March 2014
Shale Technology Review

Best practices lead to successful shale fracturing

In liquids-rich shales, several fracturing designs may appear to work early in the development, but comparison of the produced fluids to what is known to be in the reservoir should be an assessment criterion.

George E. King / Apache Corporation
Multi-stage fracturing is one of the recent technologies leading to the greater recovery of shale resources. Image courtesy of Baker Hughes.

Describing best practices for any major function in the development or operation of shale wells is not a completely practical exercise: there are simply too many variances between different shales and even within the same shale, both areally and vertically. But even with the heterogeneity within shales, there are similarities in shale selection, productive target assessment, well construction, fracturing, flowback behaviors and production operations that encourage discussion of objectives and actions for improvements in large developments. One cannot speak of fracturing without mentioning the pre-fracturing developments that make fracturing technically successful, or the post-fracturing actions that make fracturing profitable.

There are many technically based actions and activities that have coalesced into what we call modern shale development. There is no single individual or company that can reasonably claim ownership of many of the inventions, innovations and adaptations that have incrementally, but steadily changed the landscape of gas and oil production in North America and will eventually change today’s unconventional producers into tomorrow’s conventional energy supply.

Technology impact in shale development has driven the recovery of original gas in place (OGIP) from a starting point of about 1% for shale developments in the eras through the 1990s, to between 10% and 30%+ today, depending on the specific shale, Fig. 1. Not every technology has worked in every shale, but every shale is likely to yield its resources to a specific technology. Whether the shale becomes profitable or not depends on the cost of leasing acreage, the necessary drilling and production technologies and the price obtained for the produced fluids. The early field-based research was largely trial and error, often practiced in the secret, to maintain competitive advantage as large areas of shale were leased and evaluated through the drill bit and frac pump.


Fig. 1. Some of the technical milestones in the development of shale gas fracturing.



The most economic shale wells have several things in common: the wells are in the best part of the shale, well construction yields a well suitable for multiple fractures, fracturing develops very large amounts of undamaged contact area with the reservoir and both natural fractures and created fractures remain stable through the life of the well. The next statement should be obvious: we’re not there yet for some shale gas wells and probably not there yet for many, if not nearly all, liquid-rich shale wells. Part of the reason is lack of knowledge and part is reluctance by the industry to drop the dream (nightmare?) of the “cookie cutter” approach of a manufactured well in every corner of every shale. Optimizing economics requires optimizing (but, perhaps, not maximizing) recovery. Technology is the answer; “use it or perish” appears to be the mantra.

Beginning with the first multiple fractured vertical wells in 1953 and the first multi-fractured deviated wells in 1974, the wellbore, be it vertical, deviated or horizontal, is viewed as a platform for fracture delivery.1-4 Lengthening the wellbore brings the opportunity for more fracture area contact. All of this is money down the drain, however, if the well or the hydraulic fractures are targeted to the wrong area.

Cumulative production maps of shales show a two order of magnitude range of hydrocarbon production even in the heart of proven shales such as the Barnett, where productivity ranking in the defined Wise-Denton-Tarrant three-county core area shows about 20% of the area was highly productive and roughly 25% of surveyed area was poorly productive.5 If early shale wells are considered, roughly one-third all wells were uneconomic, another one-third were marginal and the top third were so successful that they often carried the shale development project.


The answer to this “how” question is in recognizing how fluids move within the shales. This, in turn, will enable the selection of fracturing initiation points that will take advantage of both the flow path and which fluids actually move. Since shale matrices are so low in permeability that pure matrix flow could be expressed in times of millions of years, the created fracture contact with micro-cracks, fissures and natural fractures must be maximized. Current thinking is to shorten the fracture separation and focus on contacting as much formation area as possible along the wellbore. These short-spaced-laterals (SSL) define an attempt to create more contact and drainage area, although they still must match up with natural flow channels provided by the rock.

In liquids-rich shales, several fracturing designs may appear to work early in the development, but comparison of the produced fluids to what is known to be in the reservoir should be an assessment criteria.6-8 In many cases, lower permeability (matrix plus natural fractures) may flow gas and light-end liquids (C2 to C5), stranding the C6+ oils in the reservoir. This removal of drive fluid (gas) and solvent (C2-C5) indicates coming problems for high-liquid recovery.9

The position of fracture initiation points along the length of the lateral has been defined by simple mathematical proportioning, brittleness and ductility groupings, stress analysis and by aligning the fracs with hydrocarbon shows from a gas or oil analyzer as part of the surface mud log. The “best way” for now is, most likely, an operator preference. The mineralogy, stress or hydrocarbon show techniques are shallow readings at the depth of wellbore investigation and shales has been known to change significantly in the vertical and lateral directions by variations in depositional environment and/or post depositional events such as uplift. The show technique, particularly with analysis of the ratio of methane to the heavier gas fraction, may indicate the presence of natural fractures and hydrocarbon liquids-rich zones during drilling.6-8

The first wells in an area are nearly always the best wells in a shale, unless fit-for-purpose technology improves. Infield wells drilled years or even months after offset well production begins are usually lower productivity, but new wells with production compared on a bbl/frac or bbl/ft of lateral should be comparable, unless the drainage areas of the first wells are larger than expected. Occurrence of regional fractures or natural fracture sets is one reason for a larger than expected drainage area. Some fields are affected and some are not (Fig. 2), which may indicate a presence or lack of drainage enhancement.10


Fig. 2. Max. monthly production over time for adjacent fields—normalized per ft. of lateral length. Left: Declining first production over time—perhaps extended drainage area by fracture or channel. Right: Roughly equivalent first month production over time, indicating contained drainage area.

Regional fractures and even natural fractures with closed appearance, but wide extent in the reservoir, may effectively drain some of the trapped pressure. Therefore, a significant amount of the free gas or liquids may be produced by early wells. Natural fractures, even those closed and filled with calcite, are orders of magnitude higher permeability than the shale matrix.


Lateral length has been constantly increasing since the first 2,000-ft multi-fractured lateral in the Devonian shale was reported in 1989.11 Shale drilling is reportedly easy, although staying in the best part of the zone is a technical art. There are limits on lateral length, defined more by the ability to deliver a frac treatment and lease shape than by current technical ability. Every foot of wellbore creates a friction drop, and both rates and pressures are necessary to create an effective shale frac. Friction reducer performance, not just presence, is a key factor.

Wellbore position, including the orientation relative to the specific stresses of an area in the field, the vertical position within the pay thickness and the angle of the wellbore, may influence fracture initiation, fracture development type, liquid flow potential and ability to effectively remove liquids, both of production flow and liquids condensing from the gas stream.

Complex fracturing involves preferential opening of the natural fracture system, where natural fracture density and formation stresses will allow. When the minimum and maximum horizontal stresses are within about 5% and multiple natural fracture sets are present, the natural fractures, even those that appear sealed with calcite, can open at about 60% of the pressure necessary to fracture shale sections with no natural fractures.12 Field testing has shown that low-viscosity, slick-water fracturing fluid can penetrate, if the surface injection rate is ramped up in small steps of about 10 bbl/min from the start of pumping until the full injection rate is achieved.13

The issue of pumping rate is largely bracketed into the region of the highest rate that can be pumped while staying in zone. Step increase of surface injection rate may also help keep the frac in zone, particularly where complex fracturing is possible. Early fracturing work in the Devonian was done at 15 to 20 bbl/min, but the goal of high-productivity shale gas wells remained elusive until Mitchell Energy raised the rate in a near 100 bbl/min slick water job in the Barnett about 1999. Fracturing rates lower than about 30 bbl/min may not produce effective stimulation in some shales.14

Where the difference of horizontal minimum and maximum stresses are low and the formation contains multiple fracture joint sets, the possibility of achieving complex fracturing is increased. In these cases, fracture orientation is important. In any case, the wellbore should be oriented at an angle to the primary fracture direction; however, if the secondary fracture direction is perpendicular to the primary fracture direction and parallel to the wellbore, the secondary fracture direction may open a longitudinal frac along the wellbore and compromise subsequent fracture stage initiation and isolation.


Completion selection between cased, cemented and perforated (CCP) vs. packer & sleeve completions has grown more important with the acceptance of pad wells and the growth of a large number of frac stages. Although there are advantages and disadvantages to each system, the selection criteria usually revolve around time saved, equipment cost, number of wells available to stimulate, reliability of fracturing the specific zone and need for access after the frac. The percent of wells using CCP has fallen from around 95% in the 2008 time period to roughly 85% by the end of 2013.

Over the past decade, the shale gas frac stage length has fallen from 500+ ft to between 150 and 250 ft with closer spaced, more dense coverage of the wellbore becoming common, especially in liquids-rich shales. Perforation clusters as close as 25 ft have been attempted in the Barnett, but perf cluster spacings of 50 to 100 ft appear more common in current applications. The stress shadow effect, where a growing fracture stresses the formation, reducing potential for other fractures to initiate or grow, is a real effect, but naturally fractured formations tend to allow closer fractures, especially in complex fracturing. One key to breaking down multiple perforation clusters is limiting the number of perforations to achieve hydraulic diversion.15 While perforations show a minor amount of friction at rates of 0.5 to 1.0 barrel of surface injection rate per minute per perforation (bpm/perf), field and lab work have shown that maximum diversion, with manageable friction, occurs between about 2.0 and 2.5 bpm/perf (using ~0.5-in. entrance hole, deep-penetrating perforators).16


Fracturing processes in shales play by a different set of rules than most other formations, although the mechanics of fracture application are very similar. Slick water, with its ability to open and penetrate natural fracture systems, is usually the first fracture fluid considered, but its limitation of proppant transport is considered problematic for liquids-rich shales. Initial gel fracs in gas shales could produce some gas, but the switch to slick water was driven by both cost savings and higher IP and EUR achievable using slick water frac jobs.

Some operators approach liquids-rich shale with a hybrid frac job, using a slick-water pad and a low sand concentration, slick-water slurry to open natural fractures and initiate complex fracturing, and then switching to a gelled or cross-linked fluid for the last 30% to 50% of the job to carry larger proppant and higher proppant loadings. Other operators, however, have retained the slick water fracs in liquids-rich shales. Clear demonstrations of best systems for an area will take more time due to technique variations in fracture application. Personal preference will always play a role in well performance.


Proppant selection is another area where price and personal preference dominate. Liquids-rich shales usually benefit from larger, higher-quality proppant. A number of tests with good well populations support this contention,17 although, across the gas shales and even some liquid-rich shales, the dominant proppant size and type is 100 mesh (with a range of 70 to 140 mesh). The 100-mesh sand is a poor proppant in lab tests and a better bridging agent, but its performance in gas shale fracturing is undeniable. Its primary functions include blocking downward growth of fractures and wedging open natural fractures. Analysis of production flow indicates moderately well propped main fractures and poorly propped secondary fractures.


Pumping fracs in shale benefits from experience and comparison with earlier well production performance. While planar fracturing in shale is reasonably well described by bi-wing fracture models, prediction of complex fracture propagation by fracture models does not appear to agree with field data. Thus, complex fracturing design does not currently appear to be model friendly, at least to this writer. Microseismic mapping of the shear fracturing component of complex fracturing is one piece of the needed data, but the tensile fracturing component behaviors remain a bit of a mystery—perhaps awaiting technology development.

Net pressure increase during the frac appears to be useful. Limited net pressure increase and microseismic data support the basic contention that net pressure gains to the end of the job signal that the fracture has remained in zone, although the exact fit to a Nolte-Smith plot is sometimes elusive. The rate of net pressure increase has been shown in several jobs to preview whether a fracture treatment will screen-out. A sharp change in pressure trends in slick water jobs indicates that a potential screen out is rapidly approaching.


The last discussion in this short article is developing a knowledge base garnered from analysis of fracture flowback data and information available from comparing salinity changes, ion ratios and time-based fluid composition.18, 19 Specific ion concentration, ion ratios and total salinity are often functions of contact time and area-to-volume ratios. Interpretation of these data is helpful from both a diagnostic and optimization perspective. Shales are, indeed, very different when backpressure is applied during backflow of a fracturing treatment. Some shales require this control to improve recovery and others are damaged by even short shut-ins.


In conclusion, while absolute best practices are a moving target, some well-defined starting points are available that can help drive early optimization with the technology available. wo-box_blue.gif 


  1. Kruger, R. F., “Advances in well completion and stimulation during JPT’s first quarter century,” JPT, Dec. 1973, pp 1447-1462.
  2. Clark, J. B., C. R. Fast, G.C. Howard, “A multiple-fracturing process for increasing the productivity of wells,” API, presented at the Spring meeting of the Mid Continent District, Division of Production, Wichita, Kan., March 1952.
  3. Neill, G. H., R. W. Brown, C. M. Simmons, “An inexpensive method of multiple fracturing,” API, presented at the Spring meeting of the Southwestern District, Division of Production, Dallas, Texas, March 1957.
  4. Strubhar, M. K., J. L. Fitch, E. E. Glenn, “Multiple, vertical fractures from an inclined wellbore – a field experiment,” JPT, May 1975, pp 641-647, based on paper presented at SPE 49th annual meeting, Houston, Tex., Oct 6-9, 1974.
  5. Gonzales, R., “Barnett shale report,”, 2006.
  6. Pixler, B. O., “Formation evaluation by analysis of hydrocarbon ratios,” JPT, June 1969, pp 665-670.
  7. Kandel, D., R. Quagliaroli, G. Segalini, B. Barraud, “Improved integrated reservoir interpretation using gas while drilling data,” SPE Reservoir Evaluation and Engineering, December 2001.
  8. Elshahawi, H., M. Hows, C. Dong, L. Venkataramanan, O. Mullins, D. McKinney, M. Flannery, M. Hashem, “Integration of geochemical, mud-gas, and downhole-fluid analysis for the assessment of compositional grading,” SPE paper 109684, presented at the SPE ATCE, Anaheim, Ca., Nov. 11-16, 2007.
  9. Whitson, C. H., S. Sunjerga, “PVT in liquids-rich reservoirs,” SPE paper 155499, presented at the SPE ATCE, San Antonio, Texas, Oct. 8-10, 2012.
  10. LaFollette, R. F., W. D. Holcomb, J. Aragon, “Impact of completion system, staging, and hydraulic fracturing trends in the Bakken formation of the Eastern Williston basin,” SPE paper 152530 presented at the SPE Hydraulic Fracturing Technology Conference, The Woodlands, Texas, Feb. 6-8, 2012.
  11. Yost II, A.B., W. K. Overby Jr., “Production and stimulation analysis of multiple hydraulic fracturing of a 2,000-ft horizontal well,” SPE paper 19090, presented at SPE Gas Technology Symposium, Dallas, Texas, June 7-9, 1989.
  12. Gale, J.F.W., R. M. Reed, J. Holder, J., 2007, “Natural fractures in the Barnett shale and their importance for hydraulic fracture treatments,” AAPG Bulletin, 91, pp. 603-622.
  13. Overbey, W.K., A. B. Yost II, D. A. Wilkins, “Inducing multiple hydraulic fractures from a horizontal wellbore,” SPE paper 18249, presented at the SPE ATCE, Houston, Texas, Oct. 2-5, 1988.
  14. King, G.E., “Thirty years of gas shale fracturing: What have we learned?,” SPE paper 133456, presented at the SPE ATCE in Florence, Italy, Sept. 20-22, 2010.
  15. Murphy, W.B., A. H. Juch, “Pin-point sandfracturing—A method of simultaneous injection into selected sands,” JPT, Nov. 1960, pp 21-24, based on paper presented at 3rd annual Venezuelan meeting of SPE, Oct. 14-16, 1959, Caracas, Venezuela.
  16. Conway, M., Conversation on hydraulic diversion, August 2010.
  17. Besler, M.R., J. W. Steele, T. Egan, J. Wagner, “Improving well productivity and profitability in the Bakken—A summary of our experiences drilling, stimulating and operating horizontal wells, SPE paper 110679, SPE ATCE, Anaheim, Ca., Nov. 11-14, 2007.
  18. Bearinger, D.: “Message in a bottle,” SPE paper 168891, presented at the SPE Unconventional Resources Technology Conference, Denver, Co., Aug. 12-14, 2013.
  19. Ezulike, O., E. Ghanbari, H. Dehghanpour, D. Bearinger, “Flowback analysis for determining load recovery and its effects on early-time hydrocarbon production rate,” presented at SPE Advanced Technical Workshop on Fracturing Flowback, San Antonio, Texas, Nov. 7, 2013.
About the Authors
George E. King
Apache Corporation
George E. King is a registered professional engineer in Texas and Oklahoma with 43 years of experience since joining Amoco Research Center in 1971. His technical work has provided advances in shale developments, environmental risk reduction in well construction and fracturing, sand control reliability and underbalanced perforating. He has a BS degree in chemistry from Oklahoma State, and a BS degree in chemical engineering and an MS degree in petroleum engineering from University of Tulsa, where he also taught fracturing and completions courses as an adjunct professor. He is Apache’s Distinguished Engineering Advisor.
Connect with World Oil
Connect with World Oil, the upstream industry's most trusted source of forecast data, industry trends, and insights into operational and technological advances.