IPTC 17439

Emerging Technologies and the Future of Hydraulic Fracturing Design in

Unconventional Gas and Tight Oil
Ernesto Fonseca, Shell International Exploration and Production Inc

Copyright 2014, International Petroleum Technology Conference

This paper was prepared for presentation at the International Petroleum Technology Conference held in Doha, Qatar, 20–22 January 2014.

This paper was selected for presentation by an IPTC Programme Committee following review of information contained in an abstract submitted by the author(s). Contents of the paper, as
presented, have not been reviewed by the International Petroleum Technology Conference and are subject to correction by the author(s). The material, as presented, does not necessarily
reflect any position of the International Petroleum Technology Conference, its officers, or members. Papers presented at IPTC are subject to publication review by Sponsor Society
Committees of IPTC. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of the International Petroleum Technology
Conference 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 where and by whom the paper was presented. Write Librarian, IPTC, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A., fax +1-972-952-9435

Abstract

As industry leaders seek solutions to the myriad of challenges facing unconventional projects, research and
development efforts continue in the area of multistage fracturing design. These research activities are focused on
the identification of novel approaches to hydraulic fracturing (HF) design that can lead to more effective design
practices in the future.

This paper describes emerging technologies within the building blocks of HF design for multi-stage horizontal
wells, and proposes how application of these technologies can lay a foundation for the development and evolution
of future industry practices. The industry debate on HF design commonly revolves around a perceived need for
more powerful software tools that can capture complex fracture geometries. This perspective or position within the
broader industry is due in large part to the cloud of microseismic induced events that is oftentimes observed to
envelop a treatment area. Some interpret that these events arise from very complex fracture geometry, and
therefore modeling tools with increased specialization are needed.

Rather than focusing only on the ability of the software to replicate complex fracture geometries, a higher-level
and integrated view of HF design is recommended. As shown in Fig. 1, this integrated approach considers the
progressive analysis and application of subsurface diagnostics and modeling capabilities, and how they can
influence meaningful decisions in the area of HF design. In reality, the capability for HF design is as strong as the
weakest of these three components and need not rely solely on modeling capability.

Modeling
Capabilities
• Planar 3D

Subsurface
Hydraulic Frac
Diagnostics Design
•Microseismic
• Temperature & Meaningful
Production Logs Decisions
•Dfit & radiocative • Fluid Type,
proppant tracers optimum volume
and rate, landing
depth,

Fig. 1: Integration of the primary building blocks in hydraulic fracturing design.

2 IPTC 17439

For example, consider the scenario where an advanced set of subsurface diagnostics are found to be limited by
software capabilities due to the inability of the software to replicate the diagnosed phenomena with credible
physics. This situation limits the use of subsurface diagnostics because the field observations are not mapped to
the modeling capabilities and a relevant decision. In a similar scenario, modeling capabilities may be underpinned
by credible data and diagnostics, but also lack the ability to influence a critical design parameter such as pump
rate, well landing depth or fluid choice.

This paper explores how emerging technologies within these building blocks are evolving and how this progress is
resulting in new and relevant engineering choices in the design of hydraulic fractures. These choices include
design for lateral sweet-spotting, better approaches in sequence and spacing of wells and fractures, and re-
fracturing decisions for horizontal wells.

Brief Description of the State of the Art in Hydraulic Fracturing

The ‘state of the art’ in diagnostics and modeling capabilities has been used to generate relevant engineering
design choices for hydraulic fractures in multi-stage horizontal wells. As summarized below and illustrated in Fig.
1, this expertise can be sorted into each building block of the design process.

• Subsurface Diagnostics:
o Microseismic is used for calibration of vertical stress profiles generated from core and sonic log
data to match vertical frac growth.
o Microseismic can also be used to infer fracture lengths via comparisons against a range of
simulation outcomes for credible leak-off and pumped volumes.
o When fractures are created in appraisal or developmental vertical wells, an estimation of fracture
height can be measured with either temperature logs, permanent distributed temperature sensing
fiber optics, or proppant tracers.
o A diagnostic fracture injection test (DFIT) is commonly conducted for pore pressure estimation,
and is also needed for calculating the vertical stress profile. Inference of permeability and
fracture closure can be challenging considering the significant effect on the water imbibition of
shale and possible complex leak-off due to the presence of natural fractures.
o Production logs give a quantitative view of how effectively fractures were generated in the lateral
section and radioactive traced proppant provides, among other learning’s, a qualitative
assessment of location of propped fractures and how well fractures propagate away from and
perpendicular to the well.
• Modeling Capabilities:
o Planar 3D models are commonly available in the industry. These models assume fracture growth
happens in a plane and they provide estimates of height, length and width of fractures. The
models also assume fractures are symmetrical on both sides of the well and can account for
shadow stress from neighboring fractures by reflecting a reduction in width and/or length.
• Meaningful Decisions:
o Initial screening evaluation of slick-water, gel fracs or hybrids can be completed with planar 3D
models [1].
o Landing depths and rate sensitivities; and in particular, with the presence of shale laminations,
carbonate layers and differential vertical strengths within the pay zone as this is an important
design exercise.
o After capturing indications of fluid performance at different rates, a more careful analysis is
usually conducted by evaluating the point of diminishing economic returns with respect to job size
for a single frac.
o When the optimum(s) single fracture design is concluded, estimating fracture spacing considering
recovery economics and shadow stress yields a final design ready for integration with the well
and completion (# of perforation, sleeve(s), etc.) design.
o HF design process is hardly sequential, as it is coupled to many design parameters, and
depending on geology (layering, zone thickness, permeability, etc), some steps become more
IPTC 17439 3

relevant than others do. As a result, some steps in the design process can receive more or less
emphasis.

There are technology advancements emerging within the areas of subsurface diagnostics and modeling that can
significantly influence future progress in HF design. In the area of subsurface diagnostics, there are low cost
surveillance techniques that use drilling data to characterize natural fractures and rock strength. Another
diagnostic tool seeing increased application is the interpretation of Distributed Acoustic Sensing (DAS) for use in
interpreting and understanding fracture interference and fracture divergence. In the area of hydraulic fracture
modeling capabilities, the emergence of software packages that handle multi-well and multi-cluster design
capabilities are starting to appear in the literature [2] and will certainly play a key role in future design efforts. In
addition, these software simulators are incorporating non-planar 3D capabilities. New progress in the areas of
diagnostics and modeling are opening doors to designs that incorporate new meaningful decisions like lateral
sweetspotting, better approaches on the sequence and spacing of fractures, and support of re-fracturing
decisions for horizontal wells.

Emerging Technologies in Subsurface Diagnostics

1. Rock properties while Drilling

There are low cost technologies emerging within industry that use drilling data to improve the definition of
important production properties along the horizontal wellbore. One such new tool is the software fracid_ubd [3]
that is being used for fracture characterization. Another emerging technique is “rock strength determination while
drilling” that is in final development (see below). There are established methods in the industry to identify natural
fractures near the wellbore; such as, image logs and the determination of rock strength properties using sonic
logs. However, the benefit derived from using these data has not yet offset their cost impact of using them on
every single well. As a result, it is important that progress is being made in the detection of natural fractures and
rock strength using data already available from mud and drilling operations, as a low-cost source of data to
improve HF design.

There may be a correlation between the density of pre-existing natural fractures and productivity in many plays,
whether positive or negative. As shown in Fig. 2 below, fracid_ubd identified natural fracture planes in the
northeast direction and these are consistent in location in both wells. The software highlights the presence of a
fracture based on simultaneous gas and mud pit increases in a well drilled underbalanced, using a proprietary
statistical inference algorithm.

Fig. 2: fracid_ubd identification of natural fractures [3]

4 IPTC 17439

This analysis suggests these higher density fracture areas may be more productive, and the ability of the
fracid_ubd software to now pinpoint these previously unidentified productive areas should open the door to new
design options for operators, such as:
• The placement of more fractures in the side of the well with a higher concentration of natural fractures, in
order to target the more productive zone of the well and increase the return on investment, or
alternatively
• The placement of fewer fractures in this area, since a more productive well section may require less
intensive stimulation treatments.

These are clearly competing options, and the decision of which option to pursue will depend on the extent of
increased productivity by the well section with the highest concentration of natural fractures and the resulting
economics of that choice. In other words, the estimated amount of incremental production will determine the
theoretical economic optimum and indicate which choice is better than the other. Nevertheless, the analysis
shows that this new surveillance capability has generated a new and meaningful decision for consideration in HF
design.

On the topic of rock strength, there continues to be an ongoing debate over what type of rock is preferred. Hard,
brittle rock can offer an advantage [4] because it is more likely to be naturally fractured and thus respond better to
stimulation, versus a softer rock that has the advantage of being easier to prop open. In either case, for those who
hold the view that a certain threshold of rock strength is the best, the emerging capability to now characterize rock
strength at a low cost in every lateral establishes the opportunity to focus on more favorable positions along the
lateral. As shown in Fig. 3, the software DWOB-DROCK developed by Dr. Geir Hareland, Harcon Inc. and Shell
International Exploration and Production Inc., takes drilling data and provides the prediction of Young’s modulus
(YM). For this particular case the average values of unconfined compressive strength (UCS) and YM predicted
with rate of penetration (ROP) models were found to be 99.57 MPa and 29.64 GPa respectively. These values
are consistent with the laboratory measurements for this formation was found to range between 13.78-41.35 GPa.
The discrepancy in prediction of ROP models versus those derived from logs could be a result of uncertainties in
drilling data that are explored in more in depth in a separate publication [5].

UCS (MPa) Young's Modulus (E) (GPa)

50 70 90 110 130 150 0 10 20 30 40 50 60
2640 2640

2670 2670
Measured Depth (m)

2700 2700

2730 2730

2760 2760

ROP Models ROP Model

Onyia, 1988 Horsrud, 2001
Horsrud, 2001 Sayers, 2010
2790 2790

Fig. 3: Comparison of UCS (left) and YM (right) obtained from ROP models with those obtained from sonic logs.
IPTC 17439 5

In either case, resolving the uncertainty around model calibration can be addressed using techniques suggested
in the literature [5]. A calibrated model can be useful in the selection of better stimulation targets. In addition, it
can facilitate development of a stronger correlation between rock strength type and production using a
comparison to production logs.

The accepted approach within industry for fracture spacing has typically been the application of a geometric
reference, meaning that the same space is provided between fractures and this dimension is determined based
on an optimization of financial recovery as a function of spacing distance. Formation average permeability is
usually the key variable behind this relationship, whereas having the capability to sweet-spot a horizontal well
based on density of natural fractures and/or rock properties is a choice that is becoming available in the industry.
It is possible that a more focused placement of fractures will become the norm because of this choice. A primary
downside to this strategy is that since fracture spacing is typically close (60 ft); lateral sweetspotting may generate
fractures that are much closer to each other if the total number of fractures in the lateral remains the same. The
solution to this dilemma is to reduce the total number of fractures; however, this option also carries uncertainty
because it leaves some zones without stimulation.

2. Distributed Acoustic Sensing (DAS)-Fracture Interference

Using DAS diagnostics [6] in the field has provided insight into the phenomena of uneven proppant distribution
within a frac stage. As shown in Fig. 4 below, proppant flow and stimulation was greater at the heel location of the
cluster. This correlates very well with the density of induced microseismic events. Geomechanical modeling also
suggests that shadow stress from the previous stage triggers this behavior. Stress from the previous stage
causes the fracture growth to favor the fractures located the farthest distance from the previous stage. This
phenomenon of shadow stress introduces the need for engineered solutions that will compensate for this effect in
the HF design. Lowering the rate to reduce shadow stress is a classical approach to minimize this effect. Zipper
fracturing is another technique that can be used to offset the impact of shadow stress. With zipper fracturing, the
design provides for alternating the fracturing activity between wells in every stage in order to allow time to
dissipate the stress. This approach increases the time between fractures allowing leak-off to dissipate for a
longer period of time in comparison to a similar well sequence, thus reducing the stress. Although achieving an
even proppant allocation between fractures is a challenge, stimulation with significant stress shadow is viewed by
some in the industry as a positive factor because it may be opening natural fractures, thus increasing stimulated
rock volume. In any case, this inevitable effect caused by the required close proximity of the fractures in shale gas,
as a result of the low permeability, it is of significant importance and design efforts to either benefit or avoid it is
another emerging design choice to evaluate in the HF design.

Fig. 4: Proppant allocation according to DAS data [6]

6 IPTC 17439

Emerging Technologies in Design Capabilities

1. Limitations of Existing Planar 3D

The use of existing planar 3D frac simulators can provide insight into the performance of different fracturing fluids
[1], including analysis of job size and landing depths. Although the simulation results are derived using an overly
simplistic fracture model, the planar 3D frac simulator has proven very useful in comparing different fracture
geometries at different rates and fluid types when calibrated with microseismic data. This calibration can be
useful in performing sensitivity analyses on both well landing depth and rate. However, these modeling
capabilities are inherently limited and are unable to yield a more credible set of answers given the fracture
geometries and fracture response mechanisms that are not captured in existing planar 3D simulators. As shown
in Fig. 5, the simplification of simulating a half length frac does not permit fractures to accommodate shadow
stress by uneven half lengths and/or widths in the same fractures versus suppressing the fractures in the middle
of the cluster. Another limitation of 3D planar models is the inability to account for fracture rotation as also shown
in Fig.5. Fracture rotation in the outer fractures in the first stage is another possible mechanism in response to
shadow stress and may be present in stimulated wells.

Fig. 5: Possible fracture geometries not captured in existing simulators

The absence of these mechanisms restricts the spectrum of plausible answers that can be considered when
planar 3D simulators are used. This limitation in software capabilities is diminishing with the emergence of non
planar 3D frac simulators [2] to be discussed further.

2. Planar or Complex Fractures

This discussion will be introduced by addressing the notion, advocated by many, that the cloud of induced
microseismic events likely results from a very complex fracture geometry. However, the apparent microseismic
event cloud does not exclude the presence of a dominant fracture where the complexity is actually due to fluid
leak-off. Natural fractures are commonly cited in this discussion; however as pointed out by Wong, Geilikman, and
Xu [2], it is uncommon to find naturally conductive fractures in core and log data; the natural fractures that can be
observed are normally filled with minerals. Propping open natural fractures is very difficult as the mineral-filled
fractures limit possible fluid intake, and shear failure mode requires a very specific set of circumstances. The
microseismic observations and the planar main feature are reconciled by the conceptual model of reference [2],
IPTC 17439 7

identified as “complex leak-off”. The phenomenon of “complex leak-off” is explained as follows. In very low
permeability rocks, small fractures or contrasts in matrix permeability can provide significant variability in leak-off
in different parts of the rock fracture face. Thus, fracture fluid likely enters the formation with a branching pattern
rather than with an evenly advancing penetration front. This form of leak off does not necessarily infer stimulated
volume because proppant is (obviously) not carried along with the leak-off fluid. Some of the microseismic events
may indeed be caused by shear fracturing and thus be self-propping, but this may only be a small portion of the
total. Using this concept and since the modeling objective is to model a dominant fracture, the emerging non-
planar 3D simulators based on a dominant fracture are applicable for meaningful design.

3. Emerging Non- Planar 3D HF Simulators

Fig. 6 shows the capability of the non-planar 3D HF simulator, FrackOptima [2], which is a new tool, recently
introduced to industry. Its use is for simulation of multiple non-planar fractures in multistage hydraulic fracturing
along horizontal wells and includes capability to model various geometric configurations. As illustrated below, a
FrackOptima simulation of the first three clusters in two parallel wells stimulated in zipper frac mode shows how
each of the three previously discussed shadow stress mechanisms can be represented. Stage #1 shows the
traditional half wing symmetry approach of the two fractures in the middle being suppressed by the outer fractures
despite some rotation of these fractures. Stage #2 shows how the assumption of equal half wing frac growth is not
a constant in these developments given the tight spacing between wells (every 300 to 700 ft) and fracture spacing
(100 to 40 ft). The simulation clearly shows that fractures on Stage #2 grow significantly more on the non-
stimulated side of the well (top of the figure). It also shows how total fracture length of the middle fractures is
constant with each other and with offset fractures having similar length as shown in the center illustration in Fig. 5.
Finally, because Stage #3 is emerging in an already crowded fractured space, the rotation of the fracture is
intensified.

Fig. 6: Top view of multi well and multi stage HF interaction in two parallel wells stimulated in zipper-frac mode.
The three circled areas represent the perforated wellbore and are numbered by the stage-order in which they
were stimulated.

This analysis shows how novel-modeling capability can provide insight into how different modes of
accommodating shadow stress can co-exist and collectively create a significantly different outcome than
previously thought. So far, these examples have considered the impact of stresses within a 2 dimensional
perspective of fracture direction and width.
8 IPTC 17439

Fig. 7: Multistage fracture interaction

However, the accommodation of these stresses can also occur with similar diversity in the vertical direction. Fig. 7
shows key variations that are important to consider.

The most important impact of this emerging technology for multistage fracture modeling is that it opens the door
to begin phasing out the “cookie-cutter” design approach. It is clear to see that the ongoing and repetitive
application of an idealized design carries assumptions that can be critically challenged and the emergence of this
technology provides the opportunity to achieve better results.

Trends in Meaningful Decisions

From this technology-mapping exercise it is clear that improvements in diagnostics bring about a new set of
engineering choices for hydraulic fracturing design as long as the computational capabilities keep up with the
developments in subsurface diagnostics. The emerging low cost abilities to characterize the lateral section can
not only trigger a more focused effort in the placement of fractures, but also trigger an intensification of
refracturing in horizontal wells. Rock strength and natural fractures with drilling data correlated to production logs
may start to shed light in productive areas, currently non-stimulated, and previously considered to be non-viable
and not worth the refracturing efforts. Affordable and location precision completion options for horizontal wells are
being applied at a greater degree like fracturing with coiled tubing or retrofitted completions with packers. All
these technologies in a low gas price environment in North America and an inventory of sub-optimal wells typical
from the start of the development makes a compelling case for refracturing as a new and important emerging
trend. This approach can be visualized as beginning the design process with a horizontal and refracturing-friendly
well in mind using ball drop fracture sleeves systems that can later be closed with coiled tubing for a refracturing
round using the plug and perforation method.

Management of shadow stress was discussed in the modeling section of this paper. There are vast opportunities
in this area as modeling capabilities improve, and as a result, the prioritization of these first order effects still need
to be determined and more diagnostics deployed to validate the results. Besides capturing what is the most
advantageous means to avoid or exploit stress shadow, the movement away from the “cookie-cutter” design and
towards a design that captures the complete design cycle on the pad is a pivotal step. This is particularly
important because the resulting fracture geometries are a function of the sequence and timing of the events.

Conclusions and Future Thoughts

Fig. 8. below illustrates the integration of the primary building blocks in hydraulic fracturing design, The emerging
technologies trends here discussed are presented in green highlight.
IPTC 17439 9

The overarching message from this analysis is that success in HF design is very much about employing an
integrated approach to generate a wider slate of novel and relevant engineering choices early in the design
process. The analysis shows that there are emerging viable alternatives to continuing to employ conventional
wisdom and therefore only realizing incremental improvements on the decisions that are made today. By pairing
new diagnostics with the appropriate modeling capabilities, newer choices are created for use in making fracturing
decisions; and for this reason, relevant technology programs must work very closely to define the relationship
between new diagnostics capability and modeling of the fracturing phenomena at the appropriate level of
complexity. The emergence of non-planar 3D simulators with multi-cluster and multi-well capabilities introduces
valuable design opportunities around shadow stress modeling and may motivate the industry to a more
comprehensive design strategy, and away from the standardized, repetitive design approach through all the steps
required to frac a complete multi-well pad.

Refracturing of horizontal wells will eventually become more common as the industry learns more about the
existing inefficiencies of HF design and execution. With the existing large inventory of wells, challenging gas
prices in North America, and the range of completion options available to perform this operation, refracturing
horizontal wells can become a competitive means to keep facility utilization high and offer a viable alternative to
drilling new wells. Low cost lateral sweetspotting techniques are key enablers in this strategy.

Modeling
Capabilities
• Planar 3D
•Non planar 3D
•Multi cluster &
multi well

Subsurface
Hydraulic Frac
Diagnostics Design
•Microseismic
• Temperature & Meaningful
Production Logs Decisions
•Dfit & radiocative • Fluid Type, proppant size,
proppant tracers landing depth and rate, job
•Natural Fracture volume and fracture
Identification & spacing.
rock strength while •Re-fracturing
drilling
•Lateral sweet-spotting
•DAS and shadow
stress •Shadow stress
management
•Pad lifecycle design
Fig.8: Multistage fracture interaction
10 IPTC 17439

Acknowledgements

The author would like to thank Shell International Exploration and Production Inc. for the permission to publish
this paper. Additional thanks are extended to Sau-Wai Wong and Claudia Hackbarth for their valuable insights
and mentoring, as well as their efforts to champion the advancement of HF design capabilities within Shell.
Appreciation is also extended to Lisa Barnes for her contribution to the review of this paper.

References

[1] Farinas, M. and Fonseca, E. 2013. Hydraulic Fracture Simulation Case Study and Post Frac Analysis in the
Haynesville Shale. SPE Paper 163847 presented at the SPE Hydraulic Fracturing Technology Conference held in
The Woodslands, TX, USA, 4-6 February 2013.

[2] Wong, S., Geilikman, M. and Xu, G. 2013. Interaction of Multiple Hydraulic Fractures in Horizontal Wells. SPE
Paper 163982 presented at the SPE Middle East Unconventional Gas Conference and Exhibition held in Muscat,
Oman, 28-30 January 2013.

[3] Norbeck, J., Fonseca, E. and Griffiths, D. 2012. Natural Fracture Identification and Characterization While
Drilling Underbalanced. SPE Paper 154864-PP presented at the Americas Unconventional Resources
Conference held in Pittsburgh, PA, USA, 5-7 June 2012.

[4] Rickman, R., Mullen, M., Petre, E., Grieser B. and Kundert D. 2008. A Practical Use of Shale Petrophysics for
Stimulation Design Optimization: All Shale Plays are not Clones of the Barnett Shale. SPE Paper 115258-MS
presented at the 2008 SPE Annual Technical Conference and Exhibition held in Denver, Colorado, USA, 21-24
September 2008.

[5] Kerkar, P. and Hareland, G., Prediction of Rock Compressive Strength Using Downhole Weight-on-Bit and
Drilling Models. 2014. IPTC Paper IPTC-17447-MS prepared for presentation at the International Petroleum
Technology Conference scheduled for Doha, Qatar, 20–22 January 2014.

[6] Molenaar, M. and Cox, B. 2013. Field Cases of Hydraulic Fracture Stimulation Diagnostics Using Fiber Optic
Distributed Acoustic Sensing (DAS) Measurements and Analyses. SPE Paper 164030 presented at the SPE
Middle East Unconventional Gas Conference and Exhibition in Muscat, Oman, 28-30 January 2013.