GEOPHYSICS, VOL. 86, NO. 1 (JANUARY-FEBRUARY 2021); P. E13–E20, 11 FIGS., 4 TABLES.

10.1190/GEO2019-0367.1

Characterization of an electrically conductive proppant for fracture

diagnostics

Peng Zhang1, Williams Ozowe2, Rodney T. Russell2, and Mukul M. Sharma2

ABSTRACT is mixed with sand, the effective electrical conductivity of the

mixture decreases with an increasing weight percentage of sand.
Fracture diagnosis with electromagnetic (EM) and electrical Although sand degrades the contact between PC particles, the
tools requires proppants with high electrical conductivity and electrical conductivity stays reasonably high (approximately
mechanical strength. Lab measurements of the electrical and hy- 1700 S/m) when 50% sand is added. Hydraulic conductivity
draulic conductivity of proppants are critical for selecting the measurements show that when a fracture is propped with pure
best candidates. Such measurements greatly benefit simulations, PC, the measured fracture conductivity is greater than 1.20 ×
field tests, and the ultimate application of such proppants in the 10−15 m2 (4 md · ft) (dimensionless fracture conductivity
field. To that end, a new lab protocol is developed for measuring greater than 100 for a shale with 9.87 × 10−20 m2 or 100 nD
the electrical and hydraulic conductivity of proppants. The lab permeability) under a confining pressure of 4.14 × 107 Pa
setup, which mainly includes a resistivity core holder and a (6000 psi). This means that a fracture propped with PC is infi-
Hassler sleeve core holder, allows for simulation of realistic nitely conductive in a typical shale formation. When sand is
pressure and temperature conditions when making measure- added, the fracture’s hydraulic conductivity becomes even higher,
ments. Petroleum coke (PC) is proposed as a candidate proppant which clearly shows PC’s ability of sustaining high stresses. The
because of its widespread availability and low cost. Lab mea- proposed protocol provides a robust and effective method that can
surements show that the effective electrical conductivity of pure be generalized for lab testing for other candidate proppants. The
PC in a model fracture is approximately 5000 S/m, under a data presented clearly show that PC has the potential for field-
closure stress greater than 2.76 × 107 Pa (4000 psi). When PC scale applications in EM hydraulic fracture diagnostics.

INTRODUCTION are used for inferring fracture height. Unfortunately, these methods
either focus on near-wellbore proppant detection, such as radioac-
Knowledge of hydraulic-fracture geometries, especially the prop- tive tracers (Scott et al., 2010) and charged boron carbide contrast
pant distribution, is critical for optimizing fracture treatment design agents (Tang et al., 2017), or they suffer from nonuniqueness of
and improving completion efficiency in the life cycle of a reservoir. their results, such as chemical tracers (Elahi and Jafarpour, 2018).
There are a few categories of diagnostic methods that endeavor to Distributed temperature and acoustic sensing (DTS and DAS)
characterize hydraulic fractures in the industry. Tiltmeters are one of through fiber optics can provide insights into fracturing fluid dis-
the earlier diagnostic methods that have been used to infer fracture tribution into different clusters and postfracture crossflow (Sierra
azimuth (Warpinski, 1996; Cipolla and Wright, 2000), but they can et al., 2008; Sookprasong et al., 2014; Bhatnagar, 2016), but they
barely detect created fracture length and height, due to their low are not able to track the lateral extension of the proppant. Micro-
sensitivity. Proppant or fluid tracers that are typically pumped with seismic imaging and acoustic logging have been the primary tools for
fracturing fluids can be detected based on various mechanisms and identification of far-field fracture geometry or stimulated reservoir

Manuscript received by the Editor 18 June 2019; revised manuscript received 15 August 2020; published ahead of production 1 October 2020; published
online 04 January 2021.
1
Formerly The University of Texas at Austin, Hildebrand Department of Petroleum and Geosystems Engineering, Austin, Texas 78712, USA; presently
Schlumberger, Houston, Texas 77042, USA. E-mail: pengzhang14@[Link] (corresponding author).
2
The University of Texas at Austin, Hildebrand Department of Petroleum and Geosystems Engineering, Austin, Texas 78712, USA. E-mail: williamsozowe@
[Link]; [Link]@[Link]; msharma@[Link].
© 2021 Society of Exploration Geophysicists. All rights reserved.

E13

Downloaded from [Link]

by Fundação Oswaldo Cruz user
 E14 Zhang et al.

volume (Mayerhofer et al., 2010; Xu et al., 2018). Unfortunately, tions. The effective electrical and hydraulic conductivity of a
these techniques do not directly assess the proppant distribution propped fracture are dictated by the proppant, the fracturing fluid,
either, which is the main factor controlling the productivity of a as well as factors such as in situ stress in the reservoir. To better
fractured well. simulate realistic downhole conditions, a lab protocol should be
To overcome the limitations of the previous methods, electro- able to conduct consistent and accurate measurements on different
magnetic (EM) and electrical borehole measurement methods have proppants, apply a high enough confining pressure, and saturate the
been proposed to directly map proppants with distinct physical proppant with different fluids as needed. Although a few candidate
properties. Through numerical simulations and field experiments, proppants have been reported in the literature, the cost of applying
researchers have demonstrated that EM induction tools (Basu and these materials in large quantities remains questionable. Therefore,
Sharma, 2014; Yang et al., 2015; LaBrecque et al., 2016; Palisch another important goal of this work is to test a material with very
et al., 2016; Zhang et al., 2016; Shiriyev et al., 2018) and electrical high electrical conductivity and mechanical strength that is eco-
tools (Weiss et al., 2016; Gabelmann et al., 2020; Zhang et al., nomical for commercial use. The proppant material selected, petro-
2018a, 2018b) show great potential in open-hole and case-hole leum coke (PC), is cost effective and readily available in large
wells, respectively. Most of these new generation tools, depending quantities in the market. It is a by-product of oil refining with a
on their configurations, rely on proppants that demonstrate high density of 2.03 g∕cm3 and a particle size of 150–400 μm. More
electrical conductivity contrast with the formation. than 96% of its composition is carbon, which makes it thermally
Prior to the developments of field-applicable proppants, people stable and a very good electrical conductor. PC has been used
have studied “proppant analogs,” which are materials that display to make anodes for the aluminum, steel, and titanium smelting
unique EM properties but may not be acceptable for injection into industries.
deep reservoirs due to their lack of mechanical strength. For exam- The remainder of this paper is organized as follows. First, the
ple, steel shots exhibit a much higher relative permittivity than that experimental systems and the procedures are introduced. Next,
of shale at the 100 Hz frequency. Small-scale field experiments at the measured electrical and hydraulic conductivity of the PC, mixed
shallow depths indicate that this material shows a significant in- with different amounts of sand, are presented. Finally, the results
crease in the intrinsic phase response over that of the background and economics of PC’s commercial use are discussed.
(Symington et al., 2010). Coke breeze, as an electrically conductive
proppant analog, has been tested as a contrast agent as well (Sen
et al., 2018). Mixing coke breeze with sand can reduce its con-
sumption while keeping reasonably high electrical conductivity METHODS
(LaBrecque et al., 2016). Recent advancements in proppant detec- Electrical conductivity measurements
tion bring up a few new candidates that can potentially be applied
in deep wells. A ceramic proppant with an electrically conductive A resistivity core holder (Figure 1) is used to conduct electrical
coating, which is expected to possess the high mechanical strength resistivity measurements on well-consolidated cores, such as shale
of ceramic, has been tested in a horizontal well 2400 m deep and sandstone with a fracture filled with proppant in the middle
(Palisch et al., 2016). However, the exact electrical and hydraulic (axially). The measurements are conducted using a four-point probe
conductivity of this proppant have not been reported and the re- method, which can minimize the error introduced by contact resis-
ported costs are prohibitive. Using the effective medium theory tance among the electrodes, wires, etc. A core with a length of
(Berryman and Hoversten, 2013), the electrical conductivity of a 5.08 cm (2 in.) and a diameter of 3.81 cm (1.5 in.), which is rep-
coke-breeze-coated sand is estimated to be as high as 2000 S/m resented by the black cylinder in Figure 1, is wrapped with a rubber
when the volume of proppant in the slurry is greater than 70% (Hov- sleeve and constrained by two end pieces of the core holder on both
ersten et al., 2015). Yet, experimental tests are needed to verify these sides. Before injecting fluid through one of the current-carrying
estimations. electrodes, the system is evacuated to remove trapped air and a con-
Lab measurements on one or several types of conductive prop- fining pressure or closure stress is applied around the rubber sleeve
pants can provide valuable information for numerical simulations, using a hydraulic pump, to prevent the fluid bypassing the core
as well as help in selecting the desired proppant for field applica- through the gap between the core and the rubber sleeve (Figure 1).
When taking the measurements, an alternating current I of 60 Hz
is applied via current-carrying electrodes, which are isolated from
the outer shell of the core holder such that current is forced to flow
through the core with a fracture propped by PC. A voltage U is
measured on voltage-sensing electrodes in the center (Figure 1).
These electrodes are connected to two ring electrodes that are
directly in contact with the core. The spacing between the two volt-
age-sensing electrodes is 2.54 cm (1 in.). The resistance of the mea-
sured part of the core can be obtained from Ohm’s law:

U
R¼ : (1)
I
Figure 1. Configuration of the resistivity core holder. The yellow
thick lines identify the electrodes, whereas the black cylinder and
the red arrows represent the sample and the confining pressure, Assuming that the core with resistance R has a cross-sectional area
respectively. A and length L, the resistivity of the core is

Downloaded from [Link]

by Fundação Oswaldo Cruz user
 Proppant for fracture diagnostics E15

R·A kf · wf · D ΔP
ρ¼ ; (2) Q¼ · ; (4)
L μ L
where the electrical conductivity σ is calculated as the reciprocal of where the product of fracture permeability and width kf · wf is de-
the resistivity ρ. fined as the fracture conductivity; D and L are the core diameter and
This setup is also used to make electrical resistivity measure- length, respectively; and μ is the fluid viscosity. This procedure is
ments on bulk proppant for a sanity check. Because the resistivity repeated for incremental closure stresses from 6.90 × 106 Pa to
core holder was initially designed for well-consolidated cores, it is 5.52 × 107 Pa (1000–8000 psi) to study the proppant’s ability to
necessary to verify that it works properly for unconsolidated grains sustain closure stresses.
as well. Sand is packed in the cylindrical rubber sleeve, loaded to
the core holder, and sealed in place for this verification. The mea-
RESULTS
sured porosity of the sand pack is 25.2%. Brines of different con-
centrations were injected to saturate the sand pack (Sw ¼ 100%), Using the experimental systems demonstrated previously, electri-
and resistivity measurements were conducted using this resistivity cal and hydraulic conductivity measurements were conducted on
core holder. Because the diameter of the sand pack and the spacing
between the two voltage-sensing electrodes are constant, A and L in
Table 1. Measured resistivity of a brine-saturated sand pack
equation 2 are 11.40 cm2 and 2.54 cm, respectively. Table 1 lists the with respect to different brine concentration.
brine concentration c, resistivity Rw, and measured resistivity of the
saturated sand pack Ro .
A plot of Ro versus Rw shows the measured resistivity of the c (g/L) 36 70 100 150
unconsolidated sand pack, and Ro is proportional to the resistivity
of saturation fluid Rw (Figure 2). The slope of the line represents the Rw ðΩ · mÞ 0.18 0.10 0.08 0.06
formation factor F in Archie’s law (Archie, 1942): Ro ðΩ · mÞ 0.86 0.54 0.44 0.29

Ro ¼ F · Rw ; (3)

where F depends on the porosity, tortuosity factor, and cementation

exponent of the porous medium. Considering the observed tortuos-
ity factor (0.5–1.5) and cementation exponent (1.3) for unconsoli-
dated sand, the measured formation factor F (approximately 4.55) is
quite reasonable for the sand pack with a measured porosity of
25.2% (Archie, 1942). This clearly validates the applicability of this
experimental system for resistivity measurements. The effective re-
sistivity of bulk proppants and proppants within a fracture is mea-
sured with the same setup.

Hydraulic conductivity measurements

The hydraulic conductivity of proppants in a fracture is measured
using an experimental system described in Figure 3. Previous ap-
plications of this system for fracture conductivity measurements Figure 2. Measured resistivity of the saturated sand pack Ro versus
have been reported by Wu et al. (2017). A Berea sandstone core resistivity of the brine Rw . A linear relationship is observed.
of 2.54 cm (1 in.) in diameter by 20.32 cm
(8 in.) in length is split along the axial direction.
A fracture of 1 mm wide between the two halves
of the core is filled with proppant, and the core is
then loaded into a Hassler sleeve core holder
(Figure 3). The system is evacuated to remove
trapped air, and a confining closure stress is ap-
plied for 24 hours.
During the measurements, a 3% brine solution
is pumped from the accumulator through the core
at a constant flow rate Q, which can be adjusted
by the back pressure regulators (BPRs). Confin-
ing oil is injected by a hydraulic pump to apply a
closure stress on the core. For each closure stress
applied, the pressure drop ΔP across the core is
measured by a differential pressure transducer Figure 3. Experimental system for fracture conductivity measurements. Fluid is injected
and the fracture conductivity can be calculated from the accumulator into the core (shown in the dotted square). The BPRs are used to
from Darcy’s law: adjust the flow rates. A confining pressure is applied during the measurements.

Downloaded from [Link]

by Fundação Oswaldo Cruz user
 E16 Zhang et al.

PC with two different particle sizes. We also explored the possibility field, we tested a few combinations with an incremental weight
of mixing it with sand, to provide an economic analysis for field percentage of sand from 0% to 75%. Again, equation 2 is used to
applications. calculate the electrical resistivity, but A in this case is the cross-
sectional area of the fracture D · wf .
Electrical conductivity of PC For PC with a particle size of 210–420 μm, the measured elec-
trical resistivity at various confining pressures is shown in Figure 6.
The effective electrical conductivity of bulk PC and PC within The green, blue, red, and orange curves represent the results for PC
a planar fracture is measured using the experimental system in mixed with 0%, 25%, 50%, and 75% sand (210–420 μm), respec-
Figure 1. For bulk PC, the PC powder was packed in the cylindri- tively. The O-ring seal around the voltage sensing electrodes keeps
cal rubber sleeve, loaded to the core holder, and sealed in place. the confining oil from leaking from the core holder. Due to the de-
Seawater (35 g∕L NaCl solution) was injected into the proppant formation of the rubber sleeve at elevated confining pressure, the
pack after the system was evacuated. Figure 4 shows two sets of contact between the voltage sensing electrodes and the O-ring seal
measurements on PC with different particle sizes. The blue and becomes loose. The confining oil will start to leak at some point,
red curves represent the results for PC with particle sizes of which cannot be easily controlled, and the confining pressure can-
210–420 μm and 150–210 μm (corresponding to 40–70 mesh and not be increased further. The end point value on each curve denotes
70–100 mesh, respectively). The initial PC bulk density for both
the measured resistivity at the highest confining pressure achieved.
measurements is approximately 1.27 g∕cm3, with the porosity cal-
Although these numbers (Table 2) are obtained under different pres-
culated to be 37.6%. Equation 2 is used to calculate the electrical
sure levels, they are closely representative of the ultimate resistivity
resistivity. When no confining pressure is applied, the electrical re-
of the corresponding PC-sand mixture because the curves reach a
sistivity is measured as 7.0 × 10−4 Ω · m. The end point values for
plateau.
tests 1 and 2 are 1.5 × 10−4 and 1.9 × 10−4 Ω · m, as the confining
Similar to the results of bulk PC, for a given composition of PC-
pressure reaches 2.07 × 107 Pa (3000 psi) (Figure 4). Seawater has
sand mixture, the electrical resistivity decreases with an increasing
a resistivity of 0.2 Ω · m at room temperature, which is several or-
confining pressure, due to a better contact between the PC particles.
ders of magnitude greater than the resistivity of PC. For that reason,
The electrical resistivity of pure PC is 1.99 × 10−4 Ω · m under a
the measured resistivity of bulk PC saturated with seawater is about
confining pressure of 2.76 × 107 Pa (4000 psi), which is close to
the same as without seawater (Zhang et al., 2016). Particle size does
the value of bulk PC (Figure 6a). As the weight percentage of sand
not play a noticeable role in the measured results, and the average
increases, the effective resistivity of the mixture increases. This is
conductivity is approximately 6000 S/m, when the confining pres-
because sand, as an insulating material, fills the spaces between the
sure is greater than 1.38 × 107 Pa (2000 psi).
PC particles and prevents direct electrical contact between PC par-
Apart from the intrinsic conductivity, the contact between the PC
ticles. When the weight percentage of sand reaches 75%, the mix-
particles is the most important factor that controls the effective re-
ture’s resistivity becomes substantially higher (3.24 × 10−3 Ω · m
sistivity. As the closure pressure increases, the contact improves and
at a confining pressure of 3.45 × 107 Pa or 5000 psi), which will
the resistivity drops. When the confining pressure rises to approx-
affect its effectiveness in field applications (Figure 6b).
imately 1.38 × 107 Pa (2000 psi), this effect reaches a plateau, lead-
We also tested the resistivity of finer particles (150–210 μm) in a
ing to electrical conductivity of approximately 6000 S/m (Figure 4).
sandstone fracture (Figure 7). The end point values on the curves are
Electrical conductivity measurements were also conducted on
listed in Table 3. Compared to the results for coarser PC (Table 2),
PC in a fractured sandstone core, mimicking proppant in hydraulic
the overall resistivity and conductivity of different PC-sand mix-
fractures in the reservoir. The core has the same dimensions as
tures stay at the same level, which means that PC’s particle size does
described previously, and the initial fracture width wf ¼ 3 mm
not play an important role in the effective resistivity. The ultimate
(Figure 5). The procedure was repeated for incremental confining
electrical conductivity of pure PC under high confining pressure is
pressures, with particle sizes of 210–420 μm and 150–210 μm. To
approximately 5000 S/m. Even with up to 50 wt% of sand mixed,
explore the possibility of using a mixture of PC and sand in the
the proppant’s effective conductivity stays as high as 1700 S/m,
when subject to a closure stress of 2.76 × 107 Pa (4000 psi) or
higher.

Figure 4. Variation of the electrical resistivity of the PC pack with Figure 5. (a) A schematic of proppant in a fracture of width wf .
confining pressure. Seawater was injected into the pack for tests 1 (b) A fractured sandstone core propped by PC. The core length
and 2. and diameter are 5.08 and 3.81 cm, respectively.

Downloaded from [Link]

by Fundação Oswaldo Cruz user
 Proppant for fracture diagnostics E17

Hydraulic conductivity of PC closure stress, to demonstrate the relative change (Figure 8b). As the
closure stress goes up, the measured fracture conductivity drops
The hydraulic conductivity of pure PC, half PC–half sand mix- quickly, due to the compacted pore space. From 1.38 × 107 Pa
ture, and pure sand was tested under confining pressure using the to 5.52 × 107 Pa (2000–8000 psi), the fracture conductivity can de-
experimental system shown in Figure 3. The procedure was re- crease up to 80% for pure PC. Conversely, the fracture propped with
peated for proppants with two particle sizes (210–420 μm and pure sand is less affected, which indicates that the sand grains have
150–210 μm). For proppants with coarser grains (210–420 μm), a lower compressibility. Mixing 50 wt% sand with PC can increase
Figure 8 shows the measured fracture conductivity kf wf at various the fracture conductivity from 1.50 × 10−15 m3 (5 md·ft) to greater
closure stress (confining pressure). The yellow, black, and red mark- than 3.01 × 10−15 m3 (10 md·ft) under 5.52 × 107 Pa closure stress.
ers represent the results for pure PC, 50% PC-50% sand, and pure
sand, respectively. We also plot the normalized fracture conductiv-
ities with respect to the ones at less than 1.38 × 107 Pa (2000 psi)

Figure 6. Measured resistivity of PC (210–420 μm) at various con- Figure 7. Measured resistivity of PC (150–210 μm) at various con-
fining pressures. The green, blue, red, and orange curves represent fining pressures. The green, blue, red, and orange curves represent
the results for PC mixed with 0%, 25%, 50%, and 75% sand the results for PC mixed with 0%, 25%, 50%, and 75% sand
(210–420 μm), respectively. (150–210 μm), respectively.

Table 2. The measured electrical resistivity and conductivity Table 3. The measured electrical resistivity and conductivity
of PC-sand mixtures at different confining pressures. The of PC-sand mixtures at different confining pressures. The
particle size of PC in this set of measurements is 210– particle size of PC in this set of measurements is
420 μm. 150–210 μm.

wt% of Confining pressure Resistivity Conductivity wt% of Confining pressure Resistivity Conductivity
sand (Pa) (Ω·m) (S/m) sand (Pa) (Ω·m) (S/m)

0 2.76 × 107 1.99 × 10−4 5.01 × 103 0 3.17 × 107 2.12 × 10−4 4.72 × 103
25 2.07 × 107 3.64 × 10−4 2.75 × 103 25 3.45 × 107 2.64 × 10−4 3.79 × 103
50 2.76 × 107 6.28 × 10−4 1.59 × 103 50 3.45 × 107 5.73 × 10−4 1.75 × 103
75 3.45 × 107 3.24 × 10−3 3.09 × 102 75 3.45 × 107 3.18 × 10−3 3.14 × 102

Downloaded from [Link]

by Fundação Oswaldo Cruz user
 E18 Zhang et al.

The fracture conductivities for finer particles (150–210 μm) are productivity models (Prats, 1961; Friehauf, 2009) show how the
plotted in Figure 9. Clearly, a fracture propped by finer grains has a stimulation ratio J∕J0 , where J and J 0 are the productivity of a well
lower conductivity when subject to the same closure stress. In a with and without a fracture, varies with FCD (Figure 10). For all
porous medium composed of well-sorted grains, a smaller grain size of the cases investigated, the stimulation ratio reaches a maximum
leads to a larger specific surface area, which leads to more resistance (as FCD ≥ 160), which means that the fracture is already infinitely
to fluid flow. Although the particle size does not play a significant conductive (Figure 10). Therefore, the mechanical strength of PC
role in its electrical conductivity, coarser PC is preferred to achieve a (or a PC-sand mixture) under high closure stresses is not a concern
higher hydraulic conductivity. for low-permeability rocks from the perspective of hydraulic con-
ductivity.
DISCUSSION The electrical conductivity of pure PC is greater than 5000 S/m,
when subject to a closure stress higher than 2.76 × 107 Pa
To better interpret what the measured results mean for field ap- (4000 psi). Although mixing PC with sand degrades the contact
plications, we calculate the dimensionless fracture conductivity between PC particles, the electrical conductivity stays reasonably
FCD ¼ ðkf · wf Þ∕ðLf · kmatrix Þ by assuming a matrix permeability high (approximately 1700 S/m) when up to 50% sand is added. This
kmatrix ¼ 9.87 × 10−20 m2 (10−4 md) and a fracture half-length Lf of conductivity is more than three orders of magnitude higher than the
76.2 m (250 ft). For the worst-case scenario observed, when pure typical conductivity of shale (<1 S/m, Figure 11), which makes it a
PC (150–210 μm) is under a closure stress of 4.14 × 107 Pa great contrast agent that can be tracked by EM or electrical tools.
(6000 psi), the measured fracture conductivity is approximately Although PC’s mechanical strength does not restrain its applica-
1.20 × 10−15 m3 (4 md · ft) (Figure 9a). The corresponding dimen- tions in deep formations, mixing a larger fraction of PC in sand is
sionless fracture conductivity FCD is calculated as 160. Fracture

Figure 8. (a) The measured fracture conductivities of PC-sand mix- Figure 9. (a) The measured fracture conductivities of PC-sand mix-
ture (210–420 μm) at various confining pressures. (b) Fracture con- ture (150–210 μm) at various confining pressures. (b) Fracture con-
ductivities normalized by the corresponding values under ductivities normalized by the corresponding values less than
1.38 × 107 Pa (2000 psi) confining pressure. 1.38 × 107 Pa (2000 psi) confining pressure.

Downloaded from [Link]

by Fundação Oswaldo Cruz user
 Proppant for fracture diagnostics E19

pumping pure sand, while still achieving an electrical conductivity

of 1.67 × 103 S∕m (which is expected to provide a good conduc-
tivity contrast with most hydrocarbon-bearing sedimentary rocks).
Using more than 75% sand is generally not recommended, because
the effective electrical conductivity will drop one order of magni-
tude. With the signal strength proportional to the fracture’s electri-
cal conductivity, it makes detecting large fractures very challenging
(Yang et al., 2015; Zhang et al., 2018b).

CONCLUSION
A new experimental method is developed for measuring the elec-
trical and hydraulic conductivity of proppants. The lab setup mainly
Figure 10. Stimulation ratio J∕J0 versus dimensionless fracture
conductivity FCD from two models. The L and re represent the includes a resistivity core holder and a Hassler sleeve core holder,
half-length of the fracture and drainage radius of the well, respec- which allow for simulation of realistic downhole conditions when
tively. “Current model” and “Prats” represent Friehauf’s and Prats’s making electrical conductivity and fluid permeability measurements
models, respectively (Friehauf, 2009). of simulated fractures filled with electrically conductive proppant.
By conducting consistent and accurate measurements with different
proppants, this protocol can serve as a benchmark for selecting the
best candidates for a new generation of EM fracture diagnostic
proppants. Experimental results on PC show that the effective con-
ductivity of this proppant increases as the confining pressure rises,
due to the better contact between the PC particles under higher
stress. The particle size does not play a noticeable role in the
measured results. Pure PC displays an electrical conductivity of
approximately 5000 S/m when the confining pressure is greater
than 2.07 × 107 Pa (3000 psi). Even when up to 50% sand is added,
the electrical resistivity stays reasonably high (approximately
1700 S/m), which makes it an excellent resistivity contrast agent
in shale formations.
Fracture conductivity measurements show that when PC is mixed
Figure 11. Typical range of electrical resistivity for selected earth with sand, the hydraulic conductivity increases, due to sand’s lower
materials (Ley-Cooper et al., 2015). compressibility. Even when propped with pure PC, the fracture con-
ductivity remains greater than 1.20 × 10−15 m3 (4 md · ft) under a
confining pressure of 4.14 × 107 Pa (6000 psi), which means that
Table 4. The effective electrical conductivity and cost with the fracture is infinitely conductive for a typical shale permeability
respect to different weight percentages of sand. The electrical of 9.87 × 10−20 m2 (100 nD). Although the particle size does not
conductivity is averaged over the values of the PC with two
particle sizes. play a significant role in its electrical conductivity, a coarser PC is
preferred to achieve a higher hydraulic conductivity. An economic
analysis suggests that when using 50% sand and 50% PC, the cost
Electrical conductivity (S/m) is only approximately 1.3 times that of sand. Hence, PC is an ex-
wt% of sand (closure stress > 2.07 × 107 Pa) Cost cellent candidate for EM fracture diagnostics, not only due to its
0 >4.87 × 103 1.52× high electrical and hydraulic conductivity when subject to high
closure stresses, but also because of its low cost and commercial
25 >3.27 × 103 1.42×
availability.
50 >1.67 × 103 1.30×

ACKNOWLEDGMENTS
favorable for proppant mapping with EM tools. However, for com-
mercial use in large quantities, the economics should be carefully This work was supported by DOE contract DE-SC0015986. The
considered and the mass percentage of PC should be designed to authors also would like to acknowledge G. Baum from the Univer-
suit specific needs. Assuming the unit price of PC is twice that sity of Texas at Austin for his assistance on the experiments.
of sand (based on an average current estimate), and the proppant
is pumped on a volume basis, the effective electrical conductivity
and cost with respect to different weight percentages of sand are DATA AND MATERIALS AVAILABILITY
listed in Table 4. The cost is a relative number regarding the actual
cost of pumping pure sand in a specific frac job. When using 50% Data associated with this research are available and can be ob-
sand and 50% PC, the cost is only approximately 1.3 times that of tained by contacting the corresponding author.

Downloaded from [Link]

by Fundação Oswaldo Cruz user
 E20 Zhang et al.

REFERENCES Shiriyev, J., Y. Brick, P. Zhang, A. E. Yilmaz, C. Torres-Verdín, M. M.

Sharma, T. Hosbach, M. A. Oerkfitz, and J. Gabelmann, 2018, Experi-
ments and simulations of a prototype triaxial electromagnetic induction
Archie, G. E., 1942, The electrical resistivity log as an aid in determining logging tool for open-hole hydraulic fracture diagnostics: Geophysics,
some reservoir characteristics: Transactions of the AIME, 146, 54–62, 83, no. 3, D73–D81, doi: 10.1190/geo2017-0354.1.
doi: 10.2118/942054-G. Sierra, J. R., J. D. Kaura, D. Gualtieri, G. Glasbergen, D. Sarker, and D.
Basu, S., and M. M. Sharma, 2014, A new method for fracture diagnostics Johnson, 2008, DTS monitoring of hydraulic fracturing: Experiences
using low frequency electromagnetic induction: Presented at the SPE and lessons learned: Presented at the SPE Annual Technical Conference
Hydraulic Fracturing Technology Conference. and Exhibition.
Berryman, J. G., and G. M. Hoversten, 2013, Modelling electrical conduc- Sookprasong, P. A., C. C. Gill, and R. S. Hurt, 2014, Lessons learned from
tivity for earth media with macroscopic fluid-filled fractures: Geophysical DAS and DTS in multicluster, multistage horizontal well fracturing:
Prospecting, 61, 471–493, doi: 10.1111/j.1365-2478.2012.01135.x. Interpretation of hydraulic fracture initiation and propagation through
Bhatnagar, A., 2016, Overcoming challenges in fracture stimulation through diagnostics: Presented at the IADC/SPE Asia Pacific Drilling Technology
advanced fracture diagnostics: Presented at the SPE Asia Pacific Hy- Conference.
draulic Fracturing Conference. Symington, W. A., R. D. Kaminsky, W. P. Meurer, G. A. Otten, M. M.
Cipolla, C. L., and C. A. Wright, 2000, State-of-the-art in hydraulic fracture Thomas, and J. D. Yeakel, 2010, ExxonMobil’s Electrofrac™ process
diagnostics: Presented at the SPE Asia Pacific Oil and Gas Conference for in situ oil shale conversion, in Oil Shale: A solution to the liquid fuel
and Exhibition. dilemma: American Chemical Society, 185–216.
Elahi, S. H., and B. Jafarpour, 2018, Dynamic fracture characterization from Tang, H., J. E. Killough, Z. Heidari, and Z. Sun, 2017, A new technique to
tracer-test and flow-rate data with ensemble Kalman filter: SPE Journal, characterize fracture density by use of neutron porosity logs enhanced by
23, 449–466, doi: 10.2118/189449-PA. electrically transported contrast agents: SPE Journal, 22, 1034–1045, doi:
Friehauf, K. E., 2009, Simulation and design of energized hydraulic frac- 10.2118/181509-PA.
tures: Doctoral dissertation, The University of Texas at Austin. Warpinski, N. R., 1996, Hydraulic fracture diagnostics: Journal of Petroleum
Gabelmann, J., M. A. Oerkfitz, T. D. Hosback, M. M. Sharma, A. Yilmaz, Technology, 48, 907–910, doi: 10.2118/36361-JPT.
and C. Torres-Verdin, 2020, Modular electrode tool for improved hy- Weiss, C. J., D. F. Aldridge, H. A. Knox, K. A. Schramm, and L. C. Bartel,
draulic fracture diagnostics: U.S. Patent 10,578,763. 2016, The direct-current response of electrically conducting fractures
Hoversten, G. M., M. Commer, E. Haber, and C. Schwarzbach, 2015, excited by a grounded current source: Geophysics, 81, no. 3, E201–E210,
Hydro-frac monitoring using ground time-domain electromagnetics: Geo- doi: 10.1190/geo2015-0262.1.
physical Prospecting, 63, 1508–1526, doi: 10.1111/1365-2478.12300. Wu, W., R. Russell, and M. M. Sharma, 2017, An experimental method to
LaBrecque, D., R. Brigham, J. Denison, L. Murdoch, W. Slack, Q. H. Liu, study the impact of fracturing fluids on fracture conductivity in hetero-
Y. Fang, J. Dai, Y. Hu, Z. Yu, A. Kleinhammes, P. Doyle, Y. Wu, and geneous shales: Presented at the Unconventional Resources Technology
M. Ahmadian, 2016, Remote imaging of proppants in hydraulic fracture Conference, 808–821.
networks using electromagnetic methods: Results of small-scale field Xu, S., X. M. Tang, C. Torres-Verdín, and Y. D. Su, 2018, Shear-wave
experiments: Presented at the SPE Hydraulic Fracturing Technology anisotropy in cracked rocks and its application to evaluate hydraulic
Conference. fracturing from acoustic logging data: 88th Annual International Meeting,
Ley-Cooper, A. Y., T. Munday, M. Gilfedder, T. Ibrahimi, D. Annetts, and SEG, Expanded Abstracts, 281–285, doi: 10.1190/segam2018-2972006
K. Cahill, 2015, Inversion of legacy airborne electromagnetic datasets to .1.
inform the hydrogeological understanding of the northern Eyre Peninsula, Yang, K., C. Torres-Verdín, and A. E. Yılmaz, 2015, Detection and quanti-
South Australia: Goyder Institute for Water Research Technical Report fication of three-dimensional hydraulic fractures with horizontal borehole
Series, 15, 50. resistivity measurements: IEEE Transactions on Geoscience and Remote
Mayerhofer, M. J., E. Lolon, N. R. Warpinski, C. L. Cipolla, D. W. Walser, Sensing, 53, 4605–4615, doi: 10.1109/TGRS.2015.2402656.
and C. M. Rightmire, 2010, What is stimulated reservoir volume? SPE Zhang, P., M. K. Sen, M. M. Sharma, J. Gabelmann, and D. Glowka, 2018a,
Production and Operations, 25, 89–98, doi: 10.2118/119890-PA. Modeling of low-frequency downhole electrical measurements for map-
Palisch, T., W. Al-Tailji, L. Bartel, C. Cannan, M. Czapski, and K. Lynch, ping proppant distribution in hydraulic fractures in casedhole wells: SPE
2016, Recent advancements in far-field proppant detection: Presented at Journal, 23, 2–147, doi: 10.2118/189884-PA.
the SPE Hydraulic Fracturing Technology Conference. Zhang, P., Y. Brick, and M. M. Sharma, 2018b, Numerical study of an elec-
Prats, M., 1961, Effect of vertical fractures on reservoir behavior- trode-based resistivity tool for fracture diagnostics in steel-cased well-
incompressible fluid case: SPE Journal, 1, 105–118, doi: 10.2118/1575-G. bores: Geophysics, 83, no. 2, D41–D48, doi: 10.1190/geo2017-0355.1.
Scott, M. P., R. L. Johnson, A. Datey, C. B. Vandenborn, and R. A. Wood- Zhang, P., J. Shiriyev, C. Torres-Verdín, M. M. Sharma, Y. Brick, J. Massey,
roof, 2010, Evaluating hydraulic fracture geometry from sonic anisotropy and A. Yilmaz, 2016, Fracture diagnostics using a low-frequency electro-
and radioactive tracer logs: Presented at the SPE Asia Pacific Oil and Gas magnetic induction method: Presented at the 50th US Rock Mechanics/
Conference and Exhibition. Geomechanics Symposium.
Sen, N., A. Kleinhammes, Y. Wu, M. O’callaghan, and M. Ahmadian-
Tehrani, 2018, Proppant additives for hydraulic fracturing: U. S. Patent
2018/0100389. Biographies and photographs of the authors are not available.

Downloaded from [Link]

by Fundação Oswaldo Cruz user