SPE-184857-MS

Impairment of Fracture Conductivity in the Eagle Ford Shale Formation

Jesse Guerra, D. Zhu, and A. D. Hill, Texas A&M University

Copyright 2017, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE Hydraulic Fracturing Technology Conference and Exhibition held in The Woodlands, Texas, USA, 24-26 January
2017.

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
Fracture conductivity in shale formations can be greatly reduced due to water-rock interactions. The
mechanisms of water damage to fracture conductivity include clay swelling, surface softening, excessive
proppant embedment, and fines migration due to fracture surface spalling and failed proppant particles.
Fracture conductivity is influenced by closure stress, bulk and surface rock mechanical properties,
fracture surface topography, fracture surface elemental composition, rock mineralogy, proppant type and
concentration, among other factors. This paper presents a study considering several of the aforementioned
factors, centered primarily on saline water induced fracture conductivity impairment of the Eagle Ford shale
formation and its five vertical lithostratigraphic units.
Laboratory experiments were conducted to investigate the effect of flowback water on fracture
conductivity for Eagle Ford shale samples. The majority of test samples were obtained from an outcrop
located in Antonio Creek, Terrell County, Texas; while the remaining samples were obtained from downhole
core provided by an industry partner. Saline water with a similar chemical composition to the typical field
flowback water was utilized.
Fracture conductivity measurements were conducted in three stages. In the first stage, dry nitrogen was
flowed to ascertain the undamaged initial fracture conductivity. In the second stage, the saline solution
was injected into the fracture until steady state behavior was observed. In the third and final stage, dry
nitrogen was once again flowed to quantify the recovered fracture conductivity. Reported mechanical
properties from the same outcrop rock samples, consisting of Poisson's ratio and the Brinell hardness
number, were considered in this study. Additionally, reported mineralogy obtained using X-ray powder
diffraction (XRD) microscopy was taken into consideration. The elemental composition along the fracture
surface was obtained using X-ray fluorescence (XRF) microscopy, and fracture surface topography was
obtained using a laser surface scanner and profilometer.
Results support findings that bulk and surface mechanical properties influence fracture conductivity,
as well as surface topography and related attributes such as fracture surface area. Furthermore, the bulk
mineralogical composition of the rock and the elemental composition of the rock fracture surface have a
significant impact on fracture conductivity when flowing saline water in order to simulate flowback. Clay
content was observed to directly influence fracture conductivity. The results of this study show a loss of
fracture conductivity for Eagle Ford formation ranging from approximately 4 to 25 % after flowing saline
2 SPE-184857-MS

water, when compared to the initial conductivity measured by flowing dry nitrogen before saline water
exposure. This is not a large loss in conductivity due to water damage, and suggests that water damage may
not be the major cause of the large early decline rates observed in most Eagle Ford shale producing wells.

Introduction
Influencing factors on fracture conductivity damage were previously investigated reflecting realistic
downhole conditions, including the effect of fracturing fluids (Cooke 1975), measurements under in-situ
conditions (Parker et al. 1987), slickwater fracturing (Palisch et al. 2007), sensitivity of fluids in shales
(Ramurthy et al. 2011), and the effect of fracturing fluids on rock mechanical properties (Akrad et al. 2011).
Water-induced fracture conductivity loss in shales was previously investigated with respect to proppant
embedment and bulk mineralogical composition (i.e., spot analysis via XRD). Zhang et al. (2014) showed
the effect of excessive proppant embedment due to fracture surface softening after water exposure, while
Zhang et al. (2015) showed that clay content determined the fracture conductivity damage by water. The
study conducted by Zhang et al. (2015) showed evidence of shale fines migration, while crushed proppant
particle migration was not observed. In the previously mentioned studies, Barnet shale samples were
obtained from an outcrop located in San Saba, Texas, while the Eagle Ford shale samples were limited to
those gathered from roadcuts northwest of Del Rio, Texas, notably only representing unit B (Fig. 1).

Figure 1Lozier Canyon (west portion) outcrop face and its respective lithostratigraphic units.

Eagle Ford shale rock mineralogy and mechanical properties were shown to have a direct effect on
fracture conductivity. According to Enriquez-Tenorio et al. (2016), there is a direct relationship between
rock brittleness based on mineralogy and fracture conductivity, while there is an inverse relationship with the
Poisson's ratio. The formation mineralogical composition inherently has a role in determining the tendency
it has to interact with water. The interactions between water and rock can be chemical, physical, or a
combination of both (Ali, et al. 2015).
From a geochemical perspective, fracture conductivity decline due to water-rock interactions depends
broadly on four factors, namely, the chemical makeup of the fracturing fluid; the type, concentration,
and distribution of minerals along the fracture surface; the chemical makeup of the formation water
originally present; and the resulting fluid-rock interactions under downhole conditions during shut-in time
SPE-184857-MS 3

and subsequent production. These fracturing fluids when in contact with the formation rock, have the
potential dissolve or transform minerals, and even create precipitates as time progresses, which in turn have
an effect on reservoir characteristics including mechanical properties, flow pathways, and produced fluids
(Ali et al. (2015), Zhang et al. (2016)). When the well is initially put on production, these aforementioned
fluid-rock interactions dictate the resulting chemical composition of the flowback water as well as the
resulting well performance through the attained fracture conductivity.
This study aims at quantifying under laboratory conditions the degree of fracture conductivity impairment
on the Eagle Ford shale when exposed to saline water with a similar chemical composition as a typical field
flowback water. Additionally, this study represents an effort to investigate the relationship between fracture
conductivity impairment by flowback water and the fracture surface elemental characterization (via XRF)
for each lithostratigraphic unit of the Eagle Ford shale.

Experimental Description
Shale Sample Acquisition
In this study, two categories of Eagle Ford shale samples were evaluated, namely samples obtained from
an outcrop rock and samples obtained from downhole core. The outcrop rock was obtained from Lozier
Canyon and Antonio Creek, Terrell County, Texas, a site currently leased by BP (Fig. 1). This Eagle Ford
outcrop in West Texas is characterized by a heterogeneity discretized in a vertical facies succession as
described by Donovan et al. (2010, 2012). The nomenclature followed in this study is an adaptation of
Gardner et al. (2013), where five different lithostratigraphic units can be considered equivalent to the Eagle
Ford subsurface in South Texas. Outcrop rocks were gathered for each unit. The downhole core was provided
by an industry partner, representing the downhole equivalent of unit B at a depth exceeding 10,000 ft.
Typical dimensions for the gathered outcrop rocks were at a minimum 8in. in each direction, in order to
satisfy test sample dimension requirements, whereas the downhole core segments were prescribed to a 4 in.
diameter with a 1/3rd diameter slabbed (Fig. 2).

Figure 2Antonio Creek outcrop rock sample pertaining to unit B (left) and a
downhole core segment pertaining to the subsurface equivalent of unit B (right).

Test Sample Fracturing and Dimensioning

Test sample dimensions were prescribed by a modified API fracture conductivity cell, which builds upon
the standard API fracture conductivity cell defined by API RP-61 (1989), a modified ISO 13503-5:2006(E)
conductivity test. The most notable differences are in sample height due to the induced fracture on the
test sample, as well as the frequent need to utilize sandstone inserts depending on the original size of
the acquired shale sample (Fig. 3). Awaloke (2013) and Kamenov (2013) outlined in detail the baseline
4 SPE-184857-MS

experimental apparatus and procedural departures from the standard API RP-61 (1989). The overall test
sample dimensions are 1.61 in. in width by 6 in. in height by 7.10 in. in length, with rounded ends (Fig.
3b). The shale portion ranges from 2.5 in. up to the entire height, whereas the sandstone inserts serve as a
filler to ensure the overall dimensions (Fig. 3c).

Figure 3(a) Outcrop rock sample with writing reflecting it came from unit A (second letter); (b) Outcrop rock
test sample E-A-R1-S1 after fracturing and dimensioning; (c) Test sample E-A-R1-S1 in an expanded view.

Several methods of inducing a fracture on a fracture conductivity test sample were reported in literature,
such as employing masonry rock splitter blades as done by Fredd et al. (2000). In this study, a new and
improved method was designed and manufactured in conjunction with a third-party contractor, where the
fracture is created under Mode I fracture growth, ensuring that no compressive stress is applied to the
fracture surface during the process.
Test sample preparation consisted of a three-step sequence procedure. The first step consisted of cutting a
2 in. wide by 34 in. tall by 8 in. long rectangular rock sample from the outcrop rock (Fig. 3a) or downhole
core segment. The second step involved scoring along the length and all around the rectangular sample, to
aid in the fracture initiation process. The third step was to insert the rectangular sample inside two steel
sleeves (Fig. 4a), coming together at the score line, then placing the assemblage under a load frame (Fig.
4b) and applying a load until a fracture was induced (Fig. 4c). Once the fracture was created, debris were
gathered and returned to the fracture face, and the rectangular sample was then put back together by marrying
each side of the fracture, at which point, it was saw cut down to test sample dimension requirements as
shown in Fig. 3b.

Figure 4(a) Steel sleeves; (b) Rectangular sample inside steel sleeves under load frame; (c) Fractured rectangular sample.

In order to better preserve sample integrity during the process of inducing a fracture, fracture conductivity
samples obtained from outcrop rocks were fractured along the bedding (referred as Z-orientation, Fig. 5).
Although horizontal fractures are less prevalent in practice, fracture conductivity samples with Z orientation
SPE-184857-MS 5

were shown to be representative while ensuring experimental reproducibility (Zhang (2014)). On the other
hand, fracture conductivity samples obtained from downhole core were fractured vertically in X-direction
(Fig. 5), given that the downhole core was obtained in the vertical section of the well.

Figure 5Schematic of bedding plane orientation with respect to fracture conductivity test samples (core plugs also shown).

Core Plug Test Sample: Triaxial Compression Test and Mineralogy via XRD
Core plugs obtained from the same outcrop rocks as the fracture conductivity samples were created and
utilized to obtain mechanical properties as well as to determine the mineralogy of the samples. Brinell
hardness number (BHN) was measured using a GCTS PLT-100 point load test system, in addition to
measuring the Poisson's ratio employing a GCTS RTX-1500 triaxial compression testing system for the
corresponding samples according to their bedding plane orientation (Fig. 5), as shown in Figure 6 (Knorr,
2016).

Figure 6Average mechanical properties of each unit/sub-unit of the Eagle

Ford shale outcrop (color differentiated for sample orientation per Fig. 5).

After measuring the mechanical properties, the minerology was determined by performing XRD
microscopy on each core plug corresponding to each unit of the outcrop, and the results are shown in Figure
7. Throughout the units, calcite is the most predominant mineral, ranging from 55.67 to 93.71% (Fig. 7).
6 SPE-184857-MS

Sub-unit B3 was considered representative of sub-unit B3-B5, and sub-units of units A, C, and D, and E
were considered uniform. Total organic content (TOC) was obtained from previously published information
(Miceli-Romero, 2014) from the outcrop located in Lozier Canyon and Antonio Creek by utilizing Rock
Eval pyrolosis (Fig. 7). Sub-unit B1 and B2 had the highest average TOC values.

Figure 7Average mineral composition and TOC of each unit and sub-unit of the Eagle Ford shale outcrop.

Three types of clay were present in the core plugs tested; kaolinite, chlorite, and illite. The highest
contents of quartz and kaolinite were in unit B, while unit C had all three clays present.

Fracture Conductivity Test Sample: XRF Scanning on Fracture Surface

One side of the fracture surface for each fracture conductivity test was scanned by XRF microscopy by
utilizing a Horiba XGT-7000 X-Ray Analytical Microscope to obtain element distribution data. In order to
produce maps representing the element distribution along the fracture surface, a resolution of 100 m was
prescribed. Given the dimensions of the sample, four scan sections were typically required. The prescribed
group of elements considered during the scanning process was set to include Al, As, Ba, Ca, Cr, Fe, K, Mg,
Mn, Rb, S, Si, Sr, Ti, Zn, and Zr.
Background noise maps for the elements of interest were calculated and subtracted from the
corresponding element maps using a Java-based image processing software, ImageJ, resulting in gray-scale
images for the elements. These images were then stitched together to produce a final image representing
the entire fracture surface. Figure 8 shows the distribution of Ca, Fe, Si, S, and K on the fracture surface for
test sample E-B1-R1-S1. These elements were the most abundant on the fracture surface.
SPE-184857-MS 7

Figure 8Test sample E-B1-R1-S1 Ca, Fe, Si, S, and K distribution along the fracture surface (0.1 mm pixel
width). Values on right are the estimated intensity for each particular element over the fracture surface.

Variations in observed intensity were a result of varying presence of a particular element on the fracture
surface. However, the intensity was also affected by the differences in the distance from the X-ray source to
the fracture surface, which inherently was a function of the varying topography along the fracture surface. K
distribution along the fracture surface illustrates a slight interference from the aforementioned phenomena
(Fig. 8). These variations were considered an artifact of the procedure itself, and were mitigated during the
stitching process by ensuring each scan section element map had the same background intensity.
Stitched element maps over the full fracture surface were also overlaid on each other producing composite
images that served as proxies to mineral presence, as illustrated in Figure 9, where Ca, Fe, and Si were
overlaid to visualize how the fracture originated in a clay bed and progressed onto a limestone bed for test
sample E-A-R1-S1 (left to right).
8 SPE-184857-MS

Figure 9(a) Actual picture of fracture surface for test sample E-A-R1-S1; (b)
Ca, Fe, and Si overlaid on each other. Ca was colored blue, Fe red, and Si green.

From a quantitative perspective, the stitched element maps were used to estimate the average intensity
over the entire fracture surface for each particular element (Fig. 10). Based on the gathered data and the
known mineralogy found in the formation (Fig. 7), the analysis was focused solely on Ca, Fe, Si, and K. The
aforementioned elements were selected given their documented reliability as proxies to certain minerals,
including clay. Test samples obtained from downhole core carry the nomenclature Core 1, Core 2 and Core
3. As mentioned before, they are all from unit B.

Figure 10Average intensity over the fracture surface for each test sample.

Based on the gathered data and the abundance of Ca across all samples, a K/Ca element ratio was chosen
as a proxy to the presence of clay over the fracture surface. Variability in element distribution along the
fracture surface was evident across lithostratigraphic units, with a heavy presence of Ca across all test
samples, and a clustered presence of Fe, Si, and K for the zones with the highest TOC.
SPE-184857-MS 9

Fracture Conductivity Test Sample: Laser Surface Scanning on Fracture Surface

The fracture surface of each test sample was scanned using a laser profilometer system which produced
measured height (z) values for a specified mesh grid (x,y), as shown in Figure 11. z values for a rectangular
grid spacing of 0.05 in. were obtained and then processed to generate an interpolated surface based on
the measured heights. This surface served as the basis for computations including surface area and surface
roughness.

Figure 11Surface area for the bottom side of the fracture surface
of test samples: (a),(b) E-B1-S1-R1; (c) E-D-R1-S1; (d) E-E-R1-S1.

A flat test sample has a computed surface area of 10.87 in.2, with a corresponding x,y mesh grid covering
the entire flat surface. In this particular study, the x,y mesh grid covered a smaller portion of the flat surface
with only 8.81 in.2 surface area, reflecting an inward offset of 0.1 in. all around the sample. This was done
to increase the reliability of the readings and avoid false height readings attributed to the materials used
in the sample preparation that run along the edge of the sample. Given that the surface area was used for
relative comparison, considering 81% of the original x,y mesh grid sufficed. The surface roughness was
computed via the root mean square method (RRMS).
A significant variation of surface area across the test samples was observed, which resulted in some
samples having a longer and/or greater tortuous flowpath, such as sample E-E-R1-S1 (Fig. 11(d)).
10 SPE-184857-MS

A summary of estimated fracture surface area, maximum height difference (zmax) between the lowest
and highest point on the fracture surface, and fracture surface roughness for each test sample is provided
in Figure 12.

Figure 12Test sample estimated fracture surface area, maximum height

difference along the fracture, and fracture surface roughness values.

The test samples from downhole core, given their relatively small original dimensions, represented
several challenges during fracture creation, resulting in fractures that were typically at an angle with respect
to the length of the sample. This fracture surfaces had a higher surface area due to its orientation with respect
to the test sample (Fig. 12).

Proppant Type and Concentration

100-mesh sand was selected given its abundant use in the Eagle Ford shale, as well as its use in slick-water
treatments for gas wells. A sieve analysis was conducted and the results show that the sand complies with
the API standard regarding particle size distribution (Figure 13). The sand had a bulk density of 101.445
lbm/ft3 (1.625 g/cm3) and an apparent density of 163.784 lbm/ft3 (2.623 g/cm3), with a d50 of 0.007 in.

Figure 13Sieve analysis of 100-mesh sand utilized in all experiments.

A proppant areal concentration of 0.1 lbm/ft2 was chosen given its representativeness of the average
proppant concentration utilized in the Eagle Ford shale (Enriquez-Tenorio et al. (2016)).
SPE-184857-MS 11

Proppant Placement
The selected 0.1 lbm/ft2 proppant concentration of 100-mesh sand was manually placed evenly across the
bottom fracture surface. At this stage (pre-testing), the proppant was stored at room temperature, measured
at 70 F. Figure 14 shows proppant placement on the bottom-side fracture surface of test sample E-A-R1-S1.

2
Figure 14(a) Test sample E-A-R1-S1 with 0.1 lbm/ft 100-mesh sand distributed over the bottom-side fracture surface; (b) Test
2
sample E-A-R1-S1 with both fracture surfaces coming together, being propped apart by the with 0.1 lbm/ft 100-mesh sand.

Test Sample Preparation

The propped test sample was then coated with a silicone rubber and cured in an oven at 160F for 4 hours.
The previous was accomplished by inserting the propped test sample into a modular mold which was 0.0003
in. wider than the modified API fracture conductivity cell, and 0.075 in. larger than the test sample in
all directions. This interference fit ensured leakage was mitigated at high pressures. The silicone rubber
compound was the two-component Momentive RTV 627. A Momentive SS 4155 rubber adhesive primer
was also utilized.
Once the coating was cured, the coated sample was removed from the modular mold and prepared for
insertion into the modified API fracture conductivity cell. Small windows were cut through the silicon
rubber and removed to ensure connectivity between the test sample fracture and the pressure sensors as well
as the flow inlet and outlet (Fig. 15). Additionally, sealant tape was placed along and around the sample to
prevent leakage and channeling around the sample (Fig. 15).

Figure 15(a) Test sample after silicon rubber coating with small windows removed for fracture-sensor connectivity and
sealant tape to prevent channeling; (b) Prepared test sample being inserted into the modified API fracture conductivity cell.
12 SPE-184857-MS

Test Setup and Procedure

The damage phenomena was simulated under laboratory conditions by flowing saline water through
the propped test sample fracture under a constant closure stress of 4,000 psi. The short-term fracture
conductivity measurements were conducted by first flowing dry nitrogen through the fracture to ascertain
the undamaged initial fracture conductivity, followed by injecting a saline solution into the fracture until
steady state behavior was observed, and lastly, dry nitrogen was once again flowed to quantify the recovered
fracture conductivity. For the undamaged fracture conductivity measurements, dry nitrogen was selected
given that it does not react with shale. The degree of fracture conductivity impairment was measured by
comparing the initial dry nitrogen-based fracture conductivity measurement with the final dry nitrogen-
based fracture conductivity measurement to estimate the recovered fracture conductivity. Figure 16 displays
the experimental appratus schematic.

Figure 16Schematic of the setup used to evaluate fracture conductivity

by flowing dry nitrogen and/or water compositions (Zhang 2014).

The reconstituted flowback water was flowed at a constant rate of 0.2 mL/min until steady state was
reached, at which point dry nitrogen was flowed at a rate of 100 mL/min.

Saline Water Flowed Through Fracture

Saline water with a chemical composition similar to the typical Eagle Ford shale field flowback water
was utilized. A typical flowback water sample in the Eagle Ford play has a total dissolve solids (TDS)
concentration of 33,015 mg/L (Slutz et al. 2012). The reconstituted flowback water used in the experiments
was formulated to have a TDS concentration of 31,406 mg/L with a similar corresponding chemical
composition as the typical flowback water (Table 1).
SPE-184857-MS 13

Table 1Chemical composition of the reconstituted flowback water sample.

Ion Name Molecular Formula Reconstituted Flowback Field Flowback

Water Sample (mg/L) Water Sample (mg/L)

Calcium Ca2+ 1,312 1,270

Sodium Na+ 10,923 10,900

Potassium K+ 2 192

Magnesium Mg2+ 1 111

Strontium Sr2+ 203

Barium Ba2+ 10

Iron Fe 112

Chloride Cl 19,162 19,318

Nitrate NO3 1

Sulfate SO42 4 163

Bicarbonate HCO3 736

TDS 31,406 33,015

It is noted that the flowback water compositions are a function of location and time (Horner et al. 2011).
Fresher water and a higher chemical load dominate in the early flow, and as production continues, salt
concentrations increase to relatively high levels in later flow (King 2010). Flowback water flowrate, on the
other hand, decreases with time (Slutz et al. 2012). Eventually, a constant flowrate is attained constituted
by produced water only.

Experimental Results and Discussion

The Eagle Ford shale samples contained a high calcite and quartz content, meaning that technically they
are not a shale but rather an organic-rich mudstone. The Eagle Ford shale has a high level of heterogeneity,
with pure layers of shale interbedded with the organic-rich mudstone. Collecting this pure shale samples
is a nontrivial task given the inherent fragility of these layers. Nevertheless, the test samples in this study
were deemed representative, and constitute the best attainable quality of outcrop rock.

Fracture Conductivity Impairment

Fracture conductivity measurements were conducted for every unit and selected sub-units. Each sample
underwent a sequential gas-water-gas fracture conductivity test from start to finish. The typical fracture
conductivity curve consists of three regions. The first region consists of four individual readings of fracture
conductivity obtained by dry nitrogen flow at a closure stress of 1,000 psi, 2,000 psi, 3,000 psi, and 4,000 psi.
Once a closure stress of 4,000 psi was reached, it was held constant for the remainder of the test sequence.
The second region consists of a continuous fracture conductivity reading by the saline water flow until
steady state is achieved. The third region also consists of a continuous fracture conductivity monitoring
by again flowing dry nitrogen until steady state is achieved. Figure 17 and Figure 18 illustrate the typical
fracture conductivity behavior for a gas-water-gas test.
14 SPE-184857-MS

Figure 17Fracture conductivity measured by a sequential flow of dry

nitrogen, saline water, and once again dry nitrogen for test sample E-B1-R1-S1.

Figure 18F Fracture conductivity measured by a sequential flow of dry

nitrogen, saline water, and once again dry nitrogen for test sample E-B2-R1-S1.

Under a closure stress of 4,000 psi, Fig. 17 shows how test sample E-B1-R1-S1 had an initial undamaged
fracture conductivity of 565 md-ft, which was reduced to 463 md-ft after saline water was flowed through
the fracture. This represented a recovered fracture conductivity of 81.9%, or in other words, an unrecovered
fracture conductivity loss of 18.1%. Similarly, Fig. 18 shows how test sample E-B2-R1-S1 had an initial
SPE-184857-MS 15

undamaged fracture conductivity of 696 md-ft, which was reduced to 537 md-ft after saline water was
flowed through the fracture, representing an unrecovered fracture conductivity loss of 22.8%.
Figure 19 summarizes the fracture conductivity, Cf, measurements at 1,000 psi and 4,000 psi for all test
samples under all conditions according to their unit or sub-unit of origin. Dry nitrogen was flowed at 1,000
psi closure stress, whereas at 4,000 psi closure stress, dry nitrogen, saline water, and dry nitrogen were
flowed in the aforementioned sequence. By contrasting the fracture conductivity measured by dry nitrogen
before and after saline water flow, an unrecoverable percent loss of fracture conductivity was computed.

Figure 19Summary of fracture conductivity measurements at 1,000 and 4,000 psi flowing dry nitrogen and saline water.

The unrecoverable percent loss of fracture conductivity represents the degree of damage to the fracture
conductivity. The largest loss of conductivity took place in sub-units B1, B2, and B3, as well as unit C.
Downhole core test samples had very little variation in loss of fracture conductivity across the set.

Relationship with Fracture Surface Attributes

The majority of outcrop rock samples had a relatively smooth transition from low to high peaks over the
fracture surface, resulting in a direct relationship between fracture surface roughness and initial undamaged
fracture conductivity at a closure stress of 1,000 psi (Fig. 20). All but unit E exhibited such behavior, possibly
due to the fracture surface having the largest maximum height difference within this set of samples.
16 SPE-184857-MS

Figure 20Fracture conductivity measured by dry nitrogen flow

at a closure stress of 1,000 psi vs. fracture surface roughness.

The downhole core samples were heavily jagged and segmented along the fracture surface, with several
abrupt height changes, which in turn were an artifact of the downhole core itself and the subsequent fracture
creation. It is possible such abrupt changes in the fracture surface served as pinch points hindering the
undamaged fracture conductivity as the closure stress increased to 4,000 psi. Given the aforementioned
topography, surface roughness was not observed to have an impact on undamaged fracture conductivity
for this set of test samples (lower right portion of Fig. 20). For all test samples, outcrop and downhole
alike, the maximum height difference showed an inverse relationship in which low zmax resulted in higher
undamaged fracture conductivity values at 4,000 psi closure stress (Fig. 21).

Figure 21Fracture conductivity measured by dry nitrogen flow at a

closure stress of 4,000 psi vs. fracture surface maximum height difference.

Relationship to Mineralogy and Composition Along the Fracture Surface

The unrecoverable percent loss of fracture conductivity was contrasted against the amount of clay present
in the test samples (Fig. 22). The total clay was computed by adding all the clay type percentages obtained
SPE-184857-MS 17

by XRD. Similarly, quantitative element data obtained by XRF was used to estimate the element ratio K/
Ca, which served as a proxy to clay distribution along the fracture surface.

Figure 22Unrecoverable percent loss of fracture conductivity with corresponding total clay percentage and K/Ca clay proxy.

Figure 22 shows how the larger the presence of clay, the larger the value of fracture conductivity loss.
The total clay and the K/Ca proxy to clay displayed similar relative differences between the tested units and
sub-units. However, the K/Ca proxy to clay proved to have the strongest direct relationship with the percent
loss of fracture conductivity (Fig. 23). This can be attributed to the fact that the total clay was obtained from
a spot analysis via XRD on the cross-section of the core plug, whereas the K/Ca ratio was derived from
element distribution data gathered via XRF along the entire fracture surface of the fracture conductivity test
sample. Ultimately, the fracture surface mineralogy better characterizes the fracture conductivity loss, as
opposed to relying on bulk mineralogy.

Figure 23Unrecoverable percent loss of fracture conductivity vs. K/Ca proxy to clay.

The Brittleness Index, BI, was calculated based on the obtained mineralogy via XRD and the total organic
content using the equation presented by Wang and Gale (2009),

(1)
18 SPE-184857-MS

Figure 24 shows a direct relationship between BIand undamaged fracture conductivity at a closure stress
of 4,000 psi.

Figure 24Fracture conductivity measured by dry nitrogen flow at a closure stress of 4,000 psi vs. Brittleness Index.

The BI was not observed to have a definitive influence on damaged fracture conductivity nor on
unrecoverable percent loss fracture conductivity under the propped tested conditions.

Relationship with Poisson's ratio and the Brinell Hardness number

Poisson's ratio was reported to have an indirect relationship with undamaged fracture conductivity at high
closure stresses (Enriquez-Tenorio et al. 2016), however this behavior was not clearly observed in this
study. Figure 25 shows the lack of a clear correlation between the undamaged fracture conductivity and the
Poisson's ratio at a closure stress of 4,000 psi. This can be attributed to the mid-range level of closure stress
imposed thought the experiments.

Figure 25Fracture conductivity measured by dry nitrogen flow at a closure

stress of 4,000 psi vs. Poisson's ratio (values according to sample orientation).
SPE-184857-MS 19

Similarly, the Brinell harness number did not have a direct relationship with the undamaged fracture
conductivity at a closure stress of 4,000 psi (Fig. 26). This is possibly caused by the fact that Brill hardness
number was obtained from the core plug cross-section and not along the entire fracture surface, thus not
accounting for variability across the surface.

Figure 26Fracture conductivity measured by dry nitrogen flow at a closure stress

of 4,000 psi vs. Brinell hardness number (values according to sample orientation).

Other factors that might affect fracture conductivity are the presence of kaolinite- and smectite-
rich altered ash beds that create fracture conductivity pinch points, which in turn affect the potential
well productivity (Xu et al. 2016). Collecting test samples that contain the aforementioned ash beds
presents difficult challenges in sample handling and subsequent fracture creation. Furthermore, water-rock
interactions in the Eagle Ford shale yields water that has particles with a high tendency for precipitation,
which in turn can result in agglomeration of colloids, blocking the fracture aperture (Ali et al. 2015).

Conclusions
Drawing from the extensive data collected for each unit and relevant sub-unit of the Eagle Ford shale,
the degree of fracture conductivity impairment when exposed to saline water representing a typical field
flowback water was quantified and presented in terms of percentage loss. Additionally, a direct relationship
between clay presence on the fracture surface and the unrecoverable fracture conductivity loss after saline
water exposure was observed. The following are the main conclusions drawn from this study, listed in order
of importance:
1. The observed range in fracture conductivity loss due to water damage was between approximately 4
to 25% for the Eagle Ford shale. This range is relative low when compared to other shale plays that
have a higher clay content, such as the Barnett shale where fracture conductivity loss can be as high
as 88% due to a 55% clay content (Zhang et al. 2014, 2015). This suggests that fracture conductivity
loss due to water damage on the fracture surface might not be a major factor in causing the large early
decline rates observed in most Eagle Ford shale producing wells.
2. Unrecoverable loss of fracture conductivity due to water damage was directly influenced by the clay
presence along the fracture surface. The K/Ca proxy to clay had the strongest direct relationship with
the percent loss of fracture conductivity. Furthermore, mineralogy obtained along the fracture surface
20 SPE-184857-MS

via XRF proved more useful than relying on bulk mineralogy obtained by spot analysis via XRD on
the cross-section of the core plug.
3. Fracture surface topography has an effect on fracture conductivity as the closure stress increases. High
surface roughness typically correlated with a high initial undamaged fracture conductivity (at 1,000
psi closure stress). On the other hand, the maximum height difference across the fracture surface was
observed to be inversely related to undamaged fracture conductivity at 4,000 psi closure stress.
4. Brittleness inferred by mineralogy via XRD showed a direct relationship with undamaged fracture
conductivity (closure stress of 4,000 psi).
5. The Brinell hardness number did not exhibit a strong correlation to the undamaged fracture
conductivity (closure stress of 4,000 psi). This can be attributed to the fact that the indentation tests
were not gathered along the fracture surface, but rather on the cross-section of the core plugs.
6. The Poisson's ratio did not exhibit a strong correlation to the undamaged fracture conductivity (closure
stress of 4,000 psi). Such behavior can be attributed to the mid-range level of closure stress used in
the experiments.

Acknowledgements
The authors would like to acknowledge the Department of Energy, and the Research Partnership to Secure
Energy for America (RPSEA) for their support of this work (RPSEA project 11122-07-TAMU-Zhu). The
authors would also like to acknowledge BP for graciously allowing access to the outcrop site. Additionally,
the authors would like to thank Dr. Michael Tice for providing access to the downhole core samples, as well
as the use of the X-ray Fluorescence Analytical Microscope (Horiba Scientific XGT-7200 series).

Nomenclature
d50 Median proppant diameter, in.
zmax Maximum difference in measured heights across fracture surface, in.
Cf Fracture conductivity, md-ft

References
Akrad, O. M., Miskimins, J. L., and Prasad, M. 2011. The Effects of Fracturing Fluids on Shale Rock Mechanical
Properties and Proppant Embedment. Presented at the SPE Annual Technical Conference and Exhibition, Denver,
Colorado, USA, 30 October2 November. Paper SPE-146658-MS. http://dx.doi.org/10.2118/146658-MS
Ali, M., and Hascakir, B. 2015. Water-Rock Interaction for Eagle Ford, Marcellus, Green River, and Barnett Shale
Samples. Society of Petroleum Engineers. Presented at the SPE Eastern Regional Meeting, Morgantown, West
Virginia, USA, 1315 October. Paper SPE-177304-MS. http://dx.doi.org/10.2118/177304-MS
API RP 61, Recommended Practices for Evaluating Short Term Proppant Pack Conductivity. First Edition, October 1,
1989. Washington, DC: API.
Awoleke, O. 2013. Dynamic Fracture Conductivity: An Experimental Investigation Based on Factorial Analysis. PhD
Dissertation, Texas A&M University, College Station, Texas. (May 2013).
Cooke Jr., C.E. 1975. Effect of Fracturing Fluids on Fracture Conductivity. Journal of Petroleum Technology, 27(10):
1273-1282. Paper SPE-5114-PA. http://dx.doi.org/10.2118/5114-PA
Donovan, A. D., and S., Staerker 2010, Sequence stratigraphy of the Eagle Ford (Boquillas) formation in the subsurface
of South Texas and the outcrops of West Texas. Gulf Coast Association of Geologic Societies Transactions, v. 60, p.
861899.
Donovan, A.D., Staerker, T.S., Pramudito, A., Li, Weiguo., Corbett, M.J., Lowery, C.M., Romero, A.M., and Gardner,
R.D. 2012. The Eagle Ford Outcrops of West Texas: A Laboratory for understanding Heterogeneities Within
Unconventional Mudstone Reservoirs. CGAGS Journal, v. 1(2012) P. 162185.
Enriquez-Tenorio, O. 2016. A Comprehensive Study of the Eagle Ford Shale Fracture Conductivity. MS Thesis, Texas
A&M University, College Station, Texas. (August 2016).
Enriquez-Tenorio, O., Knorr, A., Zhu, D. et al 2016. Relationships Between Mechanical Properties and Fracturing
Conductivity for the Eagle Ford Shale. Presented at the SPE Asia Pacific Hydraulic Fracturing Conference, Beijing,
China, 2426 August. Paper SPE-181858-MS. http://dx.doi.org/10.2118/181858-MS
SPE-184857-MS 21

Fredd, C. N., McConnell, S. B., Boney, C. L., & England, K. W. (2000)Experimental Study of Hydraulic Fracture
Conductivity Demonstrates the Benefits of Using Proppants. Society of Petroleum Engineers. Presented at the
SPE Rocky Mountain Regional/Low-Permeability Reservoirs Symposium and Exhibition, Denver, Colorado, 1215
March. Paper SPE-60326-MS. http://dx.doi.org/10.2118/60326-MS
Gardner, R., Pope, M. C., Wehner, M. P., Donovan, A. D. 2013. Comparative Stratigraphy of the Eagle Ford Group Strata
in Lozier Canyon and Antonio Creek, Terrell County, Texas. GCAGS Journal v. 2(1): 4252.
Horner, P., Halldorson, B., and Slutz, J. A. 2011. Shale Gas Water Treatment Value Chain - A Review of
Technologies, including Case Studies. Society of Petroleum Engineers. Presented at the SPE Annual Technical
Conference and Exhibition, Denver, Colorado, USA, 30 October2 November. Paper SPE-147264-MS. http://
dx.doi.org/10.2118/147264-MS
ISO 13503-5, Petroleum and Natural Gas Industries-Completion Fluids and Materials-Part 5: Procedures for Measuring
the Long-Term Conductivity of Proppants. First Edition, July 1, 2006. Geneva, Switzerland: ISO.
Kamenov, A. N. 2013. The Effect of Proppant Size and Concentration on Hydraulic Fracture Conductivity in Shale
Reservoirs. MS Thesis, Texas A&M University, College Station, Texas. (May 2013).
King, G. E. 2010. Thirty Years of Gas Shale Fracturing: What Have We Learned? Society of Petroleum Engineers.
Presented at the SPE Annual Technical Conference and Exhibition, Florence, Italy, 1922 September. Paper
SPE-133456-MS. http://dx.doi.org/10.2118/133456-MS
Knorr, A. F. 2016. The Effect of Rock Properties on Fracture Conductivity in the Eagle Ford. MS Thesis, Texas A&M
University, College Station, Texas. (May 2016).
Miceli-Romero, A.A. 2014. Subsurface and outcrop organic geochemistry of the Eagle Ford shale (Cenomanian-
Coniacian) in west, southwest, central, and east Texas. PhD Dissertation, University of Oklahoma, Norman,
Oklahoma.
Palisch, T., Duenckel, R., Bazan, L. et al 2007. Determining Realistic Fracture Conductivity and Understanding its
Impact on Well Performance Theory and Field Examples. Presented at the SPE Hydraulic Fracturing Technology
Conference, College Station, Texas, USA, 2931 January. Paper SPE 106301. http://dx.doi.org/10.2118/106301-MS
Parker, M.A. and McDaniel, B.W. 1987. Fracturing Treatment Design Improved by Conductivity Measurements Under
In-Situ Conditions. Presented at the SPE Annual Technical Conference and Exhibition, Dallas, Texas, USA, 2730
September. Paper SPE 16907. http://dx.doi.org/10.2118/16901-MS
Ramurthy, M., Barree, R.D., Kundert, D.P.et al 2011. Surface-Area vs Conductivity-Type Fracture Treatments
in Shale Reservoirs. SPE Production and Operations, 26(4): 357367. Paper SPE- 140169-MS. http://
dx.doi.org/10.2118/140169-MS
Slutz, J. A., Anderson, J. A., Broderick, R., & Horner, P. H. 2012. Key Shale Gas Water Management Strategies: An
Economic Assessment. Society of Petroleum Engineers. Presented at the International Conference on Health, Safety
and Environment in Oil and Gas Exploration and Production, Perth, Australia, 1113 September. Paper SPE-157532-
MS. http://dx.doi.org/10.2118/157532-MS
Wang, F. P. and Gale, J. F. 2009. Screening Criteria for Shale Gas Systems. Gulf Coast Association of Geological Societies
Transactions, vol. 59, 779793.
Xu, J., Fisher, K., Qiu, F. et al 2016. Impact of Ash Beds on Production in Eagle Ford Shale. Presented at the SPE
Hydraulic Fracturing Technology Conference, The Woodlands, Texas, USA, 9-11 February. Paper SPE-179110-MS.
http://dx.doi.org/10.2118/179110-MS
Zhang, Junging, 2014. Creation and Impairment of Hydraulic Fracture Conductivity in Shale Formations. PhD
Dissertation, Texas A&M University, College Station, Texas (August 2014).
Zhang, J., Ouyang, L., Hill, A. D., and Zhu, D., 2014. Experimental and Numerical Studies of Reduced Fracture
Conductivity due to Proppant Embedment in Shale Reservoirs. Presented at the SPE Annual Technical Conference and
Exhibition, Amsterdam, The Netherlands, 2729 October. Paper SPE-170775-MS. http://dx.doi.org/10.2118/170775-
MS
Zhang, J., Zhu, D., & Hill, A. D. 2015. Water-Induced Fracture Conductivity Damage in Shale Formations. Presented at the
SPE Hydraulic Fracturing Technology Conference, The Woodlands, Texas, USA, 35 February. Paper SPE-173346-
MS. http://dx.doi.org/10.2118/173346-MS
Zhang, L., Marcantonio, F., Haskakir, B. et al 2016. Solid and Soluble Products of Engineered Water/Rock Interactions
in Eagle Ford Group Chemofacies. Presented at the Berg-Hughes Center for Petroleum and Sedimentary Systems 7th
Annual Research Symposium. Graduate Student Symposium Poster Session, College Station, Texas, USA, 11 October.
Crisman Institute for Petroleum Research, Project 1.04.16.