SPE-184820-MS

Calculating the Volume of Reservoir Investigated During a

Fracture-Injection/Falloff Test DFIT

David P. Craig and Robert A. Jackson, DFITpro.com

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
Unconventional reservoirs, including shale formations and many tight-gas sands, contain natural fractures,
fissures, faults, and microfractures that contribute to the rock flow capacity; thus, core-measured
permeability, and especially crushed-core permeability measurements, may not be representative of the true
reservoir rock flow capacity. Evaluating the in-situ rock permeability, and simulating production, requires
sampling not only the matrix, but also the fractures and fissures that contribute to the unconventional rock
system permeability. A properly designed well test in an unconventional reservoir will sample a volume of
reservoir rock that is representative of the whole. In other words, a well test should sample a representative
elementary volume, which is the smallest volume of rock with properties characteristic of the whole.
Diagnostic fracture-injection/falloff tests (DFIT) have been routinely implemented since the late 1990s
to understand leakoff mechanisms, identify fracture closure stress, estimate initial reservoir pressure,
and determine permeability-thickness in unconventional reservoirs. In almost every unconventional well
completed, a DFIT is the only well test that will be completed during the well lifecycle, but historically,
the tests have been designed empirically based on analog formations and without considering the volume
of rock investigated.
We demonstrate a new method for calculating the volume of rock investigated by a DFIT, and we show
how a DFIT design can allow for sampling a representative elementary volume of the reservoir.

Introduction
Diagnostic fracture-injection/falloff tests have become the standard prefrac well testing method in
unconventional wells, and with the introduction of well test type curve analysis (Craig and Blasingame
2006; Craig 2014), additional reservoir information, including fracture conductivity, fracture half-length,
fracture-face skin, choked-fracture skin, etc., can now be extracted from the falloff data.
Seldom considered in diagnostic fracture-injection/falloff test design is the area- and volume-of-
investigation observed during a test. In traditional well testing applications, the area- or volume-of-
investigation are easily included in a well test design and analysis (Lee 1982). Typically, either the radius-of-
investigation during radial flow, or the depth-of-investigation during linear flow, is calculated from the test
parameters or the data interpretation. With the radius-of-investigation or the depth-of-investigation known,
2 SPE-184820-MS

the area- or volume-of-investigation are calculated provided the formation thickness open to flow is also
known. The radius- and depth-of-investigation are both derived from an impulse solution, and represent
the maximum distance a pressure disturbance (transient) propagates through a reservoir during the time
of a test. In most cases, a diagnostic fracture-injection/falloff test can be described mathematically as an
"impulse" test (Craig 2014), and the same techniques used to calculate radius-, depth-, area-, and volume-
of-investigation for a well test can be used in DFIT analysis.
Diagnostic fracture-injection/falloff test design has largely been empirical based on analog formations
in other fields or based on completion requirements, such as the injection volume required to insure a
horizontal well toe port is open. Reservoir pressure estimates interpreted from DFIT data are relatively
insensitive to the volume of fluid injected during a DFIT provided the injection volume is sufficient
to sample reservoir outside of the near-wellbore region. Ideally, transmissibility, permeability-thickness,
and permeability interpretations would also be insensitive to the injected volume provided the reservoir
is homogeneous. Of course, in the world of unconventional resources, the reservoirs are anything but
homogeneous.
Most unconventional reservoirs are DIANEdiscontinuous, inhomogeneous, anisotropic, and nonelastic
(Jing and Stephansson 2007). Discontinuities such as fractures, fissures, faults, and even microfractures
can have a strong impact on reservoir rock permeability. Modeling production from such unconventional
reservoirs requires either adopting a pseudohomogeneity approach, which is a term coined by Ely, Brown,
and Reed (1995) to describe the "wide-spread homogenous appearance of a heterogeneous" naturally-
fractured Mesaverde reservoir. The pseudohomogeneity approach often assumes a continuum model is
adequate for production modeling; thus, production modeling is simplified and conventional rate-transient
analysis can be applied. Alternatively, one can model a DIANE reservoir rock using a discontinuum
modeling approach, which requires the development of a discrete-fracture network model and stochastic
simulation techniques. Both continuum and discontinuum modeling are valuable and have a place in
understanding unconventional reservoir production. DFIT interpretations can supplement both modeling
techniques; however, here we limit our concerns to developing DFIT interpretations for the continuum
modeling approach.
A diagnostic fracture-injection/falloff test is a well test. It is a well test that can be used to define
permeability in an unconventional reservoir and constrain permeability in production models; however, like
any other well test, a DFIT must be designed correctly to sample a reservoir volume that is representative of
the drainage volume of a well, and the data must be interpreted correctly and preferably using a well-testing
methodology (Craig and Blasingame 2006; Craig 2014). Some, including Wallace, Kabir, and Cipolla
(2014), have argued DFIT interpretations can under- or overestimate permeability and well productivity.
It is undeniable that well tests in unconventional reservoirs can misinform via nonuniqueness issues and
ambigious reservoir signals, but recognizing DFIT and well testing limitations, a testing program can be
designed to minimize uncertainty and provide representative transmissibility, permeability-thickness, and
permeability estimates for production modeling.
In this paper, we will examine the necessity of defining a representative elementary volume in
unconventional reservoirs such that a diagnostic fracture-injection/falloff test can be designed and
implemented to sample a volume of reservoir characteristic of the whole. We will also demonstrate how the
area- and volume-of-investigation can be calculated from DFIT falloff data using standard well test analysis
impulse solutions, and we will provide examples illustrating the calculations. Finally, we will show how
the area- or volume-of-investigation can be included in a diagnostic fracture-injection/falloff test design.

Representative Elementary Volume

In general, a representative elementary volume is the smallest volume with properties statistically equivalent
and representative of the whole. Although a representative elementary volume can be defined for any
SPE-184820-MS 3

variable or property, here we are interested in finding a permeability of a volume of reservoir that is
characteristic of the much larger whole reservoir containing discontinuities, including natural fractures,
faults, fissures, and microfractures. Scale is important. Proofs of the validity of a representative elementary
volume generally attempt to relate the microscopic reservoir properties to the macroscopic reservoir (Bear
and Bachmat 1990; Jing and Stephansson 2007). For example, a proof of a valid representative elementary
volume as presented by Bear and Bachmat (1990), is based on a naturally-fractured rock idealization where
the representative elementary volume of the matrix and a representative elementary volume of the natural
fractures are used to identify the representative elementary volume of the matrix and fractures. Figure 1 is a
reproduction from Bear and Bachmat (1990), and it shows a plot of measured porosity verusus measurement
volume. At very small measurement volumes, discrete grains can dominate the porosity measurement, and
the porosity measurements will be widely scattered. Of course, the same is true for the natural fractures,
where at small measurement volumes, the fracture porosity measurements can vary widely.

Figure 1 Illustration of representative elementary volume for rock matrix and rock natural fractures.

As the measurement volume increases, a volume region will be observed where the measured property
will stabilize. The beginning of stabilization for each property corresponds to be minimum representative
elementary volume for the property, and the representative elementary volume for the whole rock
idealization, including natural fractures and matrix, is the region of overlap of the matrix porosity and
natural-fracture porosity stabilized regions denoted (Vi)matrix and (Vi)fracture, respectively in Figure 1. Although
not illustrated, the representative elementary volume can also have an upper limit where matrix or natural-
fractue heterogeneity begin to reintroduce variability into the measurements. As noted by Bear and Bachmat
(1990), it's also possible that the stabilized regions will not overlap; consequently, it is not always possible
to identify a representative elementary volume.
While the porosity of the matrix and natural fractures are important variables in unconventional reservoir
numerical modeling, our concern here is primarily the development of a representative elementary volume
for a reservoir permeability that includes both matrix and natural fractures. Additionally, our focus will be
on increasing the scale of a representative elementary volume to that describing the drainage volume of a
typical horizontal well in an unconventional reservoir.
As previously noted, many unconventional reservoirs are DIANE discontinuous, inhomogeneous,
anisotropic, and non-elastic (Jing and Stephansson 2007). Within a DIANE unconventional reservoir,
there is a scale of discontinuities that is often divided into major, intermediate, and micro discontinuities.
Discontinuities of all scales will affect fluid flow through the reservoir, and can be modeled in reservoir
4 SPE-184820-MS

simulators. Micro-scale discontinuities, like microfractures, are often assumed to be part of the rock matrix,
and the rock matrix is modeled as an equivalent continuum. Intermediate- and large scale discontinuities
can be explicitly modeled in discrete-fracture network simulations, but the simulations are computationally
demanding, slow, and often very poorly defined and constrained by measured data.
For example, a standard method for developing probability distributions of natural fracture properties
used in generating discrete-fracture network models is the mapping of fractures observed in outcrops.
Recently, Grechishnikova (2016) presented the development of a discrete-fracture network model for
the Niobrara formation of the Denver basin, which required mapping fractures exposed during a mining
operation of the Niobrara outcrop. Grechishnikova identified four discrete fracture sets in the Niobrara, and
defined probability distributions of fracture-set orientation, fracture-set intensity (density), and fracture-set
height. Since fracture aperture could not be accurately mapped, it was assumed constant, and fracture-set
trace length, which was also not accurately measured, is variable, but based on a constant height/length
aspect ratio. Figure 2, which is reproduced from Grechishnikova (2016), shows a single realization of two
of the four fracture sets in the Niobrara formation.

Figure 2 Niobrara discrete-fracture network realization after Grechishnikova (2016).

Simulating the Niobrara using the explicit discrete-fracture network model is possible, but
computationally demanding and slow. Another approach would be to develop simulation grid blocks for
the discrete-fracture network as representative elementary volumes where equivalent fracture properties,
including fracture porosity, effective permeability, and grid block size, could be generated from the model
(Decroux and Gosselin 2013). However, both approaches would only be applicable to the single discrete-
fracture network model shown above. Realistic simulation would require simulating many realizations to
understand the range of possible outcomes from other statistically equivalent discrete-fracture network
models.
An alternative, and certainly much more simplistic, continuum approach for simulating production
requires developing a permeability representative elementary volume of the drainage volume of the well. In
other words, a permeability is measured at a point along a lateral, and that permeability is used to describe the
SPE-184820-MS 5

reservoir along the entire horizontal well. Figure 3 shows an idealized horizontal well with five transverse
DFIT fractures illustrating the area-of-investigation of each diagnostic fracture-injection/falloff test.

Figure 3 Illustration of DFIT area-of-investigation for five tests along a horizontal well in a naturally-fractured reservoir.

In Figure 3, the first DFIT, labelled "1", represents a small volume injection with a relatively short
shut-in time. The pink shaded area represents an elliptical area-of-investigation of the test, and it suggests
the test would have contacted predominately rock matrix and few natural fractures. Consequently, the
interpretations of DFIT "1" would not have been representative of the entire horizontal well drainage
volume. The second DFIT illustrates a larger volume injection, but with a shorter falloff time. Although the
area-of-investigation of DFIT "2" is larger, the depth-of-investigation into the reservoir remains too small
to be representative of the entire lateral. DFIT "3" is the largest volume injection with the longest shut-
in time, and contacts a representative sample of both rock matrix and fractures. Similarly, DFIT "4" and
DFIT "5" sample a representative area-of-investigation with a shorter shut-in time than DFIT "3". It should
also be obvious that with the distribution of natural fractures along the horizontal wellbore, permeability
will vary along the lateral, but provided the area- or volume-of-investigation are sufficient, the interpreted
permeability should be representive of the drainage volume of the horizontal well after hydraulic fracturing
along the entire lateral and beginning production.
In summary, the goal of a diagnostic fracture-injection/falloff test should be to investigate a volume of
the reservoir representative of the drainage volume of the well in order that the permeability interpretation
can be used to model production from the entire multifractured horizontal well.

Calculating DFIT Volume-of-Investigation

Lee (1982) defined the radius-of-investigation during radial flow from an impuse solution in an infinite
reservoir as follows

where ri is the radius-of-investigation, k is the permeability, t is time at the end of the radial flow period,
is porosity, is viscosity, and ct is total compressibility. Since diagnostic fracture-injection/falloff tests
essentially consist of an instantaneous injection into the reservoir, the same solution can be applied to the
radial flow period of a DFIT provided pseudoradial flow is observed. The radial flow area-of-investigation,
Ai, is simply, , and the volume-of-investigation, Vi, is Vi = hAi, where h is the formation thickness
6 SPE-184820-MS

open to flow during the falloff. It's noteworthy that the volume-of-investigation during radial flow is not a
function of injection volume, that is, the injection volume doesn't appear in any of the equations.
A derivation of the radius-of-investigation assumes radial flow from the initial impulse (Lee 1982). When
a hydraulic fracture is created during the injection, and especially if a long hydraulic fracture is created
during the injection, the flow will transition from linear to elliptical to radial after the injection (impulse);
thus, the radial flow assumption from the beginning of the pressure impulse and following through the
falloff will not always be valid.
In unconventional reservoirs, radial or pseudoradial flow during a diagnostic fracture-injection/falloff
test are rarely observed. Fortunately, an impulse solution for the linear, or pseudolinear, flow period has
been derived by Behmanesh, Clarkson, Tabatabaie et al (2015), and can be used to estimate the depth-of-
investigation during the linear flow period as follows

(2)

where yi is the depth-of-investigation from the the fracture plane, and t is the time at the end of the linear
flow period or the last recorded time during the linear flow period. Figure 4 shows the geometry for the
linear flow system for either horizontal or vertical wells. After determining the depth-of-investigation, the
area-of-investigation (rectangle) is calculated as
(3)
where Lf is the fracture half-length.

Figure 4 Linear flow geometry during a diagnostic fracture-injection/falloff test in either a horizontal or vertical well.

Conventional before- and after-closure fracture-injection/falloff methodologies do not provide a suitable

means for calculating the fracture half-length open to flow during the falloff; however, the well-test analysis
methodology described by Craig and Blasingame (2006) and Craig (2014) allows the fracture half-length
to be determined from a type curve and a dimensionless time match point. Fracture half-length is calculated
from a dimensionless time match point using the following relationship
SPE-184820-MS 7

(4)

where [t/tLfD]M is the time/dimensionless-time match point. With the fracture half-length known, the area-
of-investigation can be calculated using Eq. (3).
As in the case of radial flow, the volume-of-investigation, Vi, is Vi = hAi, where h is the formation thickness
open to flow during the falloff. Unfortunately, the formation thickness open to flow during the falloff has
the most uncertainty of all the DFIT calculated parameters. In some cases, especially with clear lithologic
changes, high-stress bounding layers, low injection rates, and with an injection volume sufficient to open
the entire formation thickness, the fracture height can be assumed to be equal to the formation thickness.
Fracture propagation models, whether 2D-radial or 3D, can also be used to verify whether the DFIT injection
opened the entire formation thickness; however, fracture-height recession during fracture closure can change
the thickness open to flow, and the formation thickness used in volume-of-investigation calculations can
remain highly uncertain.

DFIT Volume-of-Investigation Examples

Three examples in the same field are used to illustrate the volume-of-investigation calculations and the
Bear and Bachmat (1990) methodology for empirically determining a representative elementary volume
from diagnostic fracture-injection/tests. The three examples are from the same formation in the same field,
and the areal spacing is illustrated in Figure 5, which shows the wells are located along a near-diagonal
line spanning 4.7 miles and covering an area of about 10 square miles. While there is some thinning of
the formation along the southwest to northeast diagonal line, all other reservoir engineering parameters are
very similar.

Figure 5 DFIT relative well locations. The northernmost well

was vertical, and the southern wells were horizontal with toe-DFIT.

The initial DFIT was pumped in the northernmost vertical well in 2003, and it consisted of a
relatively small injection of 15-bbl of treated water. Figure 6 contains the DFIT semilog derivative of the
fracture-pressure difference, td(pe-pws(t))dt, matched to the negative product of dimensionless time and the
second derivative of the constant-rate dimensionless pressure solution with respect to dimensionless time,
(-)tLfDd2pwcd/dt2pwcD/dt2LfD, for CD= 0.02 and (Sf)ch = 0.0. From a type-curve match point, the initial reservoir
8 SPE-184820-MS

pressure can be estimated, and the integrated reservoir-pressure difference, , and derivative,
t(pws(t)-pi), can be calculated from the observed DFIT data. The integrated reservoir- pressure difference
and derivative can then be matched to dimesionless pressure, pwcD, and dimensionless pressure derivative,
tLfDdpwcd/dtLfD, type curves for flow from an infinite-conductivity fracture with CD = 0.02 and (Sf)ch = 0.0
(Craig 2014). The transmissibility from the match point is kh/ = 1.750 mdft/cp, and with a reservoir fluid
viscosity, = 0.128 cp, the permeability-thickness is 0.224 mdft. The tested formation is bounded by clear
lithologic changes above and below the reservoir, so the formation thickness open to flow is estimated to
be the entire 10 ft interval.

Figure 6 Northernmost vertical well DFIT data type curve match.

Figure 6 also shows that linear flow was observed during and through the end of the falloff, so Eq. (2)
can be used to calculate the depth of investigation, which is yi = 20 ft after 20.4 hr. Using a time match point
and Eq. (4), the fracture half-length open to flow is estimated to be, Lf = 102 ft; consequently, the area-of-
investigation is, Ai = 8,038 ft2, and the volume-of-investigation is, Vi = 80,382 ft3.
The type curve matches of the two southern horizontal well DFIT are shown in Figures 7 and 8. In Figure
7, which is the southernmost well, the injection volume was 48 bbl of treated water, and the falloff was
monitored for 256.2 hr. Following the same Craig (2014) procedure, the permeability is estimated to be k
= 0.038 md, and the depth-of-investigation calculated from Eq. (2) is yi = 94 ft. From the time match point
and Eq. (4), the fracture half-length open to flow is Lf = 157 ft, and the area-of-investigation is, Ai = 58,714
ft2, which results in a volume-of-investigation of Vi = 1,232,999 ft3.
SPE-184820-MS 9

Figure 7 Southernmost horizontal well DFIT data type curve match.

Figure 8 Center horizontal well DFIT data type curve match.

The largest volume injection, Vinj = 88 bbl, was in the centermost well, which resulted in a permeability
estimate of k = 0.026 md, depth-of-investigation of yi = 77 ft, and a fracture half-length open to flow of Lf
= 215 ft. The area-of-investigation is Ai = 66,220 ft2, and the volume-of-investigation is Vi = 1,556,170 ft3.
Table 1 is included to summarize the results of the DFIT interpretations.

Table 1 Summary of DFIT injection volumes and interpretations.

Well Injection Volume (bbl) Shutin Time (hr) Permeability (md) Depth-of- Volume-of-
Investigation (ft) Investigation (ft3)

Vertical 15 20 0.022 20 80,382

Center Horizontal 88 236 0.026 77 1,556,170

Southern Horizontal 48 256 0.038 94 1,232,999

10 SPE-184820-MS

Intuitively, all the interpretations appear consistent, that is, the fracture half-length increases with the
volume of the injection, the depth-of-investigation increases with permeability and the shut-in time, and
the volume-of-investigation increases as the injection volume, permeability, and shut-in time increase.
Identifying or validating a representative elementary volume requires examining the variability of the
interpretations with respect to the area- and volume-of-investigation.
Figures 9 and 10 show graphs of permeability versus area-of-investigation and permeability-thickness
versus volume-of-investigation, respectively. Neither graph shows a convincing stabilization as required
for empirically defining a representative elementary volume as described by Bear and Bachmat (1990), but
the interpretations do illustrate the process. Ideally, a few additional measurements within the 10 square
mile area around the three test wells could define permeability or permeability-thickness stabilization
with respect to the volume-of-investigation, and a representative elementary volume could be established.
However, no additional DFIT were planned or executed in the area.

Figure 9 DFIT permeability interpretation versus the area-of-investigation.

Figure 10 DFIT permeability-thickness interpretation versus the volume-of-investigation.

SPE-184820-MS 11

DFIT Design Accounting for Volume-of-Investigation

Diagnostic fracture-injection/falloff test design, like conventional well test design, requires knowing or
estimating parameters that are the desired output of the test being designed, which obviously can make
a meaningful design difficult. In general, the design workflow proceeds backwards from the volume-of-
investigation to the radius- or depth-of-investigation. In a homogeneous reservoir, where a small-volume
DFIT is used simply to create an impulse and radial flow is anticipated throughout the test, the design
process is as follows.

Determine the desired volume-of-investigation. If the reservoir is homogeneous, the volume-of-

investigation may not be critical.
Calculate the area-of-investigation based the formation thickness open to flow during the test, Ai
= Vi/h. If the entire formation is opened to flow by the injection, then h is simply the formation
thickness. If not, the formation thickness open to flow must be estimated using lithologic changes,
boundaries, or modeling the fracture height created during propagation.
The radius-of-investigation is calculated from the area-of-investigation, .
With the radius-of-investigation known, the shut-in time can be estimated from Eq. (1), which can
be written as
.

Eq. (1) shows that the permeability must be estimated, along with other typical reservoir properties, to
determine the required shut-in time for the desired volume-of-investigation.
A volume-of-investigation diagnostic fracture-injection/falloff design where linear flow is anticipated is
somewhat more difficult because of the uncertainty in estimating the required variables. The process is as
follows.

Determine the desired volume-of-investigation. The volume-of-investigation can be determined

empirically from a planned variation in injection volumes, like the example problems above, or
the volume-of-investigation can be designed based on a discrete-fracture network model and the
calculated representative elementary volume.
Calculate the area-of-investigation based the formation thickness open to flow during the test, Ai
= Vi/h. If the entire formation is opened to flow by the injection, then h is simply the formation
thickness. If not, the formation thickness open to flow must be estimated using lithologic changes,
boundaries, or modeling the fracture height created during propagation. As previously noted, the
formation thickness open to flow, h, can be highly uncertain.
The depth-of-investigation is calculated from the area-of-investigation and the fracture half-length
open to flow during the falloff, yi = Ai/4Lf. The fracture half-length open to flow during the falloff
also creates massive uncertainty in the design. Even if the best fracture propagation models are
used to model created fracture half-length, the fracture half-length open to flow may not be known
until the data is analyzed. Often the created fracture half-length and the fracture half-length open to
flow during the falloff are not the same value. For design purposes, it may be sufficient to assume
a model created fracture half-length with a safety-factor less than 1, which will result in a fracture
half-length open to flow during the falloff that is less than the model created fracture half-length.
With the depth-of-investigation known, the shut-in time can be estimated from Eq. (2), which can
be written as

.
12 SPE-184820-MS

Eq. (2) shows that the permeability must be estimated, along with other typical reservoir
properties, to determine the required shut-in time for the desired volume-of-investigation during
linear flow. Not only will the permeability introduce uncertainty into the design, but the fracture
half-length open to flow in the previous step means that the depth-of-investigation calculation is
also highly uncertain.
The last two steps are also iterative as different injection volumes can result in different fracture
half-lengths; consequently, there will be a trade-off between injection volume, fracture half-length,
and shut-in time to achieve the desired area-of-investigation.
Experience in a field can help reduce the uncertainty, but the empirical approach of finding the volume-
of-investigation from different size injections and defining a representative elementary volume may be the
preferred methodology as opposed to design modeling.

Conclusions
Based on the techniques defined and discussed above, we offer the following conclusions.

A permeability representative elementary volume in unconventional reservoirs should be defined

such that a permeability measurement will be characteristic of the drainage volume of the well,
and a diagnostic fracture-injection/falloff test should be designed to investigate the permeability
representative elementary volume.
The area- and volume-of-investigation can be calculated from DFIT falloff data using standard
well test analysis impulse solutions.
The fracture half-length open to flow during the falloff can be calculated from a dimensionless
time match point of the integrated reservoir-pressure difference or derivative with a dimensionless
pressure or dimensionless pressure derivative type curve.
Diagnostic fracture-injection/falloff tests can be designed considering area- or volume-of-
investigation requirements.

Nomenclature
Ai = area-of-investigation, m2, ft2
CD = dimensionless storage coefficient, dimensionless
ct = total compressibility, Lt2/m, psi-1
h = formation thickness open to flow, L, ft
k = permeability, L2, md
Lf = fracture half-length, L, ft
pe = bottomhole pressure at the end of pumping, m/Lt2, psi
pi = initial reservoir pressure, m/Lt2, psi
pws = bottomhole pressure during shut-in, m/Lt2, psi
pwcD = dimensionless pressure of constant-rate solution, dimensionless
ri = radius-of-investigation, L, ft
(Sf)ch = choked-fracture skin, dimensionless
t = time, t, hr
tLfD = dimensionless time of fracture solution, dimensionless
Vi = volume-of-investigation, m3, ft3
(Vi)fracture = volume-of-investigation of fracture sample, m3, ft3
(Vi)matrix = volume-of-investigation of matrix sample, m3, ft3
SPE-184820-MS 13

Vinj = volume of injection, m3, bbl

yi = depth-of-investigation, m, ft

Greek
= porosity, dimesionless
fracture = porosity of natural fracture sample, dimesionless
matrix = porosity of matrix sample, dimesionless
= viscosity, m/Lt, cp

References
Bear, J. and Bachmat, Y. 1990. Introduction to Modeling of Transport Phenomena in Porous Media. Kluwer Academic
Publishers, Dordrecht, The Netherlands.
Behmanesh, H., Clarkson, C.R., Tabatabaie, S.H., Sureshjani, M. Heidari 2015. Impact of Distance-of-Investigation
Calculations on Rate-Transient Analysis of Unconventional Gas and Light-Oil Reservoirs: New Formulations for
Linear Flow. Journal of Candian Petroleum Technology 54 (06): 509-519. November 2015. SPE-178928-PA. http://
dx.doi.org/10.2118/178928-PA.
Craig, D.P. 2014. New Type Curve Analysis Removes Limitations of Conventional After-Closure Analysis of DFIT Data.
Presented at the SPE Unconventional Resources Conference, The Woodlands, Texas, USA, 1-3 April. SPE-168988-
MS. http://dx.doi.org/10.2118/168988-MS.
Craig, D.P. and Blasingame, T.A. 2006. Application of a New Fracture-Injection/Falloff Model Accounting for
Propagating, Dilated, and Closing Hydraulic Fractures. Presented at the SPE Gas Technology Symposium, Calgary,
Alberta, Canada, 15-17 May. SPE-100578-MS. http://dx.doi.org/10.2118/100578-MS.
Decroux, B. and Gosselin, O. 2013. Computation of Effective Dynamic Properties of Naturally Fractured
Reservoirs: Comparison and Validation of Methods. Presented at the EAGE Annual Technical Conference &
Exhibition incorporating SPE Europec, London, United Kingdom, 10-13 June 2014. SPE-164846-MS. http://
dx.doi.org/10.2118/164846-MS.
Ely, J.W., Brown, T.D., and Reed, S.D. 1995. Optimization Of Hydraulic Fracture Treatments in The Williams Fork
Formation Of The Mesaverde Group. Presented at the SPE Low Permeability Reservoirs Symposium, Denver,
Colorado, USA, 19-22 March. SPE-29551-MS. http://dx.doi.org/10.2118/29551-MS.
Grechishnikova, A. 2016. Discrete Fracture Network Model Developed from a High Resolution LIDAR Outcrop Survey of
a Naturally Fractured Unconventional Niobrara Reservoir, Denver Basin. Presented at the Unconventional Resources
Technology Conference, San Antonio, Texas, USA, 1-3 August 2016. http://dx.doi.org/10.15530/urtec-2016-2430655.
Jing, L. and Stephansson, O. 2007. Fundamentals of Discrete Element Methods for Rock Engineering Theory and
Applications. Elsevier, Amsterdam, The Netherlands.
Lee, W.J. 1982, Well Testing. Society of Petroleum Engineers of AIME, Dallas, Texas.
Wallace, J., Kabir, C.S., and Cipolla, C. 2014. Multiphysics Investigation of Diagnostic Fracture Injection Tests in
Unconventional Reservoirs. Presented at the SPE Hydraulic Fracturing Technology Conference, The Woodlands,
Texas, USA, 4-6 February 2014. SPE-168620-MS. http://dx.doi.org/10.2118/168620-MS.