. With some sin-
`gular exceptions, such as the EWR application in turbidites,
`logs in logging while drilling (LWD), as it came to be known,
`were not as accurate as the wireline. This was particularly the
`case for porosity and density sensing. But, once again, the
`elimination of rig time was a key factor in the rationalization,
`and likely also the asset unit-based “common good” mentality.
`Also important was the fact that these were early measure-
`ments, prior to fluid invasion, and left time to make reservoir
`decisions. To misquote Mick Jagger, time was on their side.
`
`The Austin Chalk
`The technology table had been set. One could now drill a
`horizontal well using MWD for positioning on the fly, an
`important attribute for precise placement, a steerable assem-
`bly to obtain the needed trajectory and make course correc-
`tions on the fly without pulling the string, and then, finally,
`the ability to evaluate the reservoir adequately without using
`wireline logs. The logs were especially costly in a high-angle
`setting because of the nonviability of true wireline-conveyed
`systems at hole inclinations much greater than 50°.
`The first modern horizontal wells are generally credited
`as being drilled by Elf Aquitane in Lacq Superieur on land
`and in Rospo Mare offshore during 1980 to 1983. But a
`basically conservative industry needed one more push to
`drive wide-scale acceptance. This was the Austin Chalk
`play. Independent oil companies operating in this Texas area
`noted that the naturally occurring fractures were particularly
`amenable to production enhancement by intersection by
`horizontal wells. The first such well was drilled in 1985, and,
`over the next decade, there was explosive growth. By 1990,
`
`JPT (cid:127) OCTOBER 2007
`
`about 1,500 horizontal wells had been drilled; by 2000, there
`were between 12,000 and 20,000 (Fig. 1), and small compa-
`nies became big companies in very short time.
`In the early 1990s, a US Department of Energy survey
`showed that costs for horizontal wells were averaging only
`about 17% more than conventional wells, and that the
`productivity increases were between two- and seven-fold.
`Curiously, though, the quantitative formation evaluation
`impetus was a small factor, although LWD still remained a
`key enabler for horizontal-well exploitation of more con-
`ventional reservoirs. But, unquestionably, the Austin Chalk
`allowed the industry to cut its teeth on debugging and opti-
`mizing the technique of horizontal-well drilling. This factor,
`of compelling economics of a special kind, was not unlike the
`situation at Cougar for wireline replacement, and once again
`underlined one of the litmus tests for disruptive technology:
`It often finds a foothold in niche situations, but then blos-
`soms to become the norm in other.
` An intriguing feature of this period was that at the same
`time major advances in drilling were being made, the drill-
`ing industry itself was under pressure. Massive personnel
`cutbacks were occurring, and overall drilling activity was in
`decline. There was a dramatic rise in horizontal-well activity,
`but a drop in the rig count. Also of interest was the drop in
`development and lifting costs per BOE in the same period, as
`cataloged by the Energy Information Administration (Fig. 2).
`This near-halving of lifting costs is at least partly attributable
`to horizontal wells, even though they were in the minority
`of total wells drilled. There is also support for the hypothesis
`that horizontal wells, together with asset decision making,
`contributed to a shift from cost/foot thinking in drilling
`to cost/barrel thinking. This period also saw the practical
`realization of 3D-seismic interpretation, which increased cer-
`tainty regarding the location of sweet spots in the reservoir
`
`119
`Ex. 2092
`IPR2016-01517
`
`2 of 3
`
`

`

`Development and Lifting Costs
`
`$10
`
`$8
`
`$6
`
`$4
`
`$2
`
`$0
`
`Cost per bbl
`
`1984
`
`1986
`
`1988
`
`1990
`
`1992
`
`1994
`
`1996
`
`1998
`
`2000
`
`Fig. 2—Developing and lifting costs fell sharply during the period.
`
`and provided a firmer basis for the increased productivity
`likely from a horizontal well.
`
`Decisions Closer to the Field
`Major changes also occurred in the industry during this time.
`Oil companies underwent restructuring, in many cases with
`the formation of asset units, as noted earlier, thus shifting deci-
`sions closer to the field. Most firms drastically reduced R&D
`spending, and the onus for development activity progressively
`shifted to the service companies, which did not materially pick
`up the R&D spending slack until the mid-1990s (Fig. 3). This
`eventually led to an exacerbation of an industry problem: the
`slow uptake of technology compared with other industries.
`
`R&D Investments – Upstream Sector
`
`1,400
`
`1,200
`
`1,000
`
`800
`
`600
`
`400
`
`200
`
`4891
`
`0002
`
`E&P Firms*
`
`Oilfield Service Firms**
`
`*U.S. E&P firms and the U.S. R&D investments of international E&P firms,
`EIA, CERA analysis
`**Traditional oil field service companies (Baker Hughes, Halliburton,
`Schlumberger, Smith, Weatherford) annual reports, CERA analysis
`
`Fig. 3—The shift in R&D spending.
`
`120
`
`Some have theorized that the shift resulted in “information
`asymmetry.” In the original concept in economics, theorized
`by Nobel Laureate George Ackerlof, this results when the
`buyer has less information or understanding than the seller
`and, as a consequence, devalues the offering. An everyday
`example is the “lemon discount.” If the seller of a used car
`shares little information about the car, the buyer will assume it
`is a “lemon” and discount its value. In our industry, the most
`manifest result is likely risk aversion—the developers having
`less understanding of the precise need and the user less pro-
`ficiency in the technology. This hypothesis was discussed and
`developed at an Applied Technology Workshop (SPE 98511,
`Rao and Rodriguez).
` Causality is difficult to establish in most walks of life. The
`circumstantial evidence supports the theory that horizontal
`wells, arguably the single biggest productivity-enhancing
`technique in the business of developing and lifting hydro-
`carbons, were enabled by a series of events acting in concert
`between the mid-1980s and mid-1990s. It began with 3D-
`seismic interpretation coming into its own, with associated
`reservoir simulations identifying the high potential of hori-
`zontal wells. Steerable drilling systems, enabled by improved
`motors and the advent of MWD, reduced the cost of horizon-
`tal wells. Quantitative LWD permitted hydrocarbon satura-
`tions to be estimated in time for completion decisions. These
`were the technology underpinnings to change.
`Asset-decision-making framework, introduced at the same
`time, was a significant factor at a behavioral level. Finally,
`compelling economics were a driver for risk taking. This could
`lead one to conclude that disruptive technologies require a
`convergence of three factors: the right combination of enabling
`technologies, compelling economics that highlighted a niche
`play at first, and industry risk takers and/or a new organiza-
`tional dynamic. This period that began in 1984 seems to have
`experienced this unique combination of circumstances, usher-
`ing in a brave new world of lower lifting costs.
`JPT
`
`JPT (cid:127) OCTOBER 2007
`Ex. 2092
`IPR2016-01517
`
`3 of 3
`
`