SPE-192898-MS

A New Collaborative Workflow to Optimize Well Placement in the First

Unconventional Horizontal Well in the United Arab Emirates

Wael Fares, Ahmet Aki, Ramy Essam, Kresimir Vican, and Ehab Negm, Halliburton; Pierre Van Laer, ADNOC

Copyright 2018, Society of Petroleum Engineers

This paper was prepared for presentation at the Abu Dhabi International Petroleum Exhibition & Conference held in Abu Dhabi, UAE, 12-15 November 2018.

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
A horizontal well was drilled in an unconventional source rock for the first time in the United Arab Emirates
using advanced logging-while-drilling (LWD) high-resolution microresistivity and near-bit gamma ray
imaging sensors. The well plan required locating a ±3 ft true vertical thickness (TVT) source rock sublayer;
this sublayer was located within a giant undeveloped unconventional carbonate reservoir in a middle
Cretaceous carbonate sequence onshore Abu Dhabi. Additional planning and drilling challenges included
the absence of offset wells in this field and the associated structural geological uncertainty. Consequently,
a vertical pilot hole was drilled to evaluate the local geological structure and to determine the reservoir
properties.
For source rock reservoirs, a unique and collaborative approach has been developed that enabled the
associated specific challenges to be addressed. This approach was built on reservoir knowledge to help
determine where and how to drill, as well as where and how to fracture, to optimize each stage of the field
development.
A well placement and geosteering methodology/workflow was developed and applied. This workflow
used azimuthal high-resolution microresistivity and near-bit azimuthal gamma ray sensors to land and
geosteer the well within the thin target layer, while maintaining the planned trajectory with minimum
borehole tortuosity using real-time drilling optimization. The high-resolution microresistivity images were
analyzed to classify image patterns to correlate to reservoir units for further field development and to better
understand the reservoir flow units. The laterolog resistivities were also determined to be better suited for
inferred Rt and Rxo measurements for petrophysical formation evaluation than the modeled electromagnetic
wave resistivity inverted values.
The well was successfully drilled, and the geosteering objectives were achieved with 100% reservoir
contact. The field development strategy for the horizontal drain in the undeveloped source rock is being
reviewed, based on these recent data, to benchmark and optimize well placement, formation evaluation, and
production technologies, to evaluate the economic potential of this unconventional reservoir.
This paper presents a case study after completing the first unconventional horizontal well in the UAE and
describes the integration of fit-for-purpose geosteering technologies and optimized drilling performance to
maximize reservoir contact and improve reservoir insight. Benchmarking of cost effective well placement
2 SPE-192898-MS

and petrophysical data gathering requirements for the future field development program are also discussed
for maximizing asset value.

Introduction
The target well was planned to be drilled as a single horizontal oil producer well in a target source rock within
reservoir A in an unconventional field located onshore Abu Dhabi. A real-time, high-resolution resistivity
image LWD tool and near-bit azimuthal gamma ray sensors were used for well-placement purposes in Field
A (Fig. 1). This field is within carbonate sequences of the Middle Cretaceous.

Figure 1—Field A location map.

The execution of the plan began with a vertical pilot well, with wireline acquisition and diagnostic
fracture injecting testing (DFIT). This vertical pilot was also located to enable microseismic acquisition
from a nearby vertical offset well with the intention to derive fracture height and minimum horizontal stress
direction.
Petrophysical analysis from the nearby offset well had identified three viable targets in the formation of
interest, two of which were recommended for testing. The data acquisition program in the vertical pilot well
was designed to complete and refine the gathered data from the nearby offset well, which is located 80 m
from the target well location, and to finalize the choice of the landing zone for the subsequent horizontal
well. Planning for drilling, stimulation, monitoring, completion, and production models was further refined
with new data gathered from the vertical pilot well.

Prejob Planning
The horizontal section was planned to cut approximately 3,000 ft, targeting thin carbonate layers with
respect to the formation dip.
The expected apparent formation dip in this area was 1.5 degrees up-dip toward the total depth (TD) of
the section. There were no indications of expected faults in this area.
A prewell study and forward modeling were performed using the pilot hole and the available nearby
well data. Log analyses demonstrated the zones of interest, which were indicated by high porosity and low
resistivity, as shown by true vertical depth (TVD) logs (Fig. 2).
SPE-192898-MS 3

Figure 2—Offset wells correlation; target zone thickness.

For source rock reservoirs, a unique, collaborative approach was developed to enable the associated
challenges to be addressed. This approach, the seismic-to-stimulation workflow, was built on reservoir
knowledge to help determine where and how to drill and where and how to fracture to optimize each field
development stage (Fig. 3).

Figure 3—Seismic-to-stimulation workflow.

This workflow represents a collaborative process that focuses on bringing together subject matter experts
from the service company and the operator to work toward a specific goal. A central tenant of the workflow
is to focus on gathering data about the reservoir to reduce uncertainty and to guide recommendations on
technology and processes for all stages of a field development. Because well location is a critical deliverable,
the initial stages of the engagement focused on peer reviewing existing data from the pilot well, including
petrophysical, geochemical, and geomechanical data, to build into a 3D earth model. This model provides
initial recommendations about well location and can later be updated with additional information to refine
the model (Fig. 4).
4 SPE-192898-MS

Figure 4—Unconventional tight oil reservoir challenges.

Based on uranium content obtained from spectral gamma ray, the target reservoir can be divided into
six alternating tight/source rock zones. Conversely, based on nuclear magnetic resonance (NMR) porosity
and T2 distribution, the reservoir can be divided into five alternating zones of "relatively" good and poor
quality (tight) zones (Fig. 5 and Fig. 6).
It was concluded that source rock 2 (S2M) shows the optimum reservoir properties in which the oil
saturation index (OSI) was greater than 100, with gases, gas/oil ratio (GOR), and permeability exhibiting
highest ratings.

Figure 5—A composite log of density, neutron, resistivity, spectral gamma ray, sonic, and NMR data showing the presence of
organic matter (indicated by zones of high uranium); the reservoir indicates an alternation of tight and source rock carbonate.
SPE-192898-MS 5

Figure 6—NMR T2 and pore bins distribution indicates that the reservoir has three zones of relatively
better quality and two zones with low quality, as indicated by red and blue rectangles, respectively.

Geomechanical Evaluation
This step calculates geomechanical attributes in terms of elastic moduli, the stress field, and brittleness
index using wireline (or while-drilling) data. Reservoir geomechanics is discussed in more detail in the
following subsections.
Processing Workflow. Fig. 7 shows the processing steps used in the workflow, including elastic rock
properties, dynamic to static calibration, brittleness index and mapping, and 3D anisotropic stress.

Figure 7—Steps used in the processing workflow.

Processing Results, Validation, and Discussion. The following describes the processing results,
validation, and discussion:

• Rock elastic moduli, Poisson’s ratio, and Young’s modulus were calculated, and then validated
with special core analysis from offset wells.
• Overburden pressure was calculated using density and sonic data.
6 SPE-192898-MS

• It is noted that the lack of significant reservoir anisotropy from cross dipole sonic measurements
give rise to significant uncertainty with calculated pore pressure and stress analysis results.
• Brittleness index results indicate three zones with relatively lower brittleness index (ductile), as
shown in Fig. 8.

Figure 8—Brittleness index; areas highlighted in yellow indicate zones with relatively lower brittleness index.

Geomechanical Interpretation
A peer review was performed of the geomechanical data gathered from the offset well. Because the stress
regime is critically important in determining the placement of the horizontal well, as well as to ensure a
proper stimulation design, a detailed review of geomechanics was performed. Fig. 9 and Fig. 10 show the
local stress orientation and stress anisotropy, respectively.
SPE-192898-MS 7

Figure 9—Local stress orientation.

Figure 10—Stress anisotropy.

The geomechanics analysis results highlight certain points. The lack of formation pressure measurements
led to uncertainty in the pore pressure and SHmax magnitude. It is believed that higher pore pressure
is likely (approximately 12 ppg) with low horizontal stress anisotropy (SH-Sh), coupled with a lack of
breakouts. This scenario makes it difficult to accurately determine stress orientation and type of shear
anisotropy. Mechanical stratigraphy was predominantly in the compressional (TF) regime across tight
carbonate intervals, and the source rock intervals exhibit strike-slip (SS) regime. Induced tensile fractures
are likely to initiate horizontally as a result of predominantly TF stress regime and depend upon mechanical
layering and depth. The mechanically weaker intervals enable the best chance for inducing near-vertical
fractures. Because of low stress anisotropy coupled with the elastic rock properties within these weak
intervals, both horizontal and vertical-planar fractures are unlikely; complex planar fractures with fissures
are most likely.
8 SPE-192898-MS

Geochemical Evaluation
This section provides information about total organic carbon (TOC), kerogen, vitrinite reflectance (Ro), as
well as hydrocarbon typing.
TOC, Kerogen, Ro. TOC, kerogen, and other geochemical attributes were calculated using different
models on wireline-acquired data. Results were calibrated and validated to the core geochemical analysis.
Vitrinite reflectance, or Ro, is obtained from maximum temperature, Tmax, using an empirical equation to
be in the range of 0.6 to 1.0. However, because of uncertainty of the Tmax value, the temperature calibration
was performed by an unconventional technology team in ADNOC, resulting in a higher calculated Ro value
(approximately 1.08).
As compared with an offset well (18 km apart and approximately 2,000 ft higher), the offset well is
believed to be more mature. Fig. 11 shows the geochemical analysis, indicating the TOC and kerogen
percentages.

Figure 11—Geochemical analysis from openhole logging shows that the TOC is
approximately 4% (track 6), and kerogen is approximately 10% (track 7). Both calculations
were compared to TOC and kerogen values obtained from the geochemical analysis.

Hydrocarbon Typing and Quantification. Based on a fluid sample obtained after fracturing the pilot
well and benchmarking the basin with the Natih formation in Oman, the reservoir fluid is believed to be
liquid hydrocarbon (oil). Geochemical core analysis, however, suggests the probability that the reservoir
may have a higher GOR than initially thought, or could potentially contain gas or condensate. It is
recommended to acquire sidewall pressurized coring to obtain in-situ pressure-volume-temperature (PVT)
analysis to provide confirmation. Core geochemical analysis also indicates that the reservoir contains zones
of mature oil-prone hydrocarbon (Fig. 12).
SPE-192898-MS 9

Figure 12—Multiple geochemical indicators showing oil, mixed, or gas prone.

Earth Modeling
A 3D earth model (Fig. 13) was created using existing data to make preliminary recommendations about
well location and horizontal wellbore placement.

Figure 13—3D grid model.

10 SPE-192898-MS

Recommendations from Earth Modeling. A reservoir model was constructed based on available data.
Regional trends were used, and horizons were controlled by 2D/3D seismic data. The objectives of reservoir
modeling included understanding where and how to drill in the reservoir and where and how to stimulate
in the reservoir. The shared model provides a common platform for all analysis and engineering projects
continuing through the Ghurab basin development. Major reservoir properties have been simulated, and a
quality index was calculated to delineate the variation of the reservoir rock. The recommendations from
the earth modeling provided coordinates for potential future well locations that enable sweet spots to be
targeted. These preliminary well locations will be refined to the actual locations as additional data is entered
into the model.
Various scenarios to drill the well were comprehensively discussed, and the strategy to achieve the plan
was clearly understood by all of the involved teams, including the drilling, completions, and reservoir
engineering domains. Prejob forward modeling of expected LWD tool responses was also conducted as part
of the earth modeling process. Proprietary geosteering software was used to simulate agreed upon geological
scenarios and estimate the various responses to drill the target well while achieving maximum reservoir
contact (Fig. 14 and Fig. 15 are a subset of these). The geosteering challenges included the earth model
uncertainty (such as formation dip and faults), uncertainty of reservoir properties, maximizing reservoir
contact within a thin reservoir (+/- 3 ft TVT), and avoiding the water zone. The outcomes included the
selection and understood limitations of the LWD tools and an undertsanding of how to optimize the well
path given the various scenarios.
Formation evaluation assessment revealed a significant induction resistivity contrast at both the top
and bottom boundaries, which because of the thin reservoir nature, would affect the induction resistivity
measurements, making traditional distance-to-bed boundary applications not viable.

Figure 14—Geosteering forward modeling.

SPE-192898-MS 11

Figure 15—Geosteering forward modeling.

High-resolution microresistivity imaging, laterolog resistivity, and near-bit azimuthal gamma ray
assessments were judged to be the optimal selection to geosteer the wells. This selection enabled early
detection of formation boundaries as a result of near-bit sensor placement, high-resolution imaging for
accurate dip picking in real-time, and three different laterolog resistivities at 3 in., 7 in., and 11 in. depths
of investigation for formation evaluation, petrophysical analysis, and interpretation purposes.
The target reservoir interval has displayed reservoir qualities (including PHIE, TOC, SW, and QI), but
with very low permeability. Because of the low permeability, it is critical to design an effective stimulation
treatment. The fracture was designed based on results of minifrac and DFIT performed in the pilot hole that
was monitored by the microseismic. In the lateral section, the fracture geometry was calculated based on
pressure history match. Post-fracture flowback and production data were used to evaluate and assess the
fracture efficiency and the contribution of each stage.

Real-Time Well Placement

The optimized bottomhole assembly (BHA), free from chemical radioactive sources, consisted of a point-
the-bit rotary steerable system (RSS), high-resolution microresistivity imaging, and near-bit azimuthal
gamma ray sensors. The RSS provided more control on steering and feedback on formation steerability. The
geosteering strategy was to use the real-time dip estimates to guide steering decisions to remain parallel to
structure. In the event of any unexpected change in formation dip that differs significantly from the modeled
dip, reservoir properties that do not match well with expectations, or the occurrence of displacement as
a result of faulting, then the geosteering strategy will be re-assessed and adjusted accordingly. The real-
time LWD data were transmitted from the rig site to the office located in Abu Dhabi by means of satellite
technology. The data integration, processing, and presentation occurred in the real-time operating center
(RTOC) for proactive decision making. The synergy between geologists, petrophysicists, and geoscience
subject matter experts located together helped to validate the incoming real-time data.
The real-time resistivity image, at-bit resistivity, and near-bit azimuthal gamma ray information were
used to land the well in the target zone. The at-bit resistivity increased as the top reservoir was drilled,
providing a broad indication of geological positioning. The near-bit gamma ray information was used for
12 SPE-192898-MS

finer definition; good correlation was shown with the nearby offset well to ensure that the well was landed
in the optimal zone from the offset well wireline assessment (Fig. 16).

Figure 16—Landing in the target zone.

The azimuthal resistivity image was used effectively to calculate the formation dip in real time, which
assisted in maintaining position in the desired zone and minimizing borehole tortuosity through smaller
continuous corrections to the trajectory in the lateral (Fig. 17). The near-bit azimuthal gamma ray would
provide the first indication of exiting the zone through lateral reservoir changes, faulting, or through the
bounding formations, and the azimuthal resistivity would provide confirmation.

Figure 17—Landing and geosteering the well in the second target layer.
SPE-192898-MS 13

The formation dip ranged from approximately from 0.5° to 1.7° up-dip from the landing point to the TD
of the well (Fig. 18) which was accomplished in one bit run with 100% reservoir contact.

Figure 18—Real-time geosteering of the well.

High-Resolution Image Interpretation

The LWD high-resolution imaging sensor was included in the BHA primarily to aid in sourceless well
placement to facilitate real-time dip picks; it was also used to recognize reservoir sublayers in real time to
maximize the reservoir exposure in this horizontal well. A post-well image interpretation was initiated to
evaluate the presence of fractures.
The image interpretation included data quality control, symmetric picks of significant bedding planes,
and stratigraphic boundaries, together with assigned pick qualities. Fig. 19 and Fig. 20 show the graphical
representations and quality ratings, together with definitions of various symmetric picks that are used in this
interpretation, in accordance with the operating company standardization.
14 SPE-192898-MS

Figure 19—Image pick definitions in accordance with requested categories and definitions.

Figure 20—Image picks quality definitions.

In this interpretation, the highest assigned pick quality was 3, whereas the majority of picks were generally
assigned a pick quality rating of 5. This pick quality rating assignment was based on that, as with other image
interpretations, "ground truthing" is desirable for picked feature interpretations. A total of 207 bedding
SPE-192898-MS 15

planes with an average dip of 1.465° in the NW-SE direction (325.438° azimuth) were identified. Fig. 21
provides a bedding plane example.

Figure 21—Example of bedding plane.

After landing the well, the trajectory was moving parallel within the desired zone of interest, with no
natural fractures encountered to TD.
16 SPE-192898-MS

Conclusions
For unconventional reservoir exploration and development, an integrated workflow was used to better
understand the reservoir properties and to design the well construction to meet the associated challenges.
Benchmarking this reservoir with North America unconventional plays was difficult because it is a
completely different type of reservoir in terms of lithology, permeability, source layer distribution, and
thermal maturity. For new plays, it is recommended to drill a pilot well to collect relevant information to:

• Better select the sweet spot (landing layer)

• Understand the geomechanical stress regime for fracture design optimization

• Estimate well deliverability based on reservoir capacity and level of maturity

The key performance indicators for sweet spot selection primarily include the reservoir quality, TOC and
level of maturity, brittlness and stress regime, and GOR and volume of free/desorbed gas.
Geosteering in tight carbonate limited the number of sensors to be used for well placement; however,
azimuthal near-bit gamma ray information, along with high-resolution resistivity imaging, were an excellent
fit to geosteer in such challenging conditions. As a lesson learned, it is recommended to add a conventional
porosity tool for better petrophysical analysis of the lateral section and to optimize the fracture design
intervals. The use of pressurized sidewall core was instrumental in collecting a fluid sample in a nano darcy
reservoir; the application of a PVT analysis helped in fracture design optimization and production prediction.
Minifrac and DFIT analysis monitored by microseismic helped to calibrate and validate the geomechanical
model and to confirm the fracture behavior in the basin stress regime.
Adding to the fracture design, the integrated workflow helped to optimize the design of all well
construction aspects in terms of drilling and well placement, drilling fluids, drilling bits type/sizes,
completion design, cement program, data acquisition while drilling/wireline and flowback, and production
management.

Acknowledgments
The authors thank the management of ADNOC for permission to publish this paper. They also thank several
ADNOC and Halliburton staff for their efforts, support, encouragement, and valuable input during this case
study.