You are on page 1of 9

ARMA 13-666

Geomechanics Considerations for Hydraulic Fracture


Productivity
Suarez-Rivera R., Burghardt J., Edelman E., Stanchits S.,
Schlumberger Innovation Center, Salt Lake City, Utah, USA
Surdi A.
TerraTek, a Schlumberger company, Salt Lake City, Utah, USA

Copyright 2013 ARMA, American Rock Mechanics Association


This paper was prepared for presentation at the 47th US Rock Mechanics / Geomechanics Symposium held in San Francisco, CA, USA, 23-26
June 2013.
This paper was selected for presentation at the symposium by an ARMA Technical Program Committee based on a technical and critical review of
the paper by a minimum of two technical reviewers. The material, as presented, does not necessarily reflect any position of ARMA, its officers, or
members. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of ARMA
is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 200 words; illustrations may not be copied. The
abstract must contain conspicuous acknowledgement of where and by whom the paper was presented.

ABSTRACT: Well productivity in unconventional reservoirs depends on contacting good reservoir quality rock with sufficient
surface area, via hydraulic fracturing, and more importantly on preserving surface area and fracture conductivity over long-term
production. Thus, understanding the created fracture system is fundamental for understanding well production. In this paper we
present fracture propagation experiments on texturally complex, organic-rich, tight shale outcrop blocks of the Niobrara formation.
After hydraulically fracturing the block we conducted unpropped fracture conductivity measurements under changing closure
stress. Subsequently we re-fractured the block with proppant laden slick water. When the desired amount of sand was transported
and just before screening out, we stopped proppant injection and conducted fracture conductivity measurements on the propped
fracture. These measurements of propped and unpropped fracture conductivity were compared to corresponding measurements on
small samples. Results show that the larger fractures, whether propped or unpropped, retain higher fracture conductivity with
increasing stress. We also observe substantial fracture complexity associated with the textural complexity in the rock. As a result
there is substantial variability in the concentration and distribution of the proppant in the fracture. Post-test analysis allowed us to
describe the fractured system as consisting of four easily identifiable regions: the wellbore, the connection between the wellbore
and the fracture system, the near-wellbore fracture, and the far-wellbore fracture network. Understanding the role of each of these
regions to hydrocarbon production helps identifying the often competing causes of production decline over time. Results show that
the requirements for optimizing hydrocarbon flow in these four regions are different, but reasonable compromises can be made to
sustain production from the overall fracture system.

Fracture complexity results primarily from complex rock


1. INTRODUCTION
fabric, which is defined by the presence, density,
Achieving economic production from nano-Darcy mechanical strength, and orientation (in relation of the
permeability, organic-rich, mudstone reservoirs requires in-situ stress) of planes of weakness in the rock. These
creating a large surface area by hydraulic fracturing, and may be associated with bedding, with particular
then sustaining adequate fracture conductivity during depositional interfaces, with the presence of mineralized
long-term production. In general, fracture complexity, or organic-filled fractures, or others. The end result is
including fracture branching, and the presence of that the rock fabric defines alternative paths of weakness
fracture connectors and step-overs, increases the surface for fracture propagation, and sources for fracture arrest
area per unit reservoir volume, which is a desirable goal. and branching.
Unfortunately, however, fracture complexity results also
Low surface hardness, and degradation of fracture
in poor proppant delivery and non-uniform placement.
surface hardness by fluid-rock interactions, contribute to
Proppant laden fluids that may move readily through
the loss of surface area and fracture conductivity. The
single fracture systems have to overcome the tortuosity
effect is most important in regions with low proppant
and narrowing at fracture connectors, and the presence
coverage. In contrast, high surface hardness reduces the
of shear-dominated step-overs, which are often
embedment between unpropped fracture faces, reduces
unfavorably oriented in relation to the in-situ stress.
proppant embedment, and helps preserve fracture
This results in poor proppant coverage, significant
conductivity. Fines migration, salt dissolution and
unpropped fracture regions, and an overall loss of
precipitation, and loss of fracture-face permeability (e.g.,
surface area and fracture conductivity that causes loss of
by water blocking or plastic deformation) are also
well productivity over time.
detrimental to hydrocarbon production and flow. All
these effects need to be understood and managed to indicate ineffective retained hydraulic continuity,
optimize stage production. and often zero indication of a sustained hydraulic
connection. This indicates that the fractures are
A recent extensive survey on current operational
rapidly losing fracture conductivity or are collapsing
experience in tight shale reservoirs [1], indicated that
at some point between the wellbores.
retention of productive fracture area and fracture
conductivity was a major economic issues in tight shale The above report indicates that the industry is
plays. This report indicates that while many production experiencing loss of fracture conductivity and loss of
trends fail to yield a unique interpretation as to the fracture surface area. It also indicates that there is no
damage mechanism, there are a number of observations comprehensive framework for understanding this
that, when considered in conjunction, compellingly problem or its various causes. The objective of this
demonstrate that fractures are not recovering the work was thus to understand the multiple causes of loss
hydrocarbon resource as effectively as commonly of fracture area and fracture conductivity, and define
anticipated. These were reported to be the following: solutions to mitigate the resulting loss of production.
For this, we conducted a large set of laboratory
Infill drilling: New wells drilled through a
experiments of fracture conductivity and hydraulic
microseismic swarm are encountering near-virgin
fracturing on small (1 inch diameter by 1 inch
reservoir pressure and appear to recover in excess of
length) to large block samples (3 ft x 3ft x 3ft) and
80% of the EUR of parent wells drilled in virgin
evaluated loss of fracture conductivity under various
rock. This indicates original parent wells have not
conditions of stress, fluid types, with and without
depleted the rock volume apparently covered with
proppant, changing proppant types, as well as
microseismic swarms.
investigating the effect of creep, temperature and sample
Re-stimulation: Horizontal wells in most resource
size. Results on the small-scale samples were reported
plays have been successfully re-stimulated,
elsewhere [2]. Results of the large-scale samples are
indicating that the initial fracture treatments were
presented here. These consist of measurements of
insufficient to drain the recoverable reserves
unpropped fracture conductivity as a function of closure
accessible from the wellbore.
stress and, after re-fracturing the block with proppant
Steep decline curves: These may be attributed to
laden slick water, repeating the fracture conductivity
collapse of fracture conductivity and/or surface area.
measurements on the propped fracture. Results show
Drawdown management: In some reservoirs,
that fractures in the larger samples, whether propped or
drawdown management seems to extend EUR,
unpropped, retain higher fracture conductivity with
suggesting that typical operational practices may be
increasing stress than the smaller samples. We also
accelerating fracture degradation.
observe substantial fracture complexity associated with
Laboratory testing: Even in ideal conditions in
the textural complexity in the rock, and as a result, a
which proppant is arranged in wide packs with
substantial variability in the concentration and
optimal packing arrangements; tremendous
distribution of the proppant in the fracture. In addition,
degradation of conductivity over time has been
post-test analysis allowed us to describe the fractured
noted in every published extended-duration
system as consisting of four easily identifiable regions:
laboratory test. In the lab, proppants continue to
the wellbore, the connection between the wellbore and
crush and compact over time. It should be logically
the fracture system, the near-wellbore fracture, and the
anticipated that actual fractures will similarly lose
far-wellbore fracture network. Understanding the role of
conductivity over time.
each of these regions to hydrocarbon production helps in
Fracture complexity: In numerous mine-back and
identifying the often competing causes of production
core-through studies it has been apparent that
decline over time. Results also show that the
fractures are not simple, wide, planar features. It is
requirements for optimizing hydrocarbon flow in these
challenging to place a uniformly wide fracture that
four regions are different, but reasonable compromises
retains hydraulic continuity and conductivity
can be made to sustain production from the overall
through highly heterogeneous or laminated intervals.
fracture system.
Cases in which hydraulic fractures connect two
wellbores: In a number of reservoirs, proppant has 2. ROCK FABRIC AND PLANES OF WEAKNESS
been pumped to the surface of adjacent wells
Observations on core, outcrops, mine back experiments
completed at exactly the same depth, irrefutably
and laboratory experiments of hydraulic fracturing
indicating that a fracture was created between the
indicate that rock fabric - and particularly the
two wells and proppant was transported through the
distribution and orientation of weak interfaces - has a
entire fracture length. However, subsequent
strong effect on fracture propagation, fracture geometry,
diagnostic tests (pulse tests, interference tests,
and containment [3-9]. Laboratory experiments of
calculation of estimated ultimate recovery) typically
fracture propagation conducted in thick glass plates with
glued interfaces [10] were used to show this behavior. only possible, however, because of the presence of
Fracture propagation on thick walled cylinders cut from planes of weakness. In the absence of these, the fracture
organic-rich mudstones with mineralized fractures also geometry in both cases would be bi-winged and planar.
demonstrated the dominant effect of weak interfaces on The lower images represent materials with planes of
fracture geometry. Most importantly, both sets of weakness oriented parallel and perpendicular to the
experiments demonstrated that for most conditions of maximum in-plane stress. Here, the stress contrast is
stress contrast and interface orientation, the interaction high in both cases. When the orientation of the planes of
of the propagating fracture and the interface created a weakness is parallel to the direction of maximum in-
stepover: a localized shear displacement. These were plane stress, the fracture follows this direction. When
longest for low stress contrast conditions and smallest the orientation of the planes of weakness is
for high stress contrast conditions, but were always perpendicular to the direction of maximum in-plane
present. In the extreme case, when there was no stress stress, the fracture follows the direction of the stress, but
contrast, or when the interface was very weak, once the it propagates in jumps, from plane of weakness to plane
fracture intersected the interface, it was temporarily of weakness, often shearing the interface during the
arrested and then propagated along the interface to the process. This fish-bone fracture geometry is observed
end of the sample. Figure 1 shows this fundamental regularly in blocks tests. However, it does not occur at
behavior. every interface.

Figure 1. Laboratory experiments of fracture propagation on


thick glass plates with glued interfaces. Fractures branch or
develop stepovers when intersecting the interfaces. Left: Low
stress contrast. Right: High stress contrast
Our laboratory experiments on medium-sized (1 ft x 1 ft
x 1.5 ft) and large (3 ft x 3 ft x 3 ft) blocks in Figure 2. Fracture propagation on a material with planes of
heterogeneous rocks indicate that propagating fractures weakness oriented oblique, parallel and perpendicular to the
are temporarily arrested at weak interfaces. This effect direction of maximum in-plane stress. The top-left image and
was also observed via acoustic emission monitoring. In the two lower images show the case for high stress contrast.
most cases, the approach of the fracture into the interface The top-right image shows the case of no stress contrast.
locally lowers the normal stress across the interface and The conceptual understanding of the effect of rock fabric
promotes a local shear slip [11-13]. From there, and on fracture geometry allows us to pose relevant
after developing the required pressure or strain energy questions for completion efficiency. For example: Are
for the next jump, they continue along a direction that some interfaces more relevant than others? How do we
satisfies the local path of least resistance. This may be identify these? Can we use our understanding of the
along or across the weak interface. This process of geologic system (e.g., sequence stratigraphy) to predict
fracture arrest and change of direction in fracture the distribution of weak interfaces? These are questions
propagation is only possible because of the presence of that should be asked for improved fracture efficiency
rock fabric (i.e., planes of weakness) [10]. and increased production.
Figure 2 shows conceptually the consequences of planes 3. LABORATORY TESTING
of weakness on fracture propagation. Two sets of
images are shown. The upper two represent materials Laboratory experiments of hydraulic fracture
with planes of weakness oriented oblique to the propagation in heterogeneous media are challenging to
maximum in-plane stress. Two conditions of fracture conduct and analyze because of internal boundary effects
propagation are described here: high stress contrast (left) during fracture initiation and external boundary
and low stress contrast (right). Under the same conditions as the fracture approaches the end of the
conditions of rock fabric and strength of the planes of specimen. Fracture initiation effects are minimized by
weakness, we observe a dramatic change in fracture cutting slots along the desired section of the wellbore;
geometry due to changes in the stress contrast. This is outer boundary effects are minimized by extending the
sample size. Thus, large-block samples provide the best fracture conductivity while changing the closure stress,
opportunity for evaluating hydraulic fracturing and then we re-fractured the block with proppant laden
propagation and the interaction of fractures with planes slickwater and measured the propped fracture
of weakness in the rock. In our studies, we use a conductivity in a similar fashion. A planar fracture with
polyaxial stress frame (Figure 3) with independent stress reasonable width and simple fracture geometry was first
control along three perpendicular directions, and with created using 1000cp glycerol at a constant injection rate
maximum capacity of 8,000 to 5,000 psi. of 1000 mL/min. The sample was subjected to
representative effective in-situ stress and with a
significant stress contrast in the plane perpendicular to
the plane of the fracture (1=4,500 psi, 2=3,000 psi,
3=1,000 psi). Borehole fluid injection at 1000 mL/min
started at approximately 138.42 minutes after the
application of stress and continued at a constant rate
until borehole breakdown was observed, at a pressure of
4,202 psi. This value exceeded the minimum and
intermediate in-situ effective stress. Figure 4 shows the
record of the measured borehole pressure, borehole
volume, flat jack volumes, cumulative AE number, fluid
volume injected to the fracture and the flow rate of fluid
entering the fracture, as a function of time. The flat jack
volumes are used to define the block deformation along
Figure 3. Large-scale polyaxial testing system (left) and large three perpendicular directions. The figure at the top
shale outcrop block (right). The system allows conducting shows the entire test, the other figures describe the
hydraulic fracture experiments on heterogeneous tight shales
region of fracture initiation and propagation and
under representative conditions of stress.
highlighting the time around the breakdown pressure.
Flatjacks are used for transmitting the load and for
continuously monitoring the block deformation during The East-West volumetric deformation at 138.67
minutes (green dotted line in the lower plots), indicates
fracturing. These measurements allow us to detect
fracture initiation and fracture breakdown. Face the onset of fracture initiation (dry fracture). This event
deformation measurements allow us to monitor whether is indicated by multiple measurements, including the
the fracture is predominantly planar or non-planar. The onset of non-linear deformation of borehole pressure,
and the onset of East-West block deformation. The
blocks are also instrumented with an array of acoustic
breakdown pressure, indicating the maximum borehole
transducers (typically 20 to 36). These transducers are
used to conduct active transmission measurements, pressure, fracture opening and possibly, the maximum
which provide a continuously updated velocity model of acceleration of fluid flow velocity into the fracture (red
the block. This allows us to account for acoustic dotted line in the lower plots), occurs 2.4 seconds later.
For this experiment, this event also coincides with the
velocity changes occurring as a result of loading and
fracturing the block. We also use the transducers onset of the North-South block deformation, indicating
concurrently in passive mode, to detect and localize opening of a second fracture oriented perpendicular to
the minimum horizontal stress. This fracture is short
acoustic emission events and use these to map the
evolution of the fracture geometry. Our broader goals are lived and closes immediately after (138.83 min). The
to observe the relationship between the rock fabric, fracture reaches the boundary of the block soon after.
fracture containment and fracture complexity, to validate Figure 5 shows results of acoustic emission localization,
interface crossing criteria models, investigate proppant highlighting the global fracture growth at the end of the
transport and proppant distribution, and better define experiment.
fracture characteristics along the near-wellbore and far-
wellbore regions.
4. RESULTS
For this project, we conducted fracture propagation
experiments on an outcrop block of the organic-rich
Niobrara formation, representative of the Niobrara
reservoir in the DJ basin. The sample configuration
simulated a vertical wellbore completion, with the
wellbore oriented perpendicular to bedding. After
fracturing the block we first measured the unpropped
along the entire length of the sample. This appears to be
arrested at approximately 8 to 10 inches from the top of
the block. However, close examination shows en-
echelon fracturing which occurred by branching and
shear displacement. Figure 7 shows a detail image of
this region. The image was rotated sideways for
convenience (notice the top and bottom directions,
indicated in the figure). Fracture complexity in this
region results from mixed-mode fracturing events
converging to meet the centered fracture propagating
across the top of the block. The fractures daylight on the
North surface of the block with widths varying from 0.5
to 5.0 mm and form clear stacking patterns. Towards the
interior of the block these coalesce and form one
dominant fracture.

Figure 6. Opposite North and South faces of the block,


displaying the fracture traces.
Figure 4. Measured borehole pressure, borehole volume, flat
jack volumes, cumulative AE number, fluid volume injected
to the fracture and the flow rate of fluid entering the fracture,
as a function of time.

Figure 7: Detail of Figure 6. The image is rotated 90 deg: the


top of the block is to the left side of the image. En-echelon
Figure 5: Results of acoustic emission localization showing fracturing and fracture branching is pervasive in this region.
the final fracture geometry.
Figure 8 (left) shows a section of the South face in
Once the fracture was generated and the conductivity Figure 6. It highlights the presence of en-echelon mix-
measurements were completed, the block was removed mode fracture propagation. The fracture is partially open
from the vessel, for sectioning and post-test fracture and discontinuous. Multiple sections of connected rock
analysis. Figure 6 shows images of the North and South (bridges) are clearly seen. The right side image shows
faces of the block. The fracture in the South face (right that the proppant distribution along the fracture length is
side image) extends clearly from the bottom of the also discontinuous. Figure 9 shows the two half-sections
sample to the top, following a general vertical of the block exposing the hydraulic fracture surfaces.
orientation, and with a slight deviation of approximately Parted horizontal planes of weakness are also seen; these
10 degrees from the middle of the sample downwards. occurred after the block was unloaded and removed from
The fracture on the North face (left side image) deviates the cell. We saw no indication of their opening during
from the vertical orientation by approximately 8 degrees testing either by the Top/Bottom displacement or by the
acoustic emission localization. Remnants of proppant
placement and their distribution are easily identified
along the surface of the block. This was flushed during
the casing coring operation. Figure 9 also shows three
distinct regions of fracturing across the block: a
wellbore-connector region, connecting the wellbore to
the fracture, which is characterized by some degree of
tortuosity, North side, as the hydraulic fracture develops
from the sand blasted slots; a near-wellbore fracture
region, characterized by a reasonably planar and smooth
fracture, and a far-wellbore fracture region, near the
edges of the block, that is associated with extensive
branching and mix-mode fracture propagation.

Figure 10: Left image: En-echelon fractures are developed


gradually from the middle of the block and towards the top
north section of the block. Two regions with step overs are
highlighted. Right images: The Red line shows the orientation
of a natural, mineralized, fracture. The lower image shows the
complex fracture pattern that result.

Figure 8. Detail from Figure 6 (right side image). The South


face also shows significant fracture complexity. Here the
fracture is discontinuous and partially open. The right side
image shows the associated discontinuous proppant
placement.

Figure 11. En-echelon fracture stacking observed on the upper


portion of the East face of the block. The thickness of the
step-overs was in all cases less than 1.0mm.
Two regions in the fracture face are bounded with red
Figure 9: East and West half sections of the block, exposing and blue rectangles. The fracture region in the red
the created fracture surfaces. The near-wellbore region, rectangle developed complexity by interaction with a
including the casing, was drilled out. A representative calcite-filled weak-plane. The orientation of this is
drawing of the openhole section of the wellbore with the indicated by the red dotted line in the upper right side of
sandblasted slots is displayed towards the center of the block the figure. A transition from few en-echelon factures, to
image. the left of the interface, and higher density of fractures,
Figure 10 shows important information on the origin and to the right of the interface, is clearly apparent in the
evolution of complex fracturing and branching. The figure. The fracture region within the blue rectangle
evolution of fracture complexity from the wellbore to the shows multiple fracture step-overs, which gradually
far-wellbore region is highlighted. grow in thickness as the fracture approached the
boundary of the block.
A detailed view of a thick step-over, seen in the mid-
section of the blue bounded area, is shown in the lower
right side of the image. We observe that fracture by calibration. Results show that the stress dependence
branches are created at weak interfaces by on fracture conductivity of the un-propped and propped
simultaneously propagation along the direction of the large fracture is substantially lower than in the small
plane of weakness, and thus creating a step-over, and sample. This is expected given the higher surface
also by crossing the weak interface, and thus roughness and more tortuous nature of the large-scale
maintaining the original direction of fracture hydraulic fracture. The smaller samples are prepared
propagation. The result is a complex fracturing pattern with smooth and flat saw cut surfaces prior to testing [2].
and the generation of considerable surface area. This indicates that the laboratory measurements on small
samples are conservative estimates of fracture
Figure 11 shows the transitional region between the
conductivity and that the actual stress dependence is
near-wellbore (right side) and the far-wellbore fracture
lower than that predicted from lab experiments. The
regions (left side), which are defined by a gradual
difference may be considerable for the un-propped
development of fracture branching and complexity. The
surfaces and smaller for the propped surfaces.
figure shows the presence of fracture branching with
three, sub-millimeter thick, stacked fractures of 7 to 8 5. FOUR KEY REGIONS OF HYDRAULIC
inches in length. Each of these layers of fractures FRACTURE PROPAGATION
contains proppant in their first half inch. Only the
Regarding the created surface area, four key regions can
dominant fracture contains proppant along the entire
be clearly identified: (1) the wellbore, (2) the
length of its surface.
wellbore/fracture connector, (3) the near-wellbore
fracture, and (4) the far-wellbore fracture network. The
particular nature of these regions changes from test to
test and may change depending on the play, the reservoir
fabric, and the distribution of rock properties. However,
we observe that these regions are always well
differentiated. Understanding the role of each of these
components to hydrocarbon production is fundamental
to understand the dominant sources of loss of
production, and their associated drivers. For example,
the far-wellbore region provides the highest surface area
in contact with the reservoir, the lowest fracture
conductivity, and offers the highest potential for
improving productivity, by preserving surface area and
improving fracture conductivity. Numerical simulations
indicate that minor improvements in fracture
conductivity in this region result in substantial
Figure 12: Un-propped and Propped fracture conductivity as a
function of confining stress improvements in production and total recovery. This is
also the region of highest water retention, highest salt
One of the goals of this program was to compare the un- dissolution and potential precipitation at the near-
propped and propped conductivity of this large-scale wellbore, if it is mobilized. In contrast, the near-
fracture, to the conductivities measured in smaller core wellbore region provides limited surface area in contact
samples. Figure 12 shows the calculated un-propped and with the reservoir. Because the majority of the proppant
propped conductivities as a function of applied stress for may reside in this region, the near-wellbore fracture
the large-scale block test and that measured on smaller- region has high conductivity and often controls the
scale core samples (of 1 inch in diameter and 1 inch overall production of the stage. Numerical simulations
length). The absolute magnitudes of fracture suggest that changes in fracture conductivity in this near-
conductivity on the large block samples were smaller wellbore region, even when substantial, do not result in
than that measured on the small samples. This was significant changes in production. This is because it is
found to be so because of an additional pressure drop already a region of high conductivity. The
between the tip of the fracture and the screen on the wellbore/fracture connector region is a small region, yet
block face, which collected fines, and another pressure highly critical, because it is highly susceptible of losing
drop at the pressure probe. Numerical modeling was connectivity. It is also a region of convergence, and thus
used to evaluate the pressure drop between the pressure a choking point. In this region, the high fluid velocity,
sensor and the end of the probe. The additional pressure during fracturing, results in low proppant content, and
drop was evaluated by equating the fracture conductivity this condition is further worsen by over flushing at the
of the two un-propped samples (small-scale and large- end of the treatment. Lack of sufficient proppant or poor
scale) at the initial pressure step. Thus, the absolute selection of rock mechanical competence in this region
values of initial fracture conductivity were made equal,
significantly increases the risk of fracture closure and propagating fracture reaches mineralized natural
loss of stage production. This is so, independently of the fractures in the block.
conditions of the other two regions. The wellbore and Fracture complexity was highest towards the
the type of completion (e.g., cased, open hole, boundaries of the sample. However, this was not
perforated, sand blasted) defines the condition of exclusively a boundary effect. Fracture complexity
fracture initiation and influences the geometry of the (branching and stepovers) were initiated at the center
wellbore/fracture connector. The wellbore location, type of the block and become gradually more prominent
of completion, presence of fractures, type of as they approached the end of the block.
perforations, use of cement dissolving acids, and even Fracture tortuosity is observed at the
the time delay between perforating and fracturing, wellbore/fracture connector region, as the fracture is
influence fracture initiation and its initial tortuosity. generated from the North side edge of the sand
Ideally one would like to define the wellbore conditions blasted slot. The other side shows a smooth
of fracture initiation such that the wellbore/fracture propagation and a planar surface with minimum
connector is simple, planar, wide, and provides a tortuosity.
maximum connectivity between the wellbore and the Simultaneous propagation of parallel fracture
near-wellbore fracture system. Unfortunately, current branches with spacing ranging from 0.5 to 5 mm is
fracture design criteria are defined to address broad observed in the near-wellbore fracture region and
objectives that are assumed to improve production (e.g., near the fracture boundaries. We believe this is
rock properties, in-situ stress, percent clay, presence of possible because these fractures are initially dry
fractures), while ignoring the different requirements of fractures and have minimal width. As fluid enters
each of the four fracturing regions for sustaining the fracture some of these widen and others remain
production. closed.
We observe a substantial variability in the
6. CONCLUSIONS
concentration and distribution of the proppant in the
In this paper we present results from large-scale fracture. In the near-wellbore fracture region
laboratory tests of hydraulic fracturing on outcrop blocks exhibiting fracture branching with three, sub-
from the organic-rich Niobrara shale, representative of millimeter thick, stacked fractures, each of these
the Niobrara reservoir in the DJ Basin. The goals of this fracture layers contain proppant in their first half
effort were: measure fracture conductivity under un- inch. However, only the dominant fracture contains
propped and propped conditions and as a function of the proppant along the entire length of its surface.
closure stress; compare these results with laboratory Comparing large-scale fracture conductivity
measurements on small samples; evaluate the geometry measurements propped and unpropped with small-
of the created fracture and investigate the effect of scale measurements shows the following: Large-
planes of weakness in the rock on the development of scale samples retain higher fracture conductivity
fracture complexity; provide additional insight on with increasing stress. This is so because of the
characteristic fracture regions within the created fracture inherent higher tortuosity on the larger samples.
and their implication to fracture productivity. Results of This means that laboratory measurements on small
this investigation can be summarized as follows: samples are conservative estimates of large-scale
Results show the generation of substantial fracture facture conductivity.
complexity, non-planar geometry and non- Post-test analysis, allowed us to describe the
continuous fracturing because of the complex rock fractured system as consisting of four easily
fabric. This happens despite the choice of high identifiable regions: the wellbore, the connection
stress contrast (i.e., stress dominated fracturing) and between the wellbore and the fracture system, the
high flow regime (i.e., high viscosity and high near-wellbore fracture, and the far-wellbore fracture
pumping rate). network.
We observe a clear separation between the onset of Understanding the role of each of these regions to
fracture initiation (representing the onset of dry hydrocarbon production helps identify the often
fracture propagation with clear indication of block competing causes of production decline over time.
dilation along the EW direction, and a rapid increase Results show that the requirements for optimizing
in the rate of acoustic emission events), and the hydrocarbon flow in these four regions are different,
onset of borehole breakdown, with fluid flow into but reasonable compromises can be made to sustain
the fracture and gradual increase in fracture width. production from the overall fracture system.
We observe substantial fracture complexity 7. ACKNOWLEDGEMENTS
associated to the textural complexity in the rock. In
The authors wish to thank Schlumberger for supporting
particular, we notice the development of branching
this effort and for permission to publish. We also wish
and step-overs (shear displacements) when the
to acknowledge the technical contribution of the 13. Gu H. X. Weng, J. Lund, M. Mack, U. Ganguly, and R.
participants of the Rock Fabric and Fracture Complexity Suarez-Rivera, 2012. Fracture Crossing Natural Fracture
consortium: Ion Ispas, BP Americas; Karen Olsen, South at Non-orthogonal Angles: A Criterion and Its Validation.
Western Energy; Ed Lake, Henry Lopez, and Mikhail SPE Production & Operations. Volume 27, Number 1
Pages pp. 20-26. February 2012.
Geilikman, Shell International Exploration and
Production Inc. and other scientists from Schlmberger,
including: Sid Green, Larry Behrmann, Redd Smith and
Nick Whitney.
8. REFERENCES
1. Vincent M., Survey on current operational experience in
tight shale reservoirs. In Appendix 1, RPSEA Sponsored
Project on Sustaining Fracture Area and Conductivity of
Gas Shale Reservoirs for Enhancing Long-Term
Production and Recovery. Final Report.
2. Ghassemi A. and Suarez-Rivera R., RPSEA Sponsored
Project on Sustaining Fracture Area and Conductivity of
Gas Shale Reservoirs for Enhancing Long-Term
Production and Recovery. Final Report.
3. Blanton T.L. 1982 An Experimental Study of Interaction
Between Hydraulically Induced and Pre-Existing
Fractures. SPE 10847.
4. Warpinski, N.R., and L.W. Teufel 1987. Influence of
geologic Discontinuities on Hydraulic Fracture
Propagation. SPE 13224. JPT: February 1987, 209220.
5. Jeffrey, R.G., L. Vandamme, and J.C. Roegiers. 1987.
Mechanical Interactions in Branched or Sub-parallel
Hydraulic Fractures. SPE 16422, 333-341.
6. Jeffrey, R.G., R.P. Brynes, and P.J. Lynch. 1992. An
Analysis of Hydraulic Fracture and Mineback Data for a
Treatment in the German Creek Coal Seam. SPE 24362.
445457.4.
7. Jeffrey, R.G., and A. Settari. 1995. A Comparison of
Hydraulic Fracture Field Experiments, Including
Mineback Geometry data, with Numerical Fracture Model
Simulations. SPE 30508. 591-606.
8. Renshaw, C.E, and D.D. Pollard 1995 An Experimental
Verified Criterion for Propagation Across Unbounded
Frictional Interfaces in Brittle Linear Elastic Materials.
International Journal of Rock mechanics Mining Science
and Geomechanics, 32, 3, 237-249.
9. Potluri, N., D. Zhu, and A.D. Hill 2005. Effect of Natural
Fractures on Hydraulic Fracture Propagation. SPE 94568.
10. Suarez-Rivera R. B. Conner, J. Kieschnick, and S. Green
2006. Laboratory Experiments of Hydraulic Fracturing on
Glass Plates Help Investigating Basic Conditions for
Fracture Branching and Fracture Containment Along
Interfaces. Paper 06-1130 presented at the 41st U.S.
Symposium on Rock Mechanics (USRMS): "50 Years of
Rock Mechanics - Landmarks and Future Challenges.",
held in Golden, Colorado, June 17-21, 2006.
11. Jeffrey, R.G. X. Zhang, and M. Thiercelin. 2009
Hydraulic fracture offsetting in naturally fractured
reservoirs: Quantifying a long-recognized process.
Presented at the 2009 SPE Hydraulic Fracturing
Technology Conference held in The Woodlands, Texas,
USA, 1921 January 2009. SPE 119351.
12. Thiercelin M. 2009 Hydraulic Fracture Propagation in
Discontinuous Media. Presented at the International
Conference on Rock Joints and Jointed Rock Masses,
Tucson, Arizona, USA, January 7-8, 2009.

You might also like