Sei sulla pagina 1di 26

SPE 166107

65 Years of Fracturing Experience: The Key To Better Productivity Is Not


What We Have Learned But What We Have Forgotten and Failed To Utilize!
Frank E. Syfan, Jr., SPE, GoFrac LLC, Terry T. Palisch, SPE, CARBO Ceramics Inc., and Jeffrey C. Dawson,
SPE, Independence Oilfield Chemicals LLC

Copyright 2013, Society of Petroleum Engineers


This paper was prepared for presentation at the SPE Annual Technical Conference and Exhibition held in New Orleans, Louisiana, USA, 30 September2 October 2013.
This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not been
reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its
officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers is prohibited. Permission to
reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.

Abstract
Many engineers today do not have the training needed to fully understand the importance of fracture mechanics principles
and are easily overwhelmed in trying to deal with proper proppant and fluid selections, perforation design and strategy, and
on-site quality control of the fracturing process. The unfortunate reality is that many fracture designs are improperly
engineered with critical reservoir and hydraulic fracture parameters either ignored or improperly addressed. Many
completions are either marginally economical or produce at reduced commercial rates. Regardless of reservoir type, it is
critically important to achieve a highly conductive hydraulic fracture that provides connectivity between the reservoir and the
wellbore.
Since its inception, fracturing and completion knowledge has expanded exponentially allowing the oil and gas industry to
develop ultra-low permeability unconventional reservoirs. During the 1980s and 1990s technology pioneers such as
Holditch, Nolte, Warpinski, Veatch, and others, further developed the principles of fracturing which recognize the
importance of critical fracture parameters and their effect upon initial productivity and ultimate recovery. These gains in
expertise have resulted in unprecedented activity in the Bakken, Eagle Ford, Barnett, Haynesville, and Marcellus with
increasing activity in new reservoirs such as the Utica, Niobrara, and Mississippian. Although each of these reservoirs is
unconventional, each is uniquely different with respect to lithology, permeability, and hydrocarbon chemistry and interaction.
This paper will challenge the industry notion that infinitely conductive fractures are being placed in many unconventional
completions. It further addresses the critical fracturing parameters required to achieve a high conductivity fracture, why they
are important, and how to achieve proper proppant and fluid treatment designs. The importance of these fundamental
principles is documented and illustrated by several case histories which demonstrate the value of achieving high conductivity
fractures and the effects of improper design. This paper should be of great value to completion and operations engineers to
help further their knowledge with regard to the importance of fracture conductivity and connectivity in all hydraulic
fracturing applications.
Introduction
In 1947, Stanolind Oil and Gas (Company succession: Stanolind Pan American Amoco BP) introduced its
Hydrafrac process to the oil and gas industry and in 1949, the first commercial fracturing application was performed in the
United States. J. B. Clark (1949) of Stanolind Oil & Gas Company introduced the concept to the industry in an article in the
Journal of Petroleum Technology in January 1949 in which he identified six certain requirements of the Hydrafrac process
which must be met to be successful:

The hydraulic fluid selected must be sufficiently viscous that it can be injected into the well at pressure high enough
to cause fracturing.

The hydraulic fluid should carry in suspension a propping agent, such as sand, so that once a fracture is formed, it
will be prevented from closing off and the fracture created will remain to serve as a flow channel for oil and gas.

SPE 166107

The fluid should be an oily one rather than a water-base fluid, because the latter would be harmful to many
formations.

After the fracture is made, it is essential that the fracturing fluid be thin enough to flow back out of the well and not
stay in place and plug the crack which it has formed.

Sufficient pump capacity must be available to inject the fluid faster than it will leak away into the porous rock
formation.

In many instances, formation packers must be used to confine the fracture to the desired level, and to obtain the
advantages of multiple fracturing.

King (2012) reported at the SPE Hydraulic Fracturing Technology Conference in February 2012 that there have been over
2.5 million fracturing treatments pumped worldwide with over one (1) million pumped in the United States. Fortunately for
the oil and gas industry, well stimulation technology has advanced extremely rapidly, especially during the past two decades.
Unfortunately, however, most production, completion, and operations engineers have not had enough time to properly
analyze and apply all the new data, materials and techniques. Many engineers today have been relegated to more of a project
management role and thus, have not been properly trained to understand the principles that govern fracture and rock
mechanics. As a result, many fail to properly design optimal fracture treatments which achieve high conductivity and high
dimensionless conductivity fractures, resulting in losses not only in initial production rate, but ultimate recovery, Net Present
Value (NPV) and Internal Rate-of-Return (IRR).
Moreover, a great number of designs today are created using spreadsheets which rely on oversimplified rules-of-thumb, such
as pounds proppant per vertical/linear foot and gallons fluid per vertical/linear foot. In efforts to save money, many
companies lower completion costs without realizing the tremendous losses in deliverability, net income, IRR, and NPV
inflicted by improper fracture and completion design.
Rock mechanics have always dictated the pressures, width, length, and height of a created hydraulic fracture. In the early
1960s, over 50 years ago, fracturing pioneers T. K. Perkins and L. R. Kern (Perkins and Kern (1961)) correctly stated that
zones with higher in-situ stress found above and below the initiating zone will cause a vertically limiting effect and that the
vertical hydraulic fracture would grow until it reached the bounding higher stress zones which will then restrict vertical
height growth. Without a thorough understanding of rock mechanics, fracture design parameters, and critical fracture
parameters, optimization of completions is virtually impossible and many completions achieve very low dimensionless
fracture conductivities. In actuality, lower fracture conductivities can and do reduce both initial production rates and ultimate
recoveries as evidenced by Lehman et al. (2003). Further, the effects of conductivity reductions and the need for maximizing
fracture conductivity was presented in a Gas Research Institute (now officially named Gas Technology Institute) Report
authored by Peterson et al. (1991), and later detailed by Palisch (2007). Syfan and Anderson (2011) presented work which
quantified the staggering losses in initial production rates, ultimate recovery, and NPV losses in the Eagle Ford and Marcellus
formations as a result of improper designs with respect to applications which exceed the recommended individual operating
ranges for various Northern White quartz and substandard proppants.
To challenge the notion that infinitely conductive fractures are being placed in many unconventional completions, four (4)
case histories were examined to illustrate the benefits of proper fracture design. The four cases include a completion which
achieved only a low-medium dimensionless fracture conductivity (Case A Marcellus Shale), a properly designed
completion which achieved high fracture conductivity (Case B Eagle Ford Shale), and two cases which represent the
benefits of improved fracture conductivity in the Bakken (Case C) and Cotton Valley (Case D). Case A was further
examined to determine the economic losses associated with the low-medium conductivity and also examined the benefits of
proper completion design to achieve a much higher conductivity. The results of each case are presented and discussed which
includes an economic assessment of the Case A depicting the gains that could have been achieved by proper fracture design.
This paper is intended to give inexperienced engineers a guide to understanding rock mechanics, fracture design, and critical
fracture parameters as they relate to modern completions in an effort to increase the overall effectiveness of the thousands of
wells drilled, completed, and hydraulically fractured annually.
Fracture Propagation Theory
The general equations and constitutive relationships governing fracture propagation are summarized below. These equations
illustrate the importance mechanical properties, injection rate, fluid rheology, and confining stress contrast have on the
fracture propagation process. The governing mass, momentum, and energy equations and constitutive relationships are
presented below (Meyer (2011)).

SPE 166107

Mass Conservation
The governing mass conservation equation, presented in Eq. 1, for an incompressible slurry in the fracture, states that the
volume of slurry injected into the fracture minus the volume loss by leakoff, Vl (t ) , and spurt loss, Vsp ( t ) , must be equal to the

fracture volume, V f (t ) .

t
0

q( ) d Vl (t ) Vsp (t ) V f (t )

(1)

Continuity

The governing continuity equation, presented below in Eq, 2, for incompressible slurry in the fracture, in terms of the fluid

velocity, v , is:


vw 2vl w t 0

(2)

Where, vl is the fluid leak-off velocity and w is the fracture width at any position.
Momentum Equation

Eq. 3, the momentum equation (equation of motion) for steady state flow, can be written as:

1 f v 2
p
2 dh

(3)

Where, f is the Darcy friction factor which is a function of the Reynolds Number, Re , and fracture wall roughness.
Width-Opening Pressure Constitutive Relationship

The crack-opening (aperture) and opening pressure relationship from Meyer (1986, 1989) for fractures in each of the
principal planes ( ) is of the form depicted in Eq. 4:
w ( x , y , z ) w

2 H ( p )
E

2 H p
E

(4)

Where, w is a generalized influence function, H is a characteristic fracture half-height (or length), w is the fracture width,

2
p is the fracture pressure, E E 2 1 is the effective Youngs modulus, is the confining stress and

p p is the net fracture pressure for each discrete fracture in the three principal planes ( ).

Fracture Propagation Models

There are numerous 2-D, Pseudo 3-D and fully 3-D models available to the petroleum industry and are well documented in
the petroleum literature. Other hydraulic fracturing models have been specifically developed for discrete fracture networks
(DFNs). Although many fracturing treatments result in relatively simple bi-wing fractures because they require less energy
to propagate, there are natural systems that force the boundary condition to create discrete fractures requiring even less
energy than a bi-wing. This is most easily understood by examining limited entry designs or by perforating large intervals in
highly deviated wellbores. Both of these are examples of creating multiple fractures as the result of a boundary condition.
All of these models are invaluable in understanding fracture propagation modeling and design in formations with spacially
varying mechanical properties.
Understanding 2-D fracture propagation theory is critical to understanding the primary effect of various parameters and in
performing parametric studies. The two most famous 2-D models are the Geertsma-de Klerk (1969) and the PerkinsKern/Nordgren (1961) type geometry models. The Geertsma-de Klerk (GDK) type fracture model is a constant height model
that assumes horizontal plane strain with the characteristic half-height equal to the fracture length (one-wing or half-length).
This model is most applicable for fracture length to half-height ratios less than one (L/H<1). The Perkins-Kern/Nordgren
(PKN) type geometry model is a constant height model with an elliptical shaped width profile vertically (vertically bounded
geometry which assumes vertical plane strain). This model is most applicable for fracture length to half-height ratios greater
than one (L/H>1).
Some of the functional relationships between various fracture parameters and their effect on fracture characteristics and
pressure responses are illustrated in Table 1 (Hagel et al. (1992)). Table 1 shows the effect of various parameters on fracture

SPE 166107

length, width and net pressure for PKN, GDK and Radial type 2-D fracture models for viscous and toughness dominated
fracture propagation. The viscous equations are for laminar flow with negligible toughness and no spurt loss. The toughness
equations are for negligible viscous dissipation. The penny shape model is referred to as the Sneddon model for toughness
controlled propagation.
Parameters with the largest exponents have the greatest influence on the specific fracture characteristic. Therefore, more
emphasis should be put on refining these critical parameters. A systematic approach is a good method of determining
parameters which best match the fracture characteristics and response. The proportionality equations in Table 1 can be used
to refine input data and to determine parameter sensitivity for 2-D type models. The parameters which affect net pressure the
most are: Young's modulus, fracture height and viscosity for the PKN model. To match the net pressure in a GDK model
only the fluid rheology or Young's modulus can be varied to get a match assuming negligible toughness. The net pressures
for the GDK and radial (penny) models are shown not to be a function of fracture height.
The connected-cluster simulation assumes that the multi-clusters create a single DFN that interacts with the other clusters.
That is, fractures from the separate clusters may coalesce with other secondary fractures created from the individual clusters.
This assumption may be the best limiting case in that secondary fractures created by individual clusters can occupy the same
secondary network system rather than assuming separate interacting fractures in the principal planes.
A DFN numerical simulation and pressure match for a Marcellus shale well was presented by Jacot et al. (2010), describing
technology integration as a methodology to enhance production and economics in horizontal Marcellus shale wells. Fig. 1
illustrates the surface and bottomhole treatment pressure match. The dominant fracture characteristics (width contours,
height, length, etc.) are shown in Figs. 2 and 3, respectively. This numerical simulation illustrates the effect of fracture
growth and propagation in complex formations.
Table 1
Two-Dimensional Hydraulic Fracture Parametric Equations

Rock Mechanics
Mineralogy

Knowledge of formation mineralogy and composition is critically important information when designing the fracturing
treatment regardless of whether the formation is predominately sandstone (quartz), fractured shale, limestone or dolomite, or

SPE 166107

a stratified combination of all. The primary components of sandstone are quartz, cementing material such as calcite, clay, or
silica, and clay (which could be smectite, chlorite, kaolinite, illite or multiple combinations of same). The primary
components of shale are quartz, calcite and clay, the clay often being mixed layered clay, illite-Smectite. However, many of
the shale prospects in development today are not technically shales, in the strictest geological sense; rather they can be
considered fine-grained clastics that exhibit ultra low permeability (Akrad et al. (2011)). Furthermore, the better prospects
possess clay content less than 35%, with higher clay content often too ductile to maintain conductive fractures, although the
Haynesville is the exception. The higher content of either quartz or calcite exhibit more brittleness and appear to retain
greater fracture conductivity. To show the variation in shales, Akrad et al. summarized the composition of several shale
reservoirs in Table 2.

Fig. 1 Pressure History Match for Well A (Jacot et al.


(2010))

Fig. 2 Stress, Width, and Length Profiles for Stage 2 Dominant


Fracture (Jacot et al. (2010))

Fig. 3 DFN Simulation: 3D Major Axis Partial View (X-Z Plane) (Jacot et al. (2010))

Table 2
Composition of Samples from Various Shales

Composition
Illite-Smectite (%)
Quartz (%)
Calcite (%)
Pyrite (%)
Dolomite (%)
Porosity (%)

Lower Bakken
47
21
0
13
10
16.0

Middle Bakken
4
11
77
1
4
0.8

Barnett
21
59
12
2
1
5.4

Eagle Ford
8
3
77
6
2
4.8

Haynesville
57
28
2
5
0
9.7

Furthermore, it is well documented in the industry that certain minerals can have highly adverse reactions with fracturing
fluids and acids, resulting in insoluble precipitates that permanently plug formation pore throats and proppant packs.
Poissons Ratio

Poissons ratio (Greek symbol (nu)) is defined as the negative ratio of transverse to axial strain. When a core sample, or
other material, is stretched to an extension or compressed to a contraction in the direction of the applied load, it corresponds
to a contraction or extension in a direction perpendicular to the applied load. The ratio between these two quantities is

SPE 166107

Poissons ratio. If the material being tested is stretched or compressed along the axial x-direction, the equation for Poissons
ration is then represented by Eq. 5.

d
d trans
d
y z
d axial
d x
d x

(5)

The effects of slight changes in Poissons ratio are well documented in the literature, e.g. Perkins and Kern (1960). In
addition, the values of Poissons ratio for most rocks reported in the literature range from about 0.05 0.25 (Birch, F.
(1942)), (Cleary, J. M. (1959)), (Wuerker, R. G. (1956)). From the existing research it is apparent that the minimum fracture
extension pressure is not very sensitive to slight changes in Poissons ratio. However, width calculations are highly
dependent on values of Poisson's ratio and if incorrectly applied can result in significant errors in calculated width, which will
affect the calculated height, created length, and fracture volume.
Youngs Modulus/Modulus of Elasticity

Modulus of Elasticity, or elastic modulus, is defined as the mathematical description of an object or substances tendency to
be deformed elastically (i.e., not permanently deformed) when a force is applied to it. The elastic modulus of an object is
defined as the slope of its stress-strain curve in the elastic deformation region. As such, the stiffer the material is, the higher
the elastic modulus. The equation for Elastic Modulus is presented below in Eq.6, where lambda () is the Elastic Modulus,
Stress () is the restoring force divided by the area to which the force is applied (for industry applications, psi), and Strain ()
is the ratio of the change caused by the stress to the original state of the object. For example, if Stress is measured in poundsforce per square inch (psi) and Strain is a dimensionless quantity, then the units of Elastic Modulus are stated in psi.
def

Stress
E
Strain

(6)

Depending on how Stress and Strain are measured, which includes the stress direction, allows for many types of Elastic
Moduli to be defined. For instance, Youngs Modulus describes the materials response to linear stress and strain. The Bulk
Modulus describes a materials response to uniform pressure and Shear Modulus describes a materials response to shearing
stresses and strains. Bulk Modulus, Shear Modulus, and Youngs Modulus are the most common.
Youngs Modulus (E), used most commonly in hydraulic fracture simulation models, describes tensile elasticity, or the
tendency of an object to deform along an axis when opposing forces are applied along that axis, and therefore is defined as
the ratio of tensile stress to tensile strain (t/t). Youngs Modulus, also known as the tensile modulus, and Poissons Ratio
are two of the most important rock mechanics values used in hydraulic fracturing. They are primary quantities which define
width calculations and therefore will also dictate calculated fracture length and fracture height. If these values are incorrectly
estimated for a given application, fracture geometry cannot be accurately calculated and again, optimization will not be
possible using predictive modeling.
In-Situ Stress

Gidley et al. (1989) discuss the effects, calculations, and role of in-situ stress in depth in SPE Monograph 12, Chapter 3.3.
The authors point out that in-situ stresses are clearly the single most important factor controlling hydraulic fracturing. The
maximum horizontal stress (Hmax), minimum horizontal stress (Hmin), and vertical (V), or overburden, stresses control
fracture azimuth, vertical/horizontal orientation, total and directional height growth, surface treating pressure, bottomhole
treating pressure, net closure stress and proppant crushing, embedment and fracture cross-sectional width profiles, just to
name a few.
Hubbert and Willis (1957) performed experiments to determine the effects of stresses on the fracturing process and stated that
if fluid pressure is applied to the rock, the pressure will increase to the point that the rock ruptures, or fractures, and the rock
parts. Thus, the formation will always fracture when a force is applied to the rock that exceeds the minimum horizontal
stress plus the pore pressure. The fracture will then propagate perpendicular to the least or minimum, horizontal stress
azimuth. Since the in-situ stresses are so critical in properly planning for and designing hydraulic fracture stimulations, the
problem has been addressed by many and is well documented in the literature.
Hubbert and Willis (1957) believed that the minimum horizontal stress was between a third and half of the effective stress
and therefore the fracture pressure could be determined from the following equation:

F S P / 3 P
Where, F is the fracture pressure, S is the overburden stress, and P is the formation pore pressure.

(7)

SPE 166107

Eaton (1969) believed that the rocks deformed elastically and thus, introduced Poissons ratio into the calculation, presented
in Eq. 8:

F
S P P
1

(8)

Eatons work was then further refined by Daines (1982) to include the variable T as the superimposed tectonic stress.
Daines formula is still widely used today as it allows fracture pressure to be calculated while drilling and is not dependent on
previous data being available from a specific region. In areas or formations where no engineering data is available on in-situ
stresses it is possible to estimate the horizontal stress which is useful not only in drilling applications but can be used to
estimate horizontal stress in fracturing operations and is presented in Eq. 9:

Pp Pp T

(9)

Closure Pressure

Nolte (1988) defined Closure Pressure as the fluid pressure required to initiate the opening of an existing fracture. Therefore,
this Closure Pressure should be equal to and counteract the stress in the rock perpendicular to the fracture plane, which is
known as the minimum principal in-situ horizontal stress

Pc min .

This statement holds true if there is no fluid

movement or residual width. The final Closure Pressure changes in the case of a propped fracture and will be equal to the
minimum horizontal stress plus the additional net pressure caused by introducing proppant to the created fracture, thus
increasing width. Many engineers mistakenly use the value of ISIP (Instantaneous Shut-In Pressure) obtained during pre-frac
testing or early fracture stimulation (e.g. pad portion of the stage) to represent Closure Pressure. It is important to note that at
the point of ISIP the fracture is still open, and thus, the actual fracture closure pressure (to zero width on an unpropped
fracture) will be a value less than the ISIP. Only if the fracture width at the moment of ISIP is, by coincidence, identical to
the width after closure on a proppant pack, can this value be used to estimate the stress on proppant. If the wrong closure
pressure is used during the stimulation treatment, then all calculations and plots, such as a Nolte or net pressure plot, will be
incorrect.
Fracture Toughness and Critical Stress (Tensile Strength)

The criteria for fracture propagation is based on the concept of stress intensity factor K I for Mode I failure. In hydraulic
fractures, propagation is assumed to occur once the stress intensity factor reaches a critical value. This critical value is
related to the propagation resistance (or energy balance) and assumed to be a material property and is given the name
fracture toughness (or critical stress intensity factor). Fracture toughness is not the same as tensile strength, but the two
parameters can be related. The basis for this relationship involves the assumption that pre-existing defects exist and induce
high stress concentrations in their vicinity. Thus the fracture will propagate when the stress intensity factor equals the
fracture toughness, K IC , or the stress intensity equals the critical stress of the rock: K I K IC or I IC , whichever is
greater, (Vejbk et al. (2013)).
Thus, the net pressure in the crack (fracture) must overcome this critical value to propagate p c K IC

H . Defining

a minimum critical stress as the minimum stress value for failure to occur based on a maximum effective crack length of H c ,
we have c

min

K IC

H c . The critical stress is the minimum stress ( c

min

) for the fracture to propagate in the vicinity

of a constant stress field. This parameter may also be thought of as the apparent tensile strength ( c

min t

) since it is the

critical stress that must be overcome for the crack to propagate (in a uniform stress field). Fig. 4 illustrates that as the fracture
dimensions increase either the maximum of the critical stress (calculated from the fracture toughness) or the minimum critical
stress (i.e., tensile strength) will determine the stress intensity factor.
Fracture Design Parameters
Fluid & Additive Design

Hydraulic fracturing fluids are designed to transmit hydraulic pressure from the surface to the tip of the propagating
fracture(s) and transport proppant from the surface to the fracture. The job design is more complicated than simply using the
fluid to transport proppant. The treatment design uses a sophisticated schedule of proppant concentration to optimize
proppant placement in the fracture, promoting high fracture conductivity. Historically, conventional reservoir fracturing
required well designed, viscous fracturing fluids that could carry high volumes of proppant needed to establish highly

SPE 166107

conductive proppant packs. The shift from conventional to unconventional reservoirs such as coal-beds, tight gas sands and
later gas and liquid-rich shales radically changed the design criteria of the fracturing fluid, particularly a shift from viscousbased fluids to high-rate capable, non-viscous fluids needed for slick-water fracturing.

Fig. 4 Critical Stress Propagation Dependence On The Minimum


Critical Stress (Tensile Strength) and Fracture Toughness

As a guide for designing well-defined fracturing fluids, the API RP-39 was established as the industry standard for testing
fracturing fluids. API RP 39 was later upgraded with the following ISO (International Organization for Standardization)
documents.

ISO 13503-1:
ISO 13503-3:
ISO 13503-4:

Measurement of Viscous Properties of Completion Fluids


Testing of Heavy Brines
Procedure of Measuring Stimulation and Gravel Pack Fluid Leakoff Under Static Conditions.

Based on success in the Barnett, fracturing designs and fracturing fluids were radically changed in a number of resource
plays. The primary driver for this shift is the belief that successful stimulation required extensive reservoir contact or fracture
networks in the shale rather than a bi-wing fracture. In addition, the permeability contrast between the fracture and the ultralow permeability of the gas shale, normally in nanoDarcy range, is so large that fracture conductivity compromises have been
tolerated. Consequently, part of these compromises included fluid designs that carry much lower proppant loadings,
normally maximum capacity ranging from 1-2 lb proppant per gallon treating fluid, as well as smaller size proppant such as
70/140 and 40/70. Based on these drivers, frac designs evolved, using high rate fluids for proppant placement and negating
the need for viscosity. This also required tremendous volumes of fluid to place the proppant and create the desired fracture
network. Fluid designs also shifted from using guar gum based fluids with viscosity that could range from 10s cP for linear
gel to 100s cP for crosslinked gels, to new fluids based on low concentrations of polyacrylamide (75 to 300 parts per
million) for friction reduction needed when pumping fluids at high rates, creating the term Slickwater. In contrast, the
viscosity of slickwater ranges in the 1s cP. However, as the slickwater fracturing process evolved, late stage pumping began
using low viscosity fluids to improve the near-wellbore conductivity with more and larger size proppant. This practice
further evolved, using complex crosslinked fluids in the hybrid fracs for the same purpose, to pack more proppant (and larger
diameter proppant) in the fracture in the late stages of the frac. The pumping of these complex fluids, normally boratecrosslinked guar gums, on the tail-end of each stage did not require long-term gel stability. These fluids were built to provide
some viscosity but also to be easily degraded. Consequently, the fluid formulations were designed without relying on ISO
13503-1 guidelines and normally based on a viscosity profile measured at a single shear rate.
As gas prices deteriorated, emphasis shifted from gas shales to liquids-rich shales, relying on many of the principles
developed for shale gas fracturing. However, oil/condensate, having substantially less mobility than gas, requires the
fractures possess higher conductivity. Crosslinked fluids are commonly used for a larger portion of the treatments targeting
oil plays, so that higher concentrations and larger diameter proppants can be placed. Borate crosslinked guar gum is the
preferred fracturing fluid for this application with gel loadings often ranging from 15 to 30 lb per 1000 gal. In order for these
fluids to provide the necessary viscosity needed for shale oil fracturing, the fluid design should be based on ISO 13503-1
guidelines and with particular focus on the Power Law indices, n (flow-behavior index) and K (consistency index), as a
guide in the rheological assessment of the fluid. In particular, the fluid should be formulated with the n range from 0.4 to
0.7 and the K exceed 0.01 lb-secn/ft2 through most of the pump time. Without using the Power Law indices as a fluid
formulation guide, borate crosslinked guar gels can appear viscous but exhibit shear thickening behavior with n exceeding
1.0, indicating a poorly designed fluid. By utilizing the Power Law indicies, the crosslinking buffer and borate crosslinker
loadings can be adjusted to provide higher-stability fluids, especially since the formulation should also include a breaker
designed to slowly degrade the fluid.

SPE 166107

The breaker schedule and type can be the most critical feature of the fluid design. There are a number of breakers that are
available including persulfate, encapsulated persulfates, chlorites, organic hydroperoxides, alkaline earth peroxides and
enzymes, each having their specific stability and temperature requirements. There is normally a compromise between initial
front-end stability for the fluid to perform its functions and the later degradation of the fluid to allow well clean-up.
Regardless of the fluid design, slickwater or viscosity-based, the fluid will also contain an array of other treatnment additives,
each needed to prevent reservoir specific flow impairment after the treatment. Common additives will include a biocide, clay
control additive, scale inhibitor and flow-back surfactant additive, each additive and loading depending on the reservoir being
treated. Furthermore, the additive application, type and loading are often scaled as a way to adjust the cost of the treatment.
To insure that adequate volumes of water are available at low cost, the water quality is often compromised requiring the
treatment to always include a biocide, particularly to control sulfate reducing and acid producing bacteria. The two main
modes of bacteria control are the use of short-lived oxidizing agents such as chlorine dioxide, peracetic acid or ozone or
longer lasting chemical agents such as dibromonitropropamide (DBNPA), Glutaraldehyde, Glutaraldehyde-surfactant-based
quaternary ammonium chloride mixtures, tetrakis (hydroxymethyl) phosphonium sulfate (THPS) and Dazomet. The
selection of the biocide is often based on previous field success or cost.
The inclusion of the treatment additives in the fracturing fluid was initially done to protect the reservoir from the invasion of
the fracturing fluid and assist to clean-up after treatment. Since hydraulic fracturing is an opportunity to place these additives
deep in the reservoir, the additives used today are expected to provide longer-term protection before production chemical
intervention is required. To determine the longevity of the additive, many companies have resorted to post-fracturing
sampling and analysis to include, for example, scale inhibitor residuals, bacteria counts, surface tension and general water
analysis. These results are used to drive a continuous improvement of the additive selection.
Proppant Design

Proppant design has been a subject that has been published on numerous occasions, and therefore, this paper will not attempt
to repeat the results and experimentation pertaining to that work. This paper will address the factors which are considered
critical to proper proppant design and references will be cited, to provide the reader with complete information.
If we were to summarize the characteristics associated with an Ideal Proppant, i.e. one that will fit all applications, the list
would include high strength and crush resistance, high conductivity, durable conductivity, complete system compatibility, no
embedment or flowback, nearly buoyant, plentiful, and low cost. While several of these criteria are extremely important,
such as crush resistance & high strength, no flowback, complete system compatibility, and cost effectiveness, the truth is that
the Ideal Proppant simply does not exist!! There is not a proppant manufactured or mined today, that can accomplish all of
these items for every application and still be cost effective.
Selecting the proper proppant (-Quartz, Resin Coated (RC) Proppant or Ceramic) cannot be performed by looking at a
single reservoir/formation parameter, such as stress, as this approach will rarely yield the optimal completion. For given
values of closure stress and also because of the rock mechanics associated with differing formations, many of the selection
factors discussed below cannot be eliminated, nor can they be used solely for optimal proppant selection. Consequently,
proper engineering is a critical step in assuring optimal conductivity within the in-situ fracture by minimizing the effects of
those selection factors over which control is limited.
Infinite vs. Finite Conductivity
The terms infinite and finite conductivity refer to the relationship between the flow capacity of the fracture to the
deliverability of the formation. Mathematically, this ratio is simply the Dimensionless Fracture Conductivity (see Equation
16 below CfD). As CfD increases, the drawdown pressure effectively transmitted to the fracture face increases and the
reservoir will deliver more hydrocarbons. However, there are diminishing returns, and at some point is CfD so high that the
fracture flow capacity outruns the deliverability of the formation. When this occurs, the conductivity is considered
infinite when compared to the formation deliverability. In other words, increasing the conductivity further will not produce
a significant increase in production. Conversely, a fracture that has finite conductivity will see an increase in production if
conductivity is increased.
For many years it was thought that many hydraulic fractures were infinite conductivity. Many also mistakenly believe the
same thing about the hydraulic fractures placed in most of todays ultra-low permeability unconventional reservoirs.
However, it is well documented that this simply is not the case. As the industry has come to understand and model the
difference between reference (API/ISO) conductivity measurements, and actual realistic conductivity of the fracture, it is now
apparent that most fractures have finite conductivity, or are conductivity-limited (Palisch 2007). This means that even in
the unconventional reservoirs, particularly those that produce liquids or condensate, increasing the fracture conductivity will
increase production - the only question is whether it is economic to do so (Palisch 2012).

10

SPE 166107

Depth/Closure Stress
The vast importance of depth and closure stress and its ultimate effect on conductivity cannot be overlooked when
considering proper proppant design. At some point all proppants undergo breakage of the individual grains through natural
stresses and depletion, temperature, induced cyclic stresses due to variable flowing conditions, or workovers executed with
less than best practices. Regardless of the source the sudden changes in effective closure stress, the strength of the proppant
material can easily be exceeded. Even the highest quality quartz proppant is weaker than most ceramic proppants, but both
can be seriously damaged by sudden changes in effective stress. This results in lost and unrecoverable flow capacity, i.e.
fracture conductivity. Additional damage to productive capacity occurs due to the movement of very fine broken grains, or
fines, which are transported through the pack by higher flow velocities resulting from damaged pack permeability. These
produced fluids are capable of moving fine proppant shards through the pack and further damaging productivity by plugging
additional conductivity.
In a paper presented by Syfan and Anderson (2011), six well conditions and proppant selection criteria were identified that
influence obtainable conductivity. The six factors include closure stress, temperature, proppant concentration, proppant
strength, proppant particle size, and proppant grain shape. A paragraph explaining the importance of proppant durability is
also included. The results of that work are presented below in the following paragraphs and are also explained in further
detail in the procedure provided in ISO 13503-5. ISO 13503-5 takes each one of these six factors into account when
performing a conductivity test on a specific proppant.
The number of specific proppant articles that address these six factors is extremely well documented with over 300 specific
articles in the petroleum literature since 1980. For times sake, a synopsis of much of the literature is available from Holditch
et al. (1989) Anderson and Phillips (1988) and Palisch (2007).
Closure Stress
Natural alpha-quartz (-quartz) Northern White (NW) proppants with roundness and sphericity values (Krumbein factors)
greater than 0.6, will begin to fail at as little as 4,000 psi, even when tested under ideal conditions with wide fractures
uniformly loaded with proppant and subjected to stable stresses for short time periods. Brown sands usually tend to fail at
much lower closure pressures due to the inclusion of additional minerals such as iron and feldspar. The quartz grains will
continue to break or crush to a point in which the amount of fines generated reaches and exceeds 10% of the original
proppant. This value, documented in ISO 13503-2 guidelines, is referred to as the K-Value.
K-value is determined by performing several crush resistance tests on a proppant sample, with the amount of fines generated
recorded for each test. Closure stress or load is increased in 1,000 psi increments on subsequent tests until such a load is
reached at which the amount of fines generated exceeds 10% by weight of the sample. Verification of the stress at 10% fines
is accomplished by lowering the stress by 1,000 psi and repeating the test on a new sample. This process of reducing the
stress and repeating the test is continued until the stress at which the fines generated exceeds 10% is bracketed by crush tests
at two consecutive stress levels. The k-value is the stress at which the amount of fines generated does not exceed 10%. For
example, if a 20/40 quartz proppant has 9.3% fines at 7,000 psi closure stress and then at 8,000 psi closure stress has 10.5%
fines, the associated K-Value for that proppant would be 7K. It has been observed that certain high quality NW mines
produce 40/70 quartz proppants capable of withstanding loads in the 11,000 psi range before the fines generated number
exceeds 10%, thus a K-Value of 11K.
It is important to clarify however, that a 7K 20/40 ISO Crush DOES NOT mean that these proppants will optimally
perform at closure stresses of these magnitudes. K-Value is simply a relative number and can only be used for comparison
purposes if the two samples have exactly the same mineralogy and sieve distribution. Coulter and Wells (1972) indicated
that as little as 5% fines in the proppant pack can reduce flow capacity as much as 60%! In the end, conductivity
requirements and performance at downhole conditions should be the primary tool for optimal proppant selection.

Durability
G. S. Penny (1987) documented that the elements of both time and temperature at closure pressure are major contributors to
long-term conductivity decline. With time, all proppants lose permeability due to proppant pack rearrangement, stress
cracking, and solubility. Pennys research led to the implementation of the modified API conductivity test, which (among
other changes) increased the test time at each stress in the API test from 15 minutes to 50 hours. This modification was later
formalized in API RP19D and ISO 13503-5. However, work by several additional authors (Cobb (1986), Hahn (1986),
Montgomery (1985)) indicated that while the bulk of the conductivity loss may occur in the first 50 hours, long term the
conductivity continues to drop albeit to a much smaller degree. In addition, the severity of this conductivity loss is dependent
on many factors, including proppant type. A quality ceramic will be least impacted, followed by quality resin coated sand,
with an uncoated sand experiencing the highest conductivity loss due to time (Handren (2007)).

SPE 166107

11

Temperature
Another consideration when choosing proppants centers on the reservoir temperature. The original API conductivity test was
designed to be performed at room temperature. However, another modification proposed by Penny (1987) was to increase
the temperature. The current API/ISO test procedure now calls for a temperature of 150 F for uncoated frac sands, and 250
F for resin coated sands and ceramics. Testing performed by the StimLab Consortium indicated that the conductivity of
sand-based proppants (uncoated and coated) is adversely affected once the temperature exceeds 200 F (ceramic proppants
were found to be unaffected by temperature, likely since they are sintered at >2000 F). It is worth noting that the impact can
be severe. As an example, at Eagle Ford conditions of 275 F and 7000 psi stress, uncoated 20/40 white sand will lose over
50% of its conductivity compared to the baseline 150 F test (Pope 2009). At Haynesville conditions the impact is even
worse.
Proppant Concentration
Proppant concentration refers to the amount of proppant per unit area of fracture wall as measured on one side only. In
customary units it is expressed in pounds of proppant per square foot of one wall of the fracture. If proppant settles to the
bottom of a vertical fracture as it enters the fracture, the width will be highly influenced by the hydraulic/dynamic width at
the time of settling, i.e., during pumping. If the proppant is suspended in the fracturing fluid until the fracture closes,
concentration will be determined by both the width during pumping and the concentration of proppant in the fluid (Palisch
2008). It is well documented that fracture conductivity increases with increasing concentration of proppant in the fracture.
Proppant Strength
It is a truism to state that the strength of proppants is of major concern in the design of propped fractures. Historically,
particle strength has been represented in terms of the load required to crush a single grain of proppant. However proppants
dont act as single grains nor are they used as single grains, but rather proppants are used in multi-layers. Since the strength
of single grains is essentially meaningless as a measure of proppant load bearing capacity in a fracture, a test procedure was
developed to observe and record the behavior of proppants in confined packs. The technique developed is called Crush
Resistance (or K-Value discussed in previous paragraphs) and attempts to measure the amount of fine particles observed after
exposing a proppant pack to a known stress or load. Within the last few years a new set of acceptability standards for use
with crush resistance testing has been published in conjunction with the ISO 13503-2. The data can also be found in API RP19C which replaces the now obsolete API RP-56 procedure.
While useful for comparison, caution must be used when using crush test results alone for selection proppant. There are
many identified nuances with the crush test that can lead one to the wrong proppant when the test results are incorrectly
employed (Palisch 2009). Crush resistance testing finds its most redeemable use as a quality control tool to assure end users
that the quality of the material mined or manufactured today is consistent with past supplies. Sand mines may use them for
qualification purposes, and ceramic manufacturers use the results to ensure consistent quality. The crush test procedures have
been discussed at great length in the literature over the last few years and many new ideas have been proposed to bring these
procedures into agreement with long-term conductivity performance.
Proppant Particle Size
If nowhere else does size matter, it is in the business of selecting proppant for the long-term productivity of the well at hand
(Syfan & Anderson 2011). The adage bigger is better is debatably true and is probably more likely. From a flow capacity
point of view bigger is better with one caveat, at a sufficiently elevated stress level the long-term conductivity of all sizes
of like proppants begin to look the same. For example, at higher stresses, a 20/40, 30/50 and 40/70 sand will eventually have
similar conductivities, as the proppants are crushed. To achieve the conductivity advantage of larger diameter proppants at
higher stresses, one must typically make the change to a stronger proppant type. While most proppants are characterized by a
mesh distribution i.e. 20/40 or 30/50, there still can be a large performance difference even within the same mesh
distribution. As an example, a well known 20/40 premium lightweight ceramic is actually a two-screen proppant, with
over 90% of the particles actually falling between the 20 and 30 Mesh screens, where as a 20/40 standard lightweight ceramic
proppant may have 90% of the particles between the 20 and 35 Mesh screens. The conductivity difference between these two
sized proppants is substantial. Furthermore, most premium resin coated sand utilizes a coarse distribution (particles
skewed to the large side of the distribution) to ensure premium conductivity.
However, with the tremendous shortage of quality white sand, premium RCS and premium ceramics, many other lower
quality products have filled the void, and in many cases, these products are skewed to the finer end of the distribution to
increase yield and drive down costs. However, this comes at a cost to performance, as just moving from a coarse to a fine
distribution can yield a noticeably large drop in conductivity.

12

SPE 166107

Proppant Grain Shape


Of the various observed properties of proppants, the significance of grain shape is arguably the least appreciated. If we
examine a typical grain of high quality, sintered ceramic proppant, the observed result is a relatively smooth, nearly spherical,
and very round particle. There is at least one property of ceramic proppants which should be considered as desirable
regardless of the type of proppant selected. The grain shape of high quality man-made ceramic proppants possess roundness
and sphericity values of 0.9 each, much higher values than are typically found in naturally occurring proppants. A nearperfectly round and spherical particle under load provides the highest sustainable porosity and therefore permeability and
conductivity possible with solid particles.
Minimum Krumbein factors as established in the ISO 13503-2 guidelines are greater than or equal to 0.6. It is noteworthy
that proppants with Krumbein factors less than 0.6, may actually exhibit higher conductivity values at very low closure
stresses (under 2,000 psi). Due to the angularity of the particles the proppant grains do not pack as well, thus increasing
porosity and permeability. However, proppant grains fail via point-to-point loading. Therefore, as closure stress increases,
the proppant grains fail quickly and result in fines in the proppant pack causing severe reductions in permeability and
conductivity. It should also be noted that while the above discussion centers on high quality proppants regardless of type,
with the rapid increase in demand over the past several years many lower quality proppants have entered the market. These
proppants may exhibit lower shape values than typically expected, dipping into the 0.7-0.8 for low quality ceramics, and less
than 0.5 for low quality quartz sands.
Formation Ductility/Embedment
The stress or load transmitted to the proppant pack during fracture closure causes shifting or rearrangement of the particle
pack as the excess liquid is squeezed into the porosity of the reservoir, or as fluid is returned to surface during flowback. As
the relative porosity reaches the range of about 30% for spherical or nearly so materials the individual grains begin to
experience mechanical stress or load from the surrounding fracture walls. When the transmitted load exceeds the embedment
strength of the reservoir rock, which is a function of the Brinell Hardness, the grains begin to embed in the face of the rock.
After embedment has gone as far as it can, the proppant grains must bear the full load of the closure stress.
The damage to fracture conductivity is also explained in detail by Penny (1987) in which he states that any embedment will
reduce the effective retained width causing a further reduction in fracture conductivity. He further describes an experiment
using 20/40 Intermediate Density Ceramic which indicated that embedment equal to 1.5 grain widths could account for
conductivity reductions in excess of 17%.
While embedment may not be avoidable in some formations, it can sometimes be reduced using curable resin coated
proppants if the proppant mass is correctly consolidated within the fracture before full drawdown stresses are applied. The
resin coating, by consolidating the proppant pack, increases the area of contact between the proppant grain and the formation
face which reduces the pressure applied to the fracture wall. Embedment can further be compensated for by decreasing the
particle (mesh) size within the fracture resulting in more grains in contact with the fracture face thereby reducing the load of
each grain. However, while smaller diameter proppants will have less embedment, there will also be a loss in permeability,
so it becomes a trade-off.
Cyclic Stress
Research identifying and quantifying the effects of cyclic fatigue failure of proppants is well documented in the industry
literature. Ouabdesselam and Hudson (1991) studied the effects of time-dependent reductions in fracture conductivity and
identified cyclic fatigue failure as a contributing factor to fracture conductivity reduction and impairment. The data from
their study concluded that proppant permeability and fracture width gradually decrease with time as the number of pressure
cycles increases as the proppant pack and formation face break down. Further, the effects of cyclic stress fatigue failure are
increased in cases where the in-situ proppant concentration is decreased. Palisch et al. (2007) demonstrated the effects of
cyclic stress fatigue failure and documented the fact that ceramic and resin-coated proppants exhibited much better resistance
to fatigue failure than northern white a-quartz proppants, as depicted in Fig. 5.
Multi-Phase Flow
Most experienced fracture design engineers are familiar with the effect of multi-phase treatment fluids on the friction
pressure losses encountered when pumping fluids through small internal diameter tubulars. These fluids include nitrogen
with gelled water foams and oil-in-water poly-emulsion fluids. These fluids exhibited higher than average fluid loss control
properties. The same type phenomena occur in propped fractures where two or more immiscible liquid or gas phases are
trying to occupy the same space. As a point of further reference, there are now sufficient correlations in the literature which
enable the calculation of gross effects of multi-phase flow and add it to the very familiar Forchheimer equation as an
additional factor to high rate gas or liquid flow.

SPE 166107

13

Documented work from the StimLab Proppant Consortium has shown that as little as 10% liquid in the produced gas stream
can increase the pressure drop per unit length within the fracture over 30 times those values measured for 100% gas flow.
These effects must be compensated for when correctly selecting a proppant.

Figure 5 Conductivity Impacts on Proppants due to cyclic loading (courtesy StimLab)

Non-Darcy Effects
Henry Darcy first developed his famous Darcy equation (Eq. 10) in the 1800s while observing fluid flow through a sand
column (Frick 1962).

P v

L
k

(10)

Where, P/L = pressure drop per length of proppant pack, = fluid viscosity, v = superficial fluid velocity (as if porosity
were 100%) and k = permeability of porous media. Since the conductivity test flow rates are laminar (2 ml/min), Darcys
Law is successfully applied to calculate the conductivity during this test. However, in 1901 Forchheimer also determined
that if a fluid rate through a porous media were increased, there comes a point where Darcys Law does not fully describe the
pressure drop (Forchheimer 1901). Eq. 11 is Forchheimers Equation:

P v

v 2
L
k

(11)

Where, = coefficient of inertial resistance, v is the fluid velocity, and = fluid density.
While Darcys Law works well in matrix flow in the reservoir, Forchheimers equation is necessary to describe the pressure
drop within hydraulic fractures. This is because in typical hydraulic fracture treatments, the fluid velocity is much higher
than that of the matrix, which, since the velocity (v) term is squared, means the Forchheimer pressure drop can quickly make
the Darcy conductivity relatively unimportant. In fact, it has been reported that non-Darcy flow can be of significance in a
dry gas fracture flowing as little as 50 MCFD (Miskimins 2005).
Critical Fracture Parameters
Fracture Skin

Fracture Face Skin


Fracture face skin is well documented in the literature from work initially performed by Ramey, Cinco, Gringarten, and
Meyer and is represented by Eq. 12:

Y k

Skin face x r 1
X k

f D

(12)

Where, Yx is the depth of the damage along the fracture face, Xf is the effective propped fracture half-length, kr is the
formation permeability, and kD is the permeability of the damage zone.

14

SPE 166107

Choked Fracture Skin


The choked skin effect as identified by Mukherjee and Economides (1991) due to the radial convergence from a vertical
fracture to a horizontal wellbore and can be written as (Meyer et al. (2010)) and presented in Eq. 13:

Sch

h 1 h
ln

x f C fD 2rw 2

(13)

Where, Sch is choked fracture skin, CfD is dimensionless fracture conductivity, h is fracture height, and rw is the wellbore
radius. This equation illustrates the parameters affecting choked skin as a result of flow convergence in a vertical fracture to
the horizontal wellbore.
Fracture conductivity does make a significant impact on the productivity index even for high conductivity fractures. This is
also amplified when one considers the radial convergence from a vertical fracture into a horizontal wellbore. The choked
fracture skin given by Eq. 13 is shown to be inversely proportional to the dimensionless conductivity. Thus even for
relatively large fracture conductivities a choked skin will still be present near the wellbore.
Fracture Half-Length and Width

The understanding of proper fracture half-length and width is well documented in the petroleum literature. However, the
understanding of optimum fracture length and the relationship between fracture fluid production, dimensionless fracture
conductivity, and effective fracture length has not been as well documented (Lolon, Elyezer P. et al. (2003)). Research
performed by Lolon et al. (2003) correctly stated that fracturing fluid that remains in the fracture and formation pore spaces
after cleaning up a fracture treatment can reduce productivity by reducing the relative permeability to gas in the region
occupied by the fracture fluid. They further stated that this invading fluid can damage the formation which results in a
significant reduction in formation permeability at the face of the fracture based on research performed by Tannich (1975) and
Holditch (1979).
Their observations and conclusion were based on fractured well simulations which examined the effects of both single phase
and multiphase modeling. In the cases of short fractures, e.g. Lf = 200 ft, the research showed that gas recovery was
approximately 18% higher for a CfD = 30 than for a CfD = 3, which again is consistent with the topic of this paper. In cases of
much longer fractures, e.g. Lf = 800 ft, it becomes apparent that fracture cleanup, and thus longer effective fractures, is much
more dependent on fracture conductivity. These observations were further substantiated by prior industry research. In their
conclusion, Lolon et al. (2003) stated Greater dimensionless fracture conductivity results in longer effective fracture lengths
and greater cumulative gas production. This is because fracture fluid recovery increases with increasing fracture
conductivity and The fracture cleans up faster with higher gas flow rates. However, the effective fracture length is still
affected more by fracture conductivity than by gas flow rate.
However, it must be pointed out that increasing fracture lengths past the optimum length, may increase gas recovery but in
turn reduce the wells Net Present Value and Rate-of-Return. Holditch et al. (1978) correctly pointed out that as effective
propped length increases, the initial flow rate and ultimate recovery are usually increased. Holditch also pointed out that the
bigger the fracture treatments the better! Summarizing, if the created and effective propped fracture are greater than the
economic optimum effective length, then while the production and ultimate recovery are better, the economic results from the
well can be reduced.
Gidley et al. (1989) outlined a procedure to calculate the optimum effective fracture length, and thus, optimum economic
benefit for a given well. In order to calculate the optimum fracture length and also well spacing for a given well and
reservoir, it is necessary to (1) predict the propped fracture lengths for a reservoir as a function of hydraulic fracturing costs,
(2) use appropriate reservoir simulation to predict the well performance as a function of fracture length, and (3) to perform an
economic assessment of each case considering total drilling and completion costs (which includes fracturing costs for each
scenario), operating expenses, and well performance. Once this is done, graphs comparing NPV to fracture length can be
constructed to determine the optimum length.
Fracture Width
Perkins and Kern (1961) presented classic discussions and calculations dealing with fracture extension pressures and
resulting fracture width for Newtonian fluids in laminar and turbulent flow, represented by Eqs. 14 and 15, respectively.
Since that time, calculation of width has been researched and published by many authors and is well documented in the
petroleum literature.

SPE 166107

15

Q L
W 0.38

1/ 4

(14)

Where, W is the maximum crack width, Q is the total pump rate, is the effective fracture fluid viscosity, L is the length of a
vertical fracture, and E is the Youngs Modulus of the rock.

Q 2 f L
W 0.6

E H

1/ 4

(15)

Where, f is the specific gravity of the fracturing fluid and H is the height of a restricted, vertical fracture.
Fracture Conductivity

The formula for fracture conductivity is extremely well documented in the petroleum literature and has the universal symbol
of w*kf, where w is the width of the created fracture in feet multiplied by the permeability (k) of the proppant pack (or
fracture, f) in millidarcies. Conductivity is arguably the most important with respect to critical fracture parameters. The
permeability-thickness of the formation becomes inconsequential, no matter how high its value, if the conductivity of the
created fracture is so finite as to limit the wells productive rate.
The laboratory measurement of long-term fracture conductivity is fraught with difficulties, inconsistencies and other
seemingly insurmountable issues. Among these is conflicting data from different sources, and the stretching of verified
techniques to provide answers the procedures were never capable of providing. In order to eliminate these difficulties and
inconsistencies, the procedures for measuring long-term fracture conductivity were outlined in detail in ISO 13503-5
guidelines.
Dimensionless Conductivity

Laplacian solution methodology for predicting the flow behavior of multiple transverse finite-conductivity vertical fractures
intercepting horizontal wellbores was utilized for the case studies included in this paper. The solution accounts for the high
initial production from multiple transverse fractures and matches the late time production decline as a result of fracture
interference.
The dimensionless fracture conductivity (CfD) is the main parameter that quantifies the hydraulic fracture effectiveness (i.e.,
the reservoir is linked to the wellbore through a finite conductivity vertical fracture). The dimensionless fracture conductivity
is given by Eq. 16.

C fD

k f wf

(16)

k x f

In the above equation, k f w f is the fracture conductivity (md-ft), k is the equivalent reservoir permeability (md), and xf is
the propped fracture half-length (ft). The dimensionless pseudosteady-state solution for a finite conductivity vertical fracture
in an infinite reservoir from Meyer et al. (2005) is given by Eq. 17 and discussed in the following paragraphs:

1 4
pD ln tDxf
2 e

f C fD

(17)

Where, the pseudo-skin function, f C fD and the effective wellbore radius, rw x f , for a uniform finite-conductivity
fracture in a homogeneous infinite reservoir are given by and presented in the following equations:

r
f C fD ln
2 and w
2
C

x f C fD
fD

The effective wellbore radius for infinite conductivity fracture C fD is rw x f 1 2 . The effective dimensionless
wellbore radius and dimensionless pseudo-skin function as function of the dimensionless fracture conductivity are shown in

16

SPE 166107

Fig. 6 and Fig. 7, respectively. These figures illustrate that rw x f and f C fD are strong functions of the dimensionless
fracture conductivity (i.e. C fD 10 ). That is, as the conductivity increases rw x f increases and f C fD decreases.

The dimensionless pseudosteady-state solution for finite conductivity vertical fracture in a closed reservoir is shown in Eq.
18:
pD 2 t DA 1 J D

(18)

Where,

I x , and rw x f e f
1 J D ln xe e f C fD , f C fD ln

C g

x f

fD

McGuire and Sikora (1960) published an iconic paper on "The Effect of Vertical Fractures on Well Productivity" that
demonstrated the productivity benefits from hydraulic fracturing as a function of fracture penetration ( I x x f xe ) and
relative fracture conductivity. Figs. 8 and 9 present a comparison of work referenced in a paper presented by Syfan and
Anderson (2011) with the published data of various authors including McGuire and Sikora as a function of fracture
penetration and dimensionless conductivity. These figures clearly indicate that as conductivity and penetration increase so
does the dimensionless productivity index. Also for full penetrating fractures ( I x 1 ), the productivity index does not
asymptote to a maximum value until the fracture conductivity approaches approximately one-hundred (i.e. C fD 100 ) for a
drainage area aspect ratio of unity ( 1 ) and approximately one-thousand ( C fD 1000 ) for an aspect ratio of ten ( 10 ).
The role of fracture conductivity as an important design criterion is often overlooked in shale plays. Even for low
permeability formations the dimensionless fracture conductivity is often much greater than the optimum value as initially
proposed by Prats et al. (1961) and later by Economides et al. (2002). The optimum dimensionless conductivity for various
penetrations and fracture drainage aspect ratios is given by C fD
4 1 1 which for a square drainage area ( 1 )
opt

is C fD

opt

2.

Figure 6 - Dimensionless effective wellbore radius for a finiteconductivity vertical fracture - Comparison with the work of
Economides, Barker, Cinco-Ley, and Riley et al.

Figure 7 Pseudo-skin function versus dimensionless


conductivity Comparison with the work of Economides,
Barker, Cinco-Ley, and Riley et al.

Case Histories

The case histories chosen to illustrate the points discussed in this paper are further documented in SPE 131732. Case A is a
Marcellus Shale example showing history matched production and an economic analysis. Case B is an Eagle Ford Shale
example discussing the history matched production and the effects of varying fracture conductivities. Cases C and D
represent the Bakken and Cotton Valley, respectively, where production rates and cumulative production volumes are

SPE 166107

17

calculated from known reservoir values and characterize changes in ultimate recovery and initial production rate from various
dimensionless fracture conductivities. The economic parameters used to calculate economic results are presented in Table 3.

Figure 8 - Dimensionless productivity index as a function of


dimensionless conductivity and penetration for a square
reservoir - Comparison with the work of Holditch, McGuire &
Sikora, and Valko et al.

Figure 9 - Dimensionless productivity index as a function of


dimensionless conductivity and penetration for a rectangular
reservoir with an aspect ratio of ten - Comparison with the
work of Valko and Economides (1998).

The values presented in Table 3 are the actual values used in the analysis of the well. The authors freely acknowledge that
while some of the values presented may have changed in todays economic conditions, but the trends that they represent are
accurate.
Table 3 Economic Parameters
Parameter
Fracture Fluid Unit Cost, $/U.S. Gal.
Proppant Unit Cost, $/Lb.
Other Well Drill & Complete Cost, $K
Gas Price, $/Mscf
Working Interest, %
Revenue Interest, %

Value
0.30
0.28
2,000
3.75
100.0
80.0

Case A Marcellus Shale

The Marcellus Shale may arguably be the largest shale in the continental United States with an aerial extend of approximately
54,000 sq. miles and extends through Pennsylvania, New York, Ohio, Maryland, and West Virginia. Engelder (2009)
predicted that the Marcellus Shale may ultimately hold 489 Tcf of natural gas reserves.
Case A references a Marcellus Shale completion that was completed with seven (7) hydraulic fracture stages containing five
(5) perforation clusters per stage over a horizontal lateral length of 2,100 ft. The reservoir and hydraulic fracture parameters
for Case A are presented in Table 4. The production data for the well in Case A were history matched using a single-phase,
analytical reservoir simulator for multiple transverse finite-conductivity fractures including a choked fracture skin. The
history matched parameters are given in Table 5.
While the history matched production only covers a period of 201 days, the predicted values for these parameters seems to be
reasonable with respect to the fracture treatment, fracture fluid efficiency, and expected reservoir permeability for this type of
reservoir. The reservoir permeability predicted from the history match is 583 nD (0.583 microdarcies) which yields a
reservoir capacity (permeability-thickness) equal to 0.094 md-ft. The propped effective fracture length was calculated to be
320 ft with a fracture conductivity of 3.77 md-ft. These values are also believed to be reasonable based on the fracture
treatment pumped. The choked fracture skin was calculated to be +0.096 which is not unreasonable considering that these
fracture treatments are over-flushed and are subjected to further over-flushing from composite plug pump downs.

18

SPE 166107

The dimensionless fracture conductivity (CfD) calculated from the history match is equal to 20.2, which represents a medium
finite-conductivity fracture. To illustrate the points made earlier in this paper, the CfD was plotted using values of 2, 20, and
200. The production rate and cumulative gas production for Case A are presented in Figs. 10 and 11.
Table 4 Case A: Marcellus Reservoir and Fracture Parameters
Description
Reservoir Depth, ft
Reservoir Thickness, ft
Hydrocarbon Porosity, %
Reservoir Pore Pressure, psi
Reservoir Temperature, oF
Drainage Area, acres
Aspect Ratio (xe/ye), dimensionless
BHFP, psi
Wellbore Radius, ft
Lateral Length, ft
Number of Stages
Number of Clusters/Stage

Value
7,876
162
4.2
4,726
175
80

1,450 530
0.3646
2,100
7
5

Table 5 Case A: Fracture and Reservoir History Matched Parameters


Description
Reservoir Permeability, nD
Reservoir Capacity (kh), md-ft
Effective Propped Length, ft
Fracture Conductivity, md-ft
Dimensionless Conductivity
Choked Skin, dimensionless
Number of Equivalent Fracs
Diffusivity, md-psi/cp

Figure 10 Case A: Predicted Gas Production Rate

Value
583.0
0.94
320
3.77
20.2
+0.096
6
4000

Figure 11 Case B: Predicted Cumulative Gas Production

The figures point out that a lower achieved value for dimensionless fracture conductivity equal to 2 or a much higher value
equal to 200 yield significant variances in gas rate and cumulative production. Initial production rates varied from just over
3.0 MMscf/day for the high conductivity case to 330 Mscf/day for the low conductivity case which yields 10-year cumulative
production equal to 2.9 Bscf and 0.9 Bscf, respectively, for the high and low conductivity cases.
The economic evaluation of Case A is based on an analysis assuming that the proppant cost is a linear function of fracture
conductivity ( k f w f ). Although this may be a very simplistic assumption it illustrates that increased proppant conductivity
does increase the total job cost. The proppant cost was taken from the actual job cost based on a history matched fracture

SPE 166107

19

conductivity of k f w f 3.47 mD-ft (i.e. C fD 20 ) (Jacot et al. (2010)). The economic analyses are based on the Net
Present Value (NPV = DWR CF) and Discounted Return on Investment (DROI = DWR/CF) where DWR is the Discounted
Well Rate and CF is the fracture treatment cost.
The fracture Net Present Value (NPV) and Discounted Return on Investment (DROI) as a function of fracture conductivity
and producing time are shown in Fig. 12 and Fig. 13, respectively. The NPV and DROI are shown to increase with time but
reach maximum values as a function of fracture conductivity. At five years, the optimum conductivity to maximize the NPV
($5MM) and DROI (1.5) are about k f w f 5 mD-ft ( C fD 29 ) and k f w f 4 mD-ft ( C fD 23 ), respectively. Figs. 12 and
13 also illustrate that for an economic period of five years, the NPV and DROI for a conductivity of 0.347 ( C fD 2 ) are
negative and less than one, respectively. The NPV and DROI for the same period but at a high conductivity of 8.68 md-ft (
C fD 50 ) are about $2MM and 1.3, respectively. Fig. 14 and Fig. 15 also illustrate for full penetration fractures designed
with the traditional rule of thumb dimensionless fracture conductivity of about two (i.e. C fD 2 ) may not be economical.

Figure 12 Case A: Fracture NPV vs. Conductivity

Figure 13 Case A: Fracture DROI vs. Conductivity

Case B Eagle Ford Shale Example

The Eagle Ford formation is a widespread Upper Cretaceous deposit in the Gulf Coast region of South-Central Texas long
known as a source rock for other plays. Only recently has it been recognized as an unconventional oil and gas play. In South
Texas, where it has hydrocarbon potential, the Eagle Ford formation is between 5,000' and 13,000' below the surface and
ranges in thickness from 50' to 300'. Although the Eagle Ford is described as a shale formation, the reservoir is actually
composed of organic-rich calcareous mudstones and chalks that were deposited during two transgressive sequences (the
Upper and Lower Eagle Ford). Total organic carbon (TOC) in the Eagle Ford formation ranges from 1-7%. The organic
richness in the Eagle Ford typically increases with depth. As such, the Lower Eagle Ford tends to be more organically rich
and produces more hydrocarbons.
The calcareous makeup of this rock makes this play significantly different than other well-known unconventional plays such
as the Barnett Shale, Haynesville Shale, and Marcellus Shale (all of which are found in primarily siliceous environments).
Case B references an Eagle Ford Shale completion that was completed with ten (10) hydraulic fracture stages containing four
(4) perforation clusters per stage over a horizontal lateral length of 4,000 ft. The reservoir and hydraulic fracture parameters
for Case B are presented in Table 6. As with the previous case, the production data for the well in Case B were history
matched using a single-phase, analytical reservoir simulator for multiple transverse finite-conductivity fractures including a
choked fracture skin. The history matched parameters are given in Table 7.
The Case B well was designed to yield high dimensionless fracture conductivity, as shown in Table 7. The Eagle Ford well
was completed with a ten stage stimulation using slickwater, linear gel, and ceramic proppant in a 4,000 foot lateral.
Composite flow through bridge plugs was used for isolation and four two-foot clusters were perforated per stage (40 clusters
total). The first and last intervals were shot at six (6) shots per foot (spf) and the middle two intervals were shot at 12 spf.
The average treating rate and pressure was about 50 bpm and 8,900 psi, respectively. The proppant concentration for each
stage increased from 0.25 to 1.5 lbm/gallon without any slickwater sweeps. Linear gel was used to place 0.75 to 1.5 pound

20

SPE 166107

concentration of proppant. Average proppant placed per stage was ~250,000 lbm and 11,300 gallons of water. Production
logs and chemical traces in the flowback samples indicated that all clusters have somewhat variable production rates;
however, the total production from each fracture treatment stage is producing approximately equal volumes. As such, the
production data history matched used forty (40) multiple transverse fractures (ten stages with four clusters per stage).
Table 6 Case B: Eagle Ford Shale Reservoir and Fracture Parameters
Description
Reservoir Depth, ft
Reservoir Thickness, ft
Hydrocarbon Porosity, %
Reservoir Pore Pressure, psi
Reservoir Temperature, oF
Drainage Area, acres
Aspect Ratio (xe/ye), dimensionless
BHFP, psi
Lateral Length, ft
Number of Stages
Number of Clusters/Stage

Value
10,875
283
5.76
8,350
285
80

3,900 1,500
4,000
10
4

Table 7 Case B: Fracture and Reservoir History Matched Parameters


Description
Reservoir Permeability, nD
Reservoir Capacity (kh), md-ft
Effective Propped Length, ft
Fracture Conductivity, md-ft
Dimensionless Conductivity
Choked Skin, dimensionless
Number of Equivalent Fracs
Diffusivity, md-psi/cp

Value
17.2
0.0049
131
0.86
382
+0.0254
40
153

While the history matched production only covers a period of approximately 93 days, the predicted values for these
parameters seems to be reasonable with respect to the fracture treatment, fracture fluid efficiency, and expected reservoir
permeability for this type of reservoir. The reservoir permeability predicted from the history match is 17.2 nD which yields a
reservoir capacity (permeability-thickness) equal to 0.0049 md-ft. The propped effective fracture length was calculated to be
131 ft with a fracture conductivity of 0.86 md-ft. These values are also believed to be reasonable based on the fracture
treatment pumped. The choked fracture skin was calculated to be +0.0254 which is considered reasonable based on flush
volumes.
The dimensionless fracture conductivity (CfD) calculated from the history match is equal to 382, which represents a high
conductivity fracture. The CfD was plotted using values of 4, 40, and 400. The production rate and cumulative gas
production for Case B are presented in Figs. 14 and 15.
The fracture stimulation treatments pumped on the Eagle Ford Shale in Case B were designed using accurate reservoir and
core data and achieved a high value of dimensionless fracture conductivity. Figures 14 and 15 illustrate this fact and indicate
that a much higher achieved value for dimensionless fracture conductivity equal to 400 yields significant increases in gas rate
and cumulative production. Initial production rates varied from just over 5.0 MMscf/day for the properly designed high
conductivity case. This is apparently due to the 250,000 lbs. of ceramic proppant placed per stage, which equates to 2.5
MMlbs for the 10 stages pumped. Lower proppant concentrations or lower conductivity proppants pumped into the well
would result in much lower fracture conductivities and initial production rates. Fig. 14 shows that much lower values of
dimensionless conductivity equal to 40 and 4 result in initial production rates of approximately 940 Mscf/day and 140
Mscf/day, greatly reducing ultimate recovery and significantly reduce the NPV and IRR of the well.
Case C Bakken Example

The Bakken is a rock formation from the Late Devonian Early Mississippian age. The play covers much of the Williston
Basin in parts of Montana and North Dakota and extends into the Canadian provinces of Saskatchewan and Manitoba. The
Bakken formation consists of three layers: an upper black shale layer, middle silty-dolomite, and a lower black layer of shale.
The shale layers are petroleum source rocks as well as seals for oil reservoirs in the middle dolomite layer and in the

SPE 166107

21

underlying Three Forks formation. The Three Forks Formation is a dolomite/mudstone that underlies the Bakken Shale. The
formation is sourced by the Bakken and produces across parts of North Dakota and Montana. The Three Forks play is a
secondary target in some areas and has proven even more prolific than the Bakken in some areas near the Nesson Anticline.

Figure 14 Case B: Predicted Gas Production Rate

Figure 15 Case B: Predicted Cumulative Gas Production

The reservoir and fracture properties used to simulate varying fracture conductivities in the Bakken formation in North
Dakota were obtained from Lolon et al. (2009) in which they performed a 3-D numerical simulation of multiple transverse
fractures intersecting a horizontal wellbore. The properties used in the Bakken evaluation are shown in Table 8. The initial
oil production rates and forecast for Case C were calculated using a single-phase, analytical reservoir simulator for multiple
transverse finite-conductivity fractures. A choked fracture skin was not included in the calculations but was calculated to be
virtually negligible (+0.00034).
Table 8 Case C: Bakken Shale Reservoir and Fracture Parameters
Description
Reservoir Depth, ft
Reservoir Thickness, ft
Reservoir Permeability, D
Porosity, %
Reservoir Pore Pressure, psi
Reservoir Temperature, oF
Reservoir Compressibility, 1/psi
Reservoir Viscosity, cp

Value
9,881
46
0.002
5.0
4,900
209
2.0 E-05
0.30

Description
Drainage Area, acres
BHFP, psi
Effective Fracture Length, ft
Fracture Conductivity, md-ft
Dimensionless Conductivity
Lateral Length, ft
Number of Transverse Fractures

Value
640
1,500
420
200
238
5,000
12

Figs. 16 and 17 illustrate the fact that increasing values for dimensionless fracture conductivity yield higher initial production
rates and ultimate recoveries. This is especially true as dimensionless fracture conductivity increases from 2.38 to 23.8.
While the Bakken is a different type of reservoir from the previous two cases in the Marcellus and Eagle Ford formations, the
laws governing reservoir performance referenced earlier on the work presented by McGuire and Sikora (1960) hold true.
Case D Cotton Valley Example

The Cotton Valley (CV) subsurface formation is a tight gas play in Northeast Texas and Northwest Louisiana located just
above the Haynesville/Bossier Shale. In East Texas much of the activity is located in Rusk, Panola, and Harrison counties. It
is Upper Jurassic and Lower Cretaceous in origin and consists of sandstone, limestone, and shale. The depth of the Cotton
Valley formation is roughly 7,800 to 10,000 feet.
The reservoir and fracture properties used to simulate varying fracture conductivites in the Cotton Valley formation in East
Texas were obtained from Lolon and Mayerhofer (2007) in which they performed a 3-D numerical simulation of multiple
transverse fractures intersecting a horizontal wellbore. The properties used in the Cotton Valley evaluation are shown in
Table 9. The initial gas production rates and forecast for Case D were calculated using a single-phase, analytical reservoir
simulator for multiple transverse finite-conductivity fractures.

Figure 16 Case C: Bakken Oil Production Rate

Figure 17 Case C: Bakken Cumulative Oil Production

Table 9 Case D: Cotton Valley Reservoir and Fracture Parameters


Description
Reservoir Depth, ft
Reservoir Thickness, ft
Reservoir Permeability, D
Porosity, %
Reservoir Pore Pressure, psi
Reservoir Temperature, oF
Drainage Area, acres
BHFP, psi
Effective Fracture Length, ft
Fracture Conductivity, md-ft
Dimensionless Conductivity
Lateral Length, ft
Number of Transverse Fractures

Value
9,000
100
0.001
7.0
6.000
285
230
1,500
1,500
114
76
2,000
7

Figs. 18 and 19 illustrate the fact that increasing values for dimensionless fracture conductivity yield higher initial production
rates and ultimate recoveries in sandstone formations. As with the previous three (3) cases, initial production rate and
ultimate gas recovery increases as dimensionless fracture conductivity increases from 0.76 to 76. A slight increase in rates
and recoveries is also illustrated as dimensionless conductivity increases from 76 to 760. Again, the Cotton Valley formation
shows significant increases in productivity and ultimate recovery again supporting the work presented by McGuire and
Sikora (1960).

Figure 18 Case D: CV Gas Production Rate

Figure 19 Case D: CV Cumulative Gas Production

SPE 166107

23

Summary

It is apparent from the provided technical information and the case histories that proper fracture design and ultimately,
fracture optimization, cannot and will not happen without sound engineering practices and an understanding of the many
technical factors involved in optimum fracture design. Furthermore, without optimum fracture design, it is a truism that
initial production rates, ultimate recovery, NPV, and rate-of-return will be compromised. The bottom line is that unless
current practices and procedures are changed, highly conductive fractures will not be achieved in a high percentage of the
stimulations pumped!
The importance of this paper centers around this key point: At the end of the day, the horsepower on location is going to rig
down, leave location and head to the next job. The fluid that was injected during the stimulation will be broken, flowed back,
removed from location and discarded. The only remaining item is the proppant pack which, depending upon design, may or
may not provide conductivity high enough to maximize initial rate and ultimate recovery. However, if the proppant pack is
properly designed, it can and will last the life of the well.
Conclusions

1. Understanding the rock mechanics associated with the primary zone of interest, in addition to the zones above and below
the primary zone, is essential to achieving high conductivity fractures.
2. The research presented by McGuire and Sikora (1960) holds true regardless of reservoir type and ultimately dictates
reservoir and production performance.
3. Fracture conductivity and dimensionless fracture conductivity ultimately govern the initial production rates and ultimate
recoveries regardless of the type of reservoir lithology.
4. The history matches of the production data from Case A (Marcellus Shale) and Case B (Eagle Ford Shale) illustrate the
importance of achieving high conductivity transverse fractures in a horizontal wellbores.
5. Increasing fracture conductivity, regardless of reservoir type, results in a significant positive impact on IRR and NPV,
Acknowledgements

The authors wish to thank the management of GoFrac LLC, CARBO Ceramics Inc., and Independence Oilfield Chemicals
LLC for the opportunity to perform this research and publish the findings. The authors would also like to thank Bruce R.
Meyer for his help and discussions on the theory of fracture propagation and production simulation analysis. The authors
would also like to acknowledge Gerald R. Coulter and James E. Briscoe for their contributions in reviewing and editing this
paper.
Nomenclature

axial
BHFP
F
CfD
o
F
m
w
md
D
nD

=
=
=
=
=
=
=
=
=
=
=
=

Axial Strain
Bottonhole Flowing Pressure
Fahrenheit
Dimensionless Fracture Conductivity
Degrees Fahrenheit
Elastic Modulus
Fracture Width
Milli-Darcy
Micro-Darcy
Nano-Darcy
Poissons Ratio
Reservoir Permeability

kh

Sch

t
T
t
t

=
=
=
=
=
=
=
=

Permeability-Thickness
Strain
Choked Fracture Skin
Stress
Tensile Stress
Tectonic Stress
Tensile Strain
Time

24

trans
E
ml
Mscf
MMscf
Tcf

SPE 166107

=
=
=
=
=
=

Transverse Strain
Youngs Modulus
Milliliter
Thousand Standard Cubic Feet
Million Standard Cubic Feet
Trillion cubic feet

References

Akrad, Ola; Miskimins, Jennifer; Prasad, Mainka: The Effects of Fracturing Fluids on Shale Rock Mechanical Properties
and Proppant Embedment, SPE 146658, Annual Technical Conference and Exhibition, Denver, CO, Oct 30-Nov 2,
2011.
Anderson, R. W.; Phillips, A. M.: Practical Application of Economic Well Performance Criteria to Optimization of
Fracturing Treatment Design, JPT, February 1988 Issue.
Birch, Francis: Editor, Handbook of Physical Constants, GSA Special Paper No. 36, (Jan. 31, 1942).
Clark, J. B.: A Hydraulic Process for Increasing the Productivity of Wells, JPT, January 1949. Vol. 1, No. 1, Pp. 1-8.
Cleary, J. M.: "Hydraulic Fracture Theory, Parts I, II, & III", Circulars 251, 252, 281, Illinois Geological Survey, Urbana, IL,
(1959).
Cobb, S. L., et al: Evaluation of Long-Term Proppant Stability, SPE 14133 presented at the SPE International Meeting on
Petroleum Engineering, March 1986.
Coulter, G. R.; Wells, R. D.: The Advantages of High Proppant Concentration in Fracture Stimulation, JPT, June 1972,
Volume 24, Number 6, Pp. 643 650.
Daines, Stephen R.: Prediction of Fracture Pressures for Wildcat Wells, JPT, Vol. 34, No. 4, April 1982, Pp. 863 872.
Eaton, Ben A.: Fracture Gradient Prediction and Its Applications in Oilfield Operations, JPT, Vol. 21, No. 10, October
1969, Pp. 1353 1360.
Economides, M., Oligney, R., and Valko, P.: Unified Fracture Design, Orsa Press, Alvin, Texas, 2002.
Engelder, T.: 2008: Report Card on the Breakout Year for Gas Production in the Appalachian Basin, Fort Worth Basin Oil
& Gas Magazine, August 2009, pp. 19-22.
Frick, T. C., et al: Petroleum Production Handbook, Millet the Printer, Dallas (1962) Vol 2, p 23-11.
Forchheimer, P.: Wasserbewegung durch Boden, ZVDI (1901) 45, 1781.
Geertsma, J.; de Klerk, F.: A Rapid Method of Predicting Width and Extent of Hydraulically Induced Fractures, JPT,
December 1968, pp. 1571 1581.
Gidley, John F.; Holditch, Stephen A.; Nierode, Dale E.; Veatch Jr., Ralph W.: Recent Advances in Hydraulic Fracturing,
SPE Monograph Series, Volume 12, 1989.
Hagel, M. W.; Meyer, B. R.: Utilizing Mini-frac Data to Improve Design and Production, CIM paper 92-40 June, 1992.
Hahn, G.: How Long Will it Prop?, Drillinng the Wellsite Publication, Vol. 47, No. 6, Issue No. 596, Energy Publications,
Dallas, TX, April 1986.
Handren, P., Palisch, T. T.: Successful Hybrid Slickwater Fracture Design Evolution An East Texas Cotton Valley Taylor
Case History, SPE 110451, presented at the SPE ATCE, Anaheim, CA, November 2007.
Holditch, S. A.: Factors Affecting Water Blocking and Gas Flow From Hydraulically Fractured Gas Well, JPT, December
1979, Pp. 1515 1524.
Holditch, Stephen A.; Jennings, James W.; Neuse, Stephen H.: The Optimization of Well Spacing and Fracture Length in
Low Permeability Reservoirs, SPE 7496, Presented at the Annual Technical Conference and Exhibition, Houston, TX,
October 1978.
Hubbert, M. King; Willis, David G.: Mechanics of Hydraulic Fracturing, Trans., AIME, Revised Manuscript February
1957, Vol. 210, Pp. 153 168.
Jacot, R. H.; Bazan, L. W.; Meyer, B. R.: Technology Integration A Methodology to Enhance Production and Maximize
Economics in Horizontal Marcellus Shale Wells, SPE 135262, September 2010.
King, George E: Hydraulic Fracturing 101: What Every Representative, Environmentalist, Regulator, Reporter, Investor,
University Researcher, Neighbor and Engineer Should Know About Estimating Frac Risk and Improving Frac
Performance in Unconventional Gas and Oil Wells, SPE 152596, Presented at the Hydraulic Fracturing Technology
Conference, February 2012.

SPE 166107

25

Lehman, L. V.; Shelley, R. F.; Crumrine, T.; Gusdorf, M.; Tiffin, J.: Conductivity Maintenance: Long Term Results from
the Use of Conductivity Enhancement Material, SPE 82241, 2003.
Lolon, E. P.; Cipolla, C. L.; Weijers, Hesketh, R. E.; Grigg, M. W.: Evaluating Horizontal Well Placement and Hydraulic
Fracture Spacing/Conductivity in the Bakken Formation, North Dakota, SPE 124905, October 2009.
Lolon, E.P.; Mayerhofer, M.J.: Application of 3-D Reservoir Simulator for Hydraulically Fractured Wells, SPE 110093,
November 2007.
Lolon, Elyezer P.; McVay, Duane A.; Schubarth, Stephen K.: Effect of Fracture Conductivity on Effective Fracture
Length, SPE 84311, Presented at the Annual Technical Conference and Exhibition, Denver, CO, October 2003.
McGuire, W. J.; Sikora, V. J.: The Effect of Vertical Fractures on Well Productivity, SPEJ, Vol. 219 401-403, 1960.
Meyer, B. R.: Design Formulae for 2-D and 3-D Vertical Hydraulic Fractures: Model Comparison and Parametric Studies,
SPE 15240 presented at the SPE Unconventional Gas Technology Symposium, Louisville, KY, May 18-21, 1986.
Meyer, B. R.: Three-Dimensional Hydraulic Fracturing Simulation on Personal Computers: Theory and Comparison
Studies, SPE 19329 presented at the SPE Eastern Regional Meeting, Morgantown, Oct. 24-27, 1989.
Meyer, B. R.; Jacot, R. H.: Pseudosteady-State Analysis of Finite Conductivity Vertical Fractures, SPE 95941, October
2005.
Meyer, B. R.; Jacot, R. H.: Impact of Stress-Dependent Youngs Moduli on Hydraulic Fracture Modeling, Meyer &
Associates, Inc., Natrona Heights, PA, 38th U.S. Rock Mechanics Symposium, Washington D.C., July 710, 2000.
Meyer, B. R.; Bazan, L. W.; Jacot, R. H.; Lattibeaudiere, M. G.: Optimization of Multiple Transverse Hydraulic Fractures in
Horizontal Wellbores, SPE 131732, February 2010.
Meyer, B. R.; Bazan, L. W.: A Discrete Fracture Network Model for Hydraulically Induced Fractures: Theory, Parametric
and Case Studies, SPE 140514, Presented at the SPE Hydraulic Fracturing Technology Conference and Exhibition, The
Woodlands, TX, January 24-26, 2011.
Miskimins, J. L.; Lopez-Hernandez, H. D.; Barree, R. D.: Non-Darcy Flow in Hydraulic Fractures: Does it Really Matter?,
SPE 96389, Presented at the SPE Annual Technical Conference and Exhibition, Dallas, TX, October 9-12, 2005.
Montgomery, C. T., et al: Proppant Selection: The Key to Successful Fracture Stimulation, SPE 12616, Journal of
Petroleum Technology, Vol. 37, No. 12, pp 2163-2172, December 1985.
Mukherjee, H.; Economides, M. J.: A Parametric Comparison of Horizontal and Vertical Well Performance, SPE
Formation Evaluation, June 1991.
Nolte, K. G.: Fracture Design Considerations Based on Pressure Analysis, SPEPE, February 1988, Pp. 22 30.
Ouabdesselam, M., Hudson, P.J.: An Investigation of the Effect of Cyclic Loading on Fracture Conductivity, SPE 22850,
presented at the SPE ATCE, Dallas, TX, October 1991.
Palisch, T. T., Duenckel, R. J, Bazan, L., Heidt, H. J.: Determining Realistic Fracture Conductivity and Understanding Its
Impact on Well Performance Theory and Field Examples, SPE 106301, presented at the SPE Hydraulic Fracturing
Technology Conference, January 2007.
Palisch, T. T., Vincent, M. C., Handren, P. J.: Slickwater Fracturing Food for Thought, SPE 115766, presented at the
SPE ATCE, September 2008.
Palisch, T. T., Duenckel, R. D, Chapman, M. A., Woolfolk, S., Vincent, M. C.: How to Use and Misuse Proppant Crush
Tests Exposing the Top 10 Myths, SPE 119242, presented at the SPE Hydraulic Fracturing Technology Conference,
January 2009.
Palisch, T. T., Chapman, M. A., Godwin, J.: Hydraulic Fracture Optimization in Unconventional Reservoirs A Case
History, SPE 160206, presented at the SPE ATCE, October 2012.
Penny, G. S.: An evaluation of the Effects of Environmental Conditions and Fracturing Fluids Upon the Long-Term
Conductivity of Proppants, SPE 16900, presented at the SPE ATCE, September 1987.
Perkins, T. K.; Kern, L. R.: Widths of Hydraulic Fractures, SPE 89, Presented at the 36th Annual Fall Meeting of SPE,
Oct. 8-11, 1961, Dallas, TX.
Peterson, R. E. et al: Planning, Implementation, Administration, and Documentation of SFE No. 4 and Related GRI Field
Research Projects, GRI Final Report GRI-91-0173, 1991.
Pope, C., Peters, B., Benton, T., Palisch, T. T.: Haynesville Shale One Operators Approach to Well Completions in this
Evolving Play, SPE 125079, presented at the SPE ATCE in New Orleans, LA, September 2009.
Prats, M.: Effect of Vertical Fractures on Reservoir Behavior - Incompressible Fluid Case, SPEJ, June 1961, pp. 105-118.
Syfan, Frank E.; Anderson, R. W.: Lower Quality Natural Quartz Proppants Result In Significant Conductivity Loss and
Reduction In Ultimate Recovery: A Case History, SPE 146827, November 2011.
Tannich, J. D.: Liquid Removal From Hydraulically Fractured Gas Wells, JPT, November 1975, Pp. 1309 -1317.

26

SPE 166107

Valko, P. P.; Economides, M. J.: Heavy Crude Production from Shallow Formations: Long Horizontal Wells Versus
Horizontal Fractures, SPE 50421, November 1998.
Vejbk, Ole V. et al.: Influence of Tensile Strength and Production Effects on Hydraulic Fracturing in Low Porosity
Carbonates: The South Arne Case, SPE 165117, prepared for presentation at the SPE European Formation Damage
Conference and Exhibition, Noordwijk, The Netherlands, June 5 7, 2013.
Wuerker, R. G.: "Annotated Table of Strength and Elastic Properties of Rocks", Paper 663-G published as separate
publication by SPE, 1956.

Potrebbero piacerti anche