Академический Документы
Профессиональный Документы
Культура Документы
Aimen Amer The number of oil and gas wells drilled horizontally continues to increase as
East Ahmadi, Kuwait
operators strive to maximize contact with target formations, develop more-efficient
Filippo Chinellato completion programs and optimize recovery from complex geologic structures.
Milan, Italy
Structural steering, a method by which operators direct horizontal and high-angle
Steve Collins
Chief Oil & Gas LLC wellbore trajectories, integrates data from deep-reading LWD resistivity tools
Dallas, Texas, USA and high-resolution imaging devices to create structural models of the geologic
Jean-Michel Denichou conditions encountered by the drill bit. This technique allows drillers to correct
Sugar Land, Texas
wellbore trajectories in anticipation of structural changes ahead of the bit and helps
Isabelle Dubourg operators better understand the formations already drilled.
Clamart, France
Roger Griffiths
Petaling Jaya, Selangor, Malaysia The drilling of oil and gas wells today has little in continue to introduce new technologies that help
common with the early days of exploration when operators drill wells that produce longer, drain
Randy Koepsell wildcatters punched holes, sometimes based on the reservoir more completely and improve return
Denver, Colorado, USA seemingly random patterns, in hopes of on investment.
discovering untapped resources. Today, modern The ability to drill high-angle and horizontal
Stig Lyngra
drilling engineers access an array of technologies well trajectories has been one of the most signifi-
Saudi Aramco
to visualize the subsurface and then command cant changes in drilling in recent years. Although
Dhahran, Saudi Arabia
sophisticated downhole hardware to precisely horizontal wells typically average two to three
Philippe Marza target reservoir sections. After analyzing data times the cost of conventional vertical wells and
Aberdeen, Scotland from various sources, well placement engineers involve higher risks, the proportion of horizontal
can adjust wellbore trajectories while drilling to wells continues to increase (below). For exam-
Doug Murray maximize reservoir contact. Service companies ple, in the US at the end of 2012, 63% of the
Abu Dhabi, UAE
70
Iwan (Bob) Roberts
Dhahran, Saudi Arabia
60
Oilfield Review Spring 2013: 25, no. 1.
Horizontal wells in the US, %
10
0
1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012
Year
> Horizontal drilling in the US. The number of wells drilled horizontally, as a
percentage of the total number of wells drilled in the US, has increased
sharply in the past decade. (Data used with permission from Baker Hughes.)
14 Oilfield Review
Spring 2013 15
1,817 wells being drilled were classified as hori- geology are crucial components of successful equally well, which makes vertical wells more
zontal and another 11% were labeled directional drilling operations. attractive because of lower costs. In a recent
wells. Only 26% of the wells were classified as Technological advances in LWD formation study from a complex, naturally fractured reser-
vertical.1 A primary reason for this shift to hori- evaluation tools are also making it possible for voir, vertical wells produced with higher oil rates
zontal and high-angle wells is that they bring engineers to achieve well placement objectives and lower water cuts than horizontal wells.4 This
potential short- and long-term rewards that ver- more effectively. Tools that probe the formation phenomenon was a result of waterflood maturity
tical wells cannot typically provide. Through a some distance from the wellbore allow drilling and unique reservoir geology.
single wellbore, high-angle and horizontal wells engineers to visualize complex subsurface geom- Nevertheless, recent drilling and geosteering
can improve drainage, access discrete compart- etry. Engineers use real-time, high-resolution technology developments have allowed access to
ments in complex reservoirs, reduce interven- image data to define structural geometry and pro- resources for which conventional vertical or geo-
tion costs, improve efficiencies and provide actively adjust drilling programs. LWD measure- metrically drilled horizontal wells would not have
exposure to more of the reservoir. Despite their ments also help operators differentiate intervals worked as well. This article discusses LWD tools
higher initial costs, horizontal wellbores often that have superior production characteristics that provide data directly impacting drilling pro-
provide operators with a method for developing from those that may not be profitable. grams, software for visualization of subsurface
reservoirs that otherwise might not be profit- To optimize the horizontal drilling process, geometry and a workflow operators use to optimize
able. Horizontal wellbores are especially impor- well placement engineers have developed work- well placement through structural steering tech-
tant in unconventional reservoirs where their flows that help them in the quest to realize drill- niques. Examples from a gas storage project in
use has been a key enabling technology in devel- ing objectives. Even before the well is spudded, Italy, an unconventional reservoir in the US and a
oping shale resources.2 these workflows play a role in helping geologists thin carbonate reservoir in the Middle East dem-
These improvements in drilling technologies and engineers identify targets and develop realis- onstrate how operators are optimizing well place-
and practices have transformed the way operators tic trajectories that avoid unnecessary drilling ment and more effectively accessing reservoirs
approach developing resources.3 In the early days complications. While drilling, the well placement and resources.
of horizontal drilling, wells were constructed team can update models with real-time informa-
based primarily on geometric well plans. When tion utilizing a 3D approach. Dimensions of Drilling
some wells were not optimally placed in target These new technologies and workflows may Technological advances in drilling were key fac-
zones, results were often disappointing. Today, not always provide the answer. The critical ele- tors in the sharp rise in the number of horizontal
geoscientists and drilling engineers understand ment in determining which well type performs and directional wells in the 1990s. In 1986, there
that the term horizontal drilling is an best is the drilling team’s understanding of how were 41 horizontal wells drilled worldwide
oversimplification of the process, and accurate reservoir geology impacts long-term performance (below left).5 Just four years later, in 1990, there
well placement involves more than turning a well of the well. In many high-permeability reservoirs, were 1,190 horizontal wells drilled, the majority
90° from vertical. Maximizing reservoir contact particularly during the early phase of develop- of which were in Texas, USA. More than 20% of
and understanding subsurface geometry and ment, vertical and horizontal wells may perform those wells were concentrated in the Austin
Chalk formation. Activity in this formation epito-
mizes the evolution that has occurred in horizon-
Horizontal Wells Drilled Worldwide tal drilling.
1,600 1,570 The Austin Chalk is a low-permeability and
Outside North America 1,400 low-to-moderate–porosity reservoir. The first
1,400
Canada 1,250
wells were drilled in this formation in the 1920s.
1,200 United States 1,190 Production of oil and gas in commercial quanti-
1,020
ties relied on the well intersecting existing inter-
1,000 connected fractures. To enhance production
Number of wells
16 Oilfield Review
Offset Log Layer Geologic
Well Log Squaring Column Model Resistivity Log
Modeled
Measured
Horizontal Departure
Formation Formation markers The markers are linked to
resistivity identified surfaces in the geologic model.
Geologic Model
> Model, compare and update method of well placement. Log data from offset wells are used to construct geologic and tool
response models. Geologists generate geologic models by first squaring the original data, creating layer columns and labeling
formation markers and surfaces (top left). Formation markers and layering are propagated from offset well data to create a
geologic model (bottom left), which may also be enhanced with seismic data. A forward modeling program predicts how logging
tools, such as a resistivity tool, will respond to formation properties (red, top right). Once the model is generated, a well trajectory
(green, bottom right) is proposed to target specific reservoir layers. As the horizontal well section is drilled, the measured
response (blue, top right) observed from the well is compared with the modeled response. (Adapted from Griffiths, reference 7.)
Some experts have attributed the global applicable, of better completion quality. However, update and is a reactive drilling process. The sec-
growth in horizontal drilling to the successes and until a well is drilled, the model, which is a func- ond relies on estimating and extrapolating the
lessons learned in these and other Texas wells.6 tion of the type and quality of available data, orientation of bedding planes from formation
However, not all horizontal wells drilled in the remains an approximation. dip data, usually with azimuthal measurements
Austin Chalk formation experienced dramatic acquired while drilling. The third method relies
increases in production. Many operators failed to Well Placement Methods on deep-reading directional data for remote
take into account the complex nature of the for- Downhole hardware for drilling horizontal wells boundary detection to proactively adjust the well-
mation, which included compartmentalization has improved considerably over the past decades, bore trajectory to maximize reservoir contact and
and subseismic faults that divided the Austin and well placement methodologies and work- avoid exiting target zones.
Chalk into multiple isolated reservoirs. flows have been developed to capitalize on new In the model, compare and update method,
Today, operators use extensive reservoir mod- technologies and drilling techniques. With these the well placement team first generates a model
els to extrapolate surface and downhole data and improvements, well placement methods continue of logging tool responses based on expected for-
predict the formations that horizontal wellbores to evolve. Today, three complementary methods mations observed in offset log data (above).
will encounter. This approach attempts to iden- are generally used in well placement.7 The first Seismic data interpretations are included in the
tify rock of better reservoir quality, and where method is characterized as model, compare and analysis to help geologists estimate the location
1. Baker Hughes: “Interactive Rig Counts,” Investor Bennetzen B, Fuller J, Isevcan E, Krepp T, Meehan R, 4. Widjaja DR, Lyngra S, Al-Ajmi FA, Al-Otaibi UF and
Relations, http://gis.bakerhughesdirect.com/RigCounts/ Mohammed N, Poupeau J-F and Sonowal K: Alhuthali AH: “Vertical Cased Producers Outperform
(accessed February 13, 2013). “Extended-Reach Wells,” Oilfield Review 22, no. 3 Horizontal Wells in a Complex Naturally Fractured Low
2. Alexander T, Baihly J, Boyer C, Clark B, Waters G, (Autumn 2010): 4–15. Permeability Reservoir,” paper SPE 164414, presented at
Jochen V, Le Calvez J, Lewis R, Miller CK, Thaeler J and Williams M: “Better Turns for Rotary Steerable Drilling,” the SPE Middle East Oil and Gas Show and Conference,
Toelle BE: “Shale Gas Revolution,” Oilfield Review 23, Oilfield Review 16, no. 1 (Spring 2004): 4–9. Manama, Bahrain, March 10–13, 2013.
no. 3 (Autumn 2011): 40–55. Downton G, Hendricks A, Klausen TS and Pafitis D: 5. Kuchuk F, Nurmi R, Cassell B, Chardac J-L and Maguet P:
3. For more on the evolution and innovations in drilling “New Directions in Rotary Steerable Drilling,” “Horizontal Highlights,” Middle East Well Evaluation
technology: Felczak E, Torre A, Godwin ND, Mantle K, Oilfield Review 12, no. 1 (Spring 2000): 18–29. Review 16 (1995): 7–25.
Naganathan S, Hawkins R, Li K, Jones S and Slayden F: 6. Kuchuk et al, reference 5.
“The Best of Both Worlds—A Hybrid Rotary Steerable 7. Griffiths R: Well Placement Fundamentals. Sugar Land,
System,” Oilfield Review 23, no. 4 (Winter 2011/2012): Texas, USA: Schlumberger (2009): 10.
36–44.
Spring 2013 17
Real-Time Modeling of formation boundaries. The team may use 3D
200 visualization software that usually includes plan-
Gamma ray, gAPI
Modeled
ahead functionality to develop wellbore trajec-
Measured tory and drilling programs. Real-time data
acquired while drilling either validate the model
0 or are used to update it in response to the new
information (left). The directional driller can
Deep resistivity, ohm.m Shallow resistivity, ohm.m
1,000
then make changes in the wellbore trajectory
100 Modeled
based on the updated model.
Measured
10 The second well placement method requires
an understanding of the orientation and magni-
tude of the formation dip. After interpreting azi-
1,000
muthal data from wellbore images, well
100 Modeled placement engineers are able to estimate and
Measured extrapolate the orientation of the target bed or
10
formation. The bit is steered to remain within the
target. If the bit is no longer in the target reser-
X,450
voir layer, the LWD data can be used to determine
whether the bit has exited the top or the bottom
X,500
of the reservoir, and the directional drilling engi-
True vertical depth, ft
Well Intersecting Ascending Layer Image log Well Intersecting Descending Layer Image log
Top
Bottom
Top
> Well placement using formation dip data. Azimuthal log data in the shapes of smiles and frowns help well placement engineers determine bit corrections.
When a wellbore crosses a bedding plane, an azimuthal logging tool response indicates whether the wellbore is exiting an ascending or descending
geologic layer. When the wellbore cuts an ascending layer (left), the first contact with the formation is at the bottom of the hole (bottom left); when the bit
exits the layer, the last contact will be at the top of the hole. When the bit cuts an ascending layer, the data appear as a frown in the image. Conversely,
measurements from a wellbore that exits a descending bedding plane appear as a smile (right). The bit can be guided up or down based on these
interpretations to ensure that the wellbore remains in or reconnects with a target zone.
18 Oilfield Review
early warning of approaching changes in the
target and surrounding layers (right). This tech- X,606
nique works best when there is sufficient resis-
tivity contrast between the bounding layer and X,608
the target. Drilling programs that optimize
Spring 2013 19
Prejob Planning decisions (left). The interpreter picks the dis-
tance to boundaries, and the boundaries are dis-
played on an eXpandGST curtain section. Image
data from tools such as the MicroScope service
provide bedding dip, fracture information and
fault detection.
The eXpandBG module imports the LWD log-
ging data, and engineers generate an updated
Structural model Offset well logs
model that includes drilling polarity logs. Polarity
logs indicate whether the well is heading toward
the bottom or toward the top of a structure. The
Correlation and Formation Tops Well Placement Models
software next computes a true stratigraphic
thickness (TST) index; TST is related to the
thickness of the reservoir section. Well place-
ment engineers can compare structural dips
while the drilling progresses with those in the
original model and quickly identify anomalies.
Curtain section The software projects the structural dip away
from the well using stratigraphic horizons, and
the geologist can label formation tops and strati-
Real-Time Dip Analyses
graphic surfaces. Armed with this information,
the well placement team can determine whether
corrections are required and in which direction
to steer.
Two crucial elements for structural drilling
are LWD data that can be used to develop realis-
tic models and software that can provide a
robust solution describing the reservoir. Without
real-time data, engineers and geologists may
TST Profiling Well Placement have difficulty understanding the geometry of
the subsurface and accurately projecting where
the next step should be taken. Unfortunately,
engineers must often make decisions with insuf-
ficient data about complex reservoirs. Until
recently, the tools for resolving these complexi-
ties did not exist for LWD operations, but this is
no longer the case.
8. Borghi M, Piani E, Barbieri E, Dubourg I, Ortenzi L and presented at the 10th Offshore Mediterranean
Van Os R: “New Logging-While-Drilling Azimuthal Conference and Exhibition, Ravenna, Italy,
Resistivity and High Resolution Imaging in Slim Holes,” March 23–25, 2011.
20 Oilfield Review
Tools of the Trade Mud Pulse Telemetry Data Rate
Along with modeling software and drilling hard- 10 3
magnitude below those of wireline logging sys- > MicroScope tool. The MicroScope LWD logging tool provides four laterolog resistivities, four
tems, the primary source for imaging data has borehole image measurements and two toroidal resistivities. The tool also includes inclination and
been wireline tools. Recently, LWD data trans- azimuthal gamma ray measurements. High-resolution borehole images from the tool can be used to
mission systems and imaging tools have been define structural conditions. From these data, faults as well as fractures can be identified. Because
the tool provides high-resolution information at multiple depths of investigation (DOIs), natural
introduced that can replicate the capabilities of fractures can often be differentiated from shallow drilling-induced fractures. Toroidal resistivity
measurements are useful for determining drilling conditions and identifying formations at the bit.
Spring 2013 21
A bit resistivity measurement, derived from
FMI Images Depth, MicroScope Images two antennas at the bottom of the tool, is also
ft
available. One antenna acts as a transmitter and
the other as a monitor. Current flows out from the
X,333 bit and returns farther up the toolstring. The
drillstring below the antennas acts as an elec-
trode, and the measured current depends on the
formation resistivity and mud properties.
X,334
Other than the large amount of data needed
to provide images, one of the biggest challenges
of producing high-resolution images using LWD
X,335 tools is the conversion of time-based to depth-
based data. Traditional LWD measurements
are indexed to pipe movement observed at the
drilling floor. This technique is not adequate
X,336
for detection of small formation features
because drillpipe movement at the surface may
not reflect small tool movements downhole.
X,337 Scientists at Schlumberger have introduced a
new algorithm to derive local depth information
based on tool revolutions rather than observed
pipe movement.10
X,338
For this technique, high-resolution data,
along with magnetometer-based tool orientation,
are recorded versus time. These data can be
X,339 viewed as strips with a constant and known thick-
ness. Converting the time-based measurements
to a depth-indexed image requires precise esti-
mations of the azimuthal and axial position of the
X,340
sensors. As the tool advances, overlapping strips
> Wireline versus LWD imaging. Data from wireline imaging tools, such as are merged and then correlated to axial tool
the FMI tool (left), have been the standard, even though running these tools movement. The technique provides a high-resolu-
in horizontal wells can be time consuming and may result in added risk of
sticking. Recently introduced LWD imaging tools, such as the MicroScope tion depth match (next page, top). The images
tool, provide images (right) with a quality comparable to that of wireline are then transmitted to the surface with minimal
tools, often in real time or stored for surface retrieval. (Adapted from resolution degradation.
Allouche et al, reference 10.)
Well placement engineers also use measure-
ments with greater depths of investigation than
those of wireline logging tools to identify dis-
the tool provides images of the borehole compa- measurement of current is a function of the for- tances to top and bottom boundaries of reservoir
rable to those of wireline tools such as the FMI mation conductivity (and its reciprocal, resistiv- sections. These measurements help engineers
fullbore formation microimager. An added bene- ity). The buttons measure azimuthal resistivity in plan wellbore trajectories so they remain within
fit of the MicroScope tool is that it can provide 56 separate bins distributed around the borehole target intervals. The PeriScope bed boundary
high-resolution resistivity images at different circumference, and the orientation of the button mapper makes a 360° measurement and can
radial depths of investigation, which allows engi- measurements is determined with respect to the detect beds as far as 6.4 m [21 ft] from the bore-
neers to distinguish natural fractures from drilling- Earth’s magnetic field, which is measured with hole. Tilted receiver coils that have directional
induced fractures. an azimuthal orientation system mounted per- sensitivity can determine bed orientation. As
The tool uses toroidal antennas as transmit- pendicular to the tool’s axis. long as there is sufficient resistivity contrast
ters to send axial currents along the collar and The full array of measurements has depths of between target beds and those adjacent to the
into the formation for resistivity measurements. investigation approximately 2.5, 7.6, 13 and zones of interest, the PeriScope tool can provide
Two electrode buttons mounted at opposite sides 15 cm [1, 3, 5 and 6 in.], measured radially out- crucial information about the position of the
of the collar provide borehole coverage as the ward from the tool surface. These data can wellbore in the formation.11
tool rotates. The current leaves the tool surface, resolve bedding planes and features as small as Modern well placement requires more than
is directed through the conductive drilling fluid an inch. Although FMI images can resolve determining the location and orientation of the
into the formation and returns to the button elec- smaller features, which are useful for texture bit within a target zone. If faults are encoun-
trodes.9 Once corrected for borehole effects, the analysis and characterization of fractures, tered, well placement engineers may not have
MicroScope image data compare favorably with
FMI images (above).
22 Oilfield Review
Noncorrelated Correlated
Images Images
2,944
Correlation depth
2,945 Surface-measured
depth
2,946
Depth
Depth, ft
2,947
Image
strip
2,948
2,949
2,950
2,951
1,498 1,499 1,500 1,501 1,502 1,503 1,504 1,505
Time, 100 s
> Correlating high-resolution measurements to depth. LWD logging depths are referenced to pipe measurements taken at the surface. For most data, this is
an acceptable acquisition method. However, the accuracy of this method is not sufficient for high-resolution measurements. To compensate for
shortcomings of traditional depth measurements, engineers at Schlumberger developed a technique that uses overlapping strips from images (left) to
create an internal depth reference based on known fixed distances between sensor buttons on the tool. Correlation takes into account the mismatch
between tool movement downhole (middle, blue) and surface movement (black). The resulting correlated images (right) are much improved compared with
the noncorrelated images. (Adapted from Borghi et al, reference 8.)
Spring 2013 23
Based on the dips identified in the vertical
well, the formation was assumed to be gently dip-
ping to the NNW at 5°. The well was to be landed
Anticline in the target interval and follow this trend.
Structural Model Produced Using eXpandBG Processing
Syncline However, the lateral was actually landed at a fold
NW SE
axis where the structure immediately turned
down to the south at 24°. The well soon exited the
Vertical well 8°
reservoir section and crossed a fault, eventually
5° Fold axis reconnecting with the reservoir but in a section
25° that dipped in the opposite direction, north at
24° Marcellus Shale
25°. The well encountered a second fault and was
Horizontal well
then back in the target formation; drilling contin-
ued along the path indicated by the original
structural model. Unfortunately, because the for-
Marcellus Shale mation was dipping more steeply than had been
Gamma ray log
modeled, the well exited the bottom of the
Marcellus Shale earlier than expected.
Major fault line A review of the well path validates the need
Gamma ray
log X° Average dip and for real-time structural data while drilling. With
direction
only azimuthal gamma ray data available for
Highly interpreting the formation structure, the drilling
tilted
Asymmetrical anticline block Gently tilted block program did not produce an optimal well path.
Had imaging and deep resistivity data been
> Complex Marcellus Shale model. The original reservoir model (inset) was created by propagating acquired with LWD tools in real time, the resul-
formation tops picked from pilot hole data, and well placement engineers designed a well trajectory tant drilling and completion programs may have
based on the model. The actual structure was quite different. The lateral well was landed in what was been quite different.
thought to be a gently upward sloping (5° NNW) layer cake formation. However, before reaching One final step in the modeling process
horizontal, the well crossed a fold axis with steeply dipping beds, and then exited the target interval. involves validation with Petrel geologic recon-
Although the directional driller hoped that continuing to drill would help determine which way to steer
the well, the well soon crossed a fault; the fault block was lifted and tilted relative to the previous struction software. This suite enables restora-
section and was also dipping in the opposite direction. The well reconnected with the target interval tion and forward modeling of complex folded
only to cross a second fault. Fortunately, the well was still within the Marcellus Shale, and drilling and faulted geologic models. By simulating
continued close to the originally projected angle based on the pilot hole, from which it was assumed
mechanical rock behavior with a comprehensive
that the formation dipped 5° NNW. The well exited the bottom of the target earlier than expected
because the formation was dipping more steeply (8° NNW) than projected. The directional driller had set of boundary conditions, the software allows
little help determining how best to steer this well because only azimuthal gamma ray information was the user to analyze complex structures. The
available for guidance. Geologists later loaded image data, acquired with an FMI tool, into the software confirmed the viability of the complex
eXpandBG module of Petrel modeling software, which generated an interpretation that explained why
the well failed to encounter the reservoir as expected.
present-day interpretation (next page, top).
Newly acquired 3D seismic data also validated
the structural model.
Production from this well was classified as dis-
Constructed with data from the pilot well, the structural models using multiwell data for input. appointing compared with that in nearby wells.
original model indicated a syncline plunging to In this example, formation dip was computed Had the structural model been updated using
the northwest and an anticline plunging to the using the local curvature axes technique to pro- LWD images, the well path may have been modi-
southeast toward the toe of the proposed lateral duce the eXpandBG structural model.13 Solutions fied or perhaps redrilled based on the new model.
section. Engineers at Chief Oil & Gas proposed a using this technique may not be unique; however, Similarly, the four-stage stimulation design may
well trajectory based on this interpretation from the interpreting geologists may intervene and have been more effective (next page, bottom).
the vertical pilot well. adjust the solution to fit with other data such as Only Stages 2 and 3 were completely within the
The horizontal section was drilled using only 3D seismic interpretations. target zone. Stages 1 and 4 covered zones that
LWD azimuthal gamma ray data. After the lateral Geologists with Chief and Schlumberger were in the target for only half the interval.
was drilled, borehole images were acquired on wire- analyzed dip sequences in both the pilot well Additionally, a section of the Marcellus Shale at
line with an FMI tool. The interpretation provided and the subsequent horizontal section and clas- the heel of the well was not stimulated, although
by eXpandBG processing highlighted significant dif- sified structures using the local curvature axes it coincides with a highly stressed interval around
ferences between the original reservoir model and technique on a Schmidt plot to resolve the the fold where FMI data indicated the presence of
the geologic structure observed in the well. structural complexity. Contrary to the original natural fractures, which often enhance produc-
The eXpandBG approach solves for geometric model, the new model revealed three distinct tion in shale reservoirs. In this case, real-time
complexities without the need for extensive input sections: an asymmetrical anticline, a highly structural data may have resulted in a well trajec-
by the interpreter. The software can also create tilted block and a third section characterized as tory that contacted more of the target formation
gently tilted (above). and led to better well production.
24 Oilfield Review
Steering for Storage Step 1: After Deposition
A structural steering workflow that did use real-
time LWD data and eXpandBG processing was
recently employed in an underground gas storage
project carried out by Stoccaggi Gas Italia
(Stogit) SpA, the gas storage division of Società
Nazionale Metanodotti (Snam). The multifield,
Step 2: Early Stage of Deformation
multiwell project was developed with the techni-
cal contribution of Eni SpA specialists. The objec-
tive of the drilling program was to expose as Maximum
much reservoir section with optimal properties compresssion
Spring 2013 25
LWD Data
MicroScope
Resistivity Images
PeriScope Data
PeriScope
Resistivity
ROP
Gamma Ray
700 800 900 1,000
X,004
Shale layers
Planned trajectory
X,006 Upper lobe Drilled trajectory
X,008
True vertical depth, m
X,010
X,012
X,014
X,016
Lower lobe Possible subseismic fault,
apparent dip
X,018
700 800 900 1,000
True horizontal length, m
> Complex geology in an Italian gas storage well. The targets for this horizontal well were two sand lobes separated by shale layers. The original model
assumed a layer cake geology, and the drilling engineers developed a trajectory (green) to pass through the upper sand lobe, cross the shale layers and
terminate in the lower lobe. Instead of encountering continuous layers, the actual trajectory (blue) encountered a fault, entered an uplifted section below
the target sand lobes and crossed another fault before reconnecting with the lower sand lobe in a downthrown section. After drilling through a section of
the lower sand lobe, the well placement engineers turned the well upward and then it again crossed the shale layer and reconnected with the downward-
dipping upper sand lobe. Traditional petrophysical measurements—resistivity (Tracks 2 and 3), gamma ray (Track 4) and formation density (not shown)—did
not provide much directional guidance. In addition, had engineers used only DTB data from the PeriScope tool (red and blue dots, bottom), they would have
had difficulty determining directional adjustments. Image data from the MicroScope tool (top) allowed geologists to identify the faults and determine dip
direction and correctly steer the well. Without the two complementary measurement systems, engineers would have had difficulty determining in which
direction to steer the well after it crossed either fault. (Adapted from Borghi et al, reference 14.)
that the trajectory had crossed the upper lobe well would have been drilled through the shale hole and acquire petrophysical data using logging
and the middle shale sections and was approach- layers that separated the lobes. The updated tools on wireline. For evaluation wells, the log-
ing the lower lobe; however, geologists inter- trajectory, modified using real-time LWD data, ging program usually consists of resistivity, neu-
preted the image data as indicating the well had intersected both lobes and maximized wellbore tron and density porosity, elemental capture
crossed an unexpected fault and was now in an contact with the reservoir section. spectroscopy and nuclear magnetic resonance
upthrown section (above). (NMR) tools. Borehole image logs are run for
Drilling through the upthrown section con- Shale Drilling fracture identification and geologic characteriza-
firmed the fault interpretation, and eventually In a Niobrara Shale exploration well in northeast tion. Acoustic logs may be run for mechanical
the wellbore crossed a second fault and again Colorado and southeastern Wyoming, USA, engi- properties, which are used in fracture stimula-
encountered the reservoir section. Log data neers used the structural drilling workflow to tion design and wellbore stability estimations.
acquired after the trajectory crossed the second resolve complex geologic conditions in a shale Operators often include conventional coring for
fault indicated that the well was penetrating the well.15 The Niobrara Shale play is an upper the pilot wells to determine lithology and
lower sand lobe, and the drilling team made the Cretaceous calcareous shale that produces oil describe fractures. Data from the pilot holes are
decision to incline the bit to more than 90° and and gas. The shales are composed of argillaceous used to characterize the reservoir, define the ori-
reconnect with the upper lobe. limestones with interblended chalk, marl and entation of target zones and identify the optimal
The well crossed back through the shale lay- bentonite. Because of the rock’s low permeability depth for landing the lateral section.
ers and eventually reached a downdip section of and porosity, production is generally higher in 15. Koepsell R, Han SY, Kok J, Munari M and Tollefsen E:
the upper sand lobe. After crossing the first fault, zones with natural fractures enhanced by hydrau- “Advanced LWD Imaging Technology in the Niobrara—
Case Study,” paper SPE 143828, presented at the SPE
the original well trajectory would have missed lic stimulation. North American Unconventional Gas Conference and
the lower lobe entirely, and a large portion of the The typical development scenario for Exhibition, The Woodlands, Texas, June 14–16, 2011.
Niobrara Shale wells is to drill a vertical pilot
26 Oilfield Review
95°
Resistive Fracture
Fault
True Dip Angle
Bed Boundary
Fault Orientation
Microfault Orientation
–5°
Top
Bed Boundary
Dynamic Image
Fault
Bottom
Resistive Fracture
Microfault
Top
1,000
Deep Resistivity
ohm.m
Shallow Resistivity
0.1
Top
Static Image
Bottom
Top
150
Gamma Ray gAPI
0
Measured
X,940
X,950
X,960
X,970
X,980
X,990
Y,000
Y,010
Y,020
Y,030
Y,040
Y,050
Depth, ft
> Resolving complex geology. Image data from the MicroScope tool can be presented in dynamic (Track 2) or static (Track 4) mode. Formation dip (top,
Track 1) can be handpicked from images or computed from these data. The green tadpoles indicate the down direction of the dip, 0° to 360° clockwise
around each tadpole, which represents north-east-south-west-north. The magnitude of dip is also computed and can be read from the log. The magenta
tadpole indicates a fracture and provides its orientation. The image data can be presented in a wrap mode that simulates the horizontal well (inset).
Bedding planes (green), faults (magenta), open fractures (blue) and healed fractures (cyan) can be visualized as they appear in the horizontal wellbore.
(Adapted from Koepsell et al, reference 15.)
Logging objectives in the horizontal section design. Engineers identify zones with structural Traditional methods of acquiring image data
differ from those of the vertical pilot wells. complexity to help them keep the wellbore in the in the lateral section require that drillpipe-
Fracture population density, type and orientation reservoir unit or determine the best path to conveyed wireline tools be deployed. Geologists
are needed for stimulation design. Measuring the reconnect if the well exits the target. Depositional used these data to identify the presence of natu-
wellbore path and orientation are crucial, espe- variations can be determined with LWD tools, ral fractures and quantify their orientation and
cially when the well crosses in and out of reser- and geologists use these data to adjust models density. However, high-resolution image data
voir layers. Identifying faults and determining that extrapolate properties from the pilot well. from the MicroScope tool eliminate the need for
their location and orientation aid completion separate wireline logging runs (above).
Spring 2013 27
An operator in the Denver-Julesburg basin target layer known as the C bench. The image tures at an azimuth of 104° and with a length of
began a campaign to develop the Niobrara data indicated open fractures with strike popula- approximately 2,400 ft [730 m].
formation with multistage hydraulic fracturing in tions oriented NW-SE and mineralized fractures The MicroScope tool provided real-time high-
horizontal wells. From the vertical pilot well logs, striking NNE-SSW. To maximize intersection with resolution images for structural and fracture anal-
geologists were able to confirm the presence of a the natural fractures, the horizontal well section ysis. Engineers created 3D models using eXpandBG
was planned perpendicular to the natural frac- software, which helped them optimize well place-
Raw dips ment and design hydraulic stimulation operations.
N
W E
Vertical pilot hole The MicroScope image data were particularly use-
S
0 90 ful in revealing the complex structural setting. In
X,200
Open fracture
addition to numerous open and healed fractures,
Healed fracture geologists identified numerous faults, a missing
X,300 Fault section and structurally deformed beds (below).
Structural dip The planned well trajectory, developed from
X,400 pilot hole and surface seismic data, resulted in
the well crossing a fault and exiting the target
Niobrara formation
C bench Fort Hays zone into unproductive marl sections below the
X,500 Limestone target. The last half of the well was below the C
bench and was drilled mostly in nonreservoir
Structural dip Codell
X,600 siltstone quality ductile shale. After these data were ana-
Depth, ft
S
W
C bench 0
target
270 90
0
180
270 90
0 0 0
180
270 90 270 90 270 90
> Stepping out from vertical. Geologists can identify the location and orientation of bedding planes and faults in vertical wells and project them away from
the wellbore, but horizontal wells often encounter unexpected geologic geometry. In this vertical well section (top), geologists identified several geologic
sections, including the target reservoir C bench section, which is a mixture of chalk and marl bounded by ductile shales and unproductive chalks and is part
of the Niobrara formation. Well placement engineers developed a trajectory to follow the target, and directional engineers landed the lateral well in the C
bench (bottom, expanded section). Horizontal drilling proceeded for approximately 2,350 ft [716 m] and the well encountered structural geometry that
differed from geologists’ expectations. The well (black) crossed at least seven major faults (magenta lines). After the first set of faults, the C bench was
found to be upthrown, which positioned the well in the lowest part of the reservoir. As drilling progressed, the well crossed a fifth major fault and was
below the target formation and completely out of the reservoir. After geologists developed the new model of the horizontal well, the operator pulled back to
the first fault section and redrilled the horizontal section with an orientation (not shown) that carried the well above the original trajectory; this repositioning
allowed the well to remain in the target interval. Geologists can also use dip data to identify other features. The stereonet plots shown across five of the
intervals can be used to identify fracture and fault type and orientation. The far left plot shows the NNW by SSE orientation of healed fractures, probably
resulting from folding. (Adapted from Koepsell et al, reference 15.)
28 Oilfield Review
Resistivity Water
90-in. Array
Oil
0.2 ohm.m 2,000
X,725
X,750
X,775
> Pilot hole logs. An oil-bearing carbonate reservoir section was identified from openhole logs acquired in a pilot hole drilled at a 30°
angle. The pay zone (gray shading) was less than 20 ft [6 m] thick based on measured depth. Corrected for well deviation, the true
vertical thickness would be even less than shown. Fluid mobility, determined from NMR (Track 5) and MDT data (not shown), indicated
a small permeable streak within the interval. The well placement team proposed drilling a horizontal well that would be steered using
PeriScope and MicroScope tools. Directional drillers use petrophysical measurements to guide them in steering wells, but there was
insufficient variation in the gamma ray (Track 1), resistivity (Track 2) or porosity (Track 3) within the target zone for these to be used. A
high-resistivity caprock overlying the zone, however, provided a boundary layer for reference in steering the well.
which were also identified in image data. nomically. The carbonate reservoir consisted of a main producing horizon. The pilot hole was
Fracture stimulations were performed in stages, thin, permeable layer sandwiched between low- drilled as a 30° slanted well across the reservoir
and the stages were designed to target similar permeability intervals overlain by a thick, nonpo- section, and Saudi Aramco carried out an exten-
rock types identified from petrophysical data. rous carbonate caprock. sive data acquisition program that included cor-
Stimulation designs also included consideration The well was drilled in a mature giant field ing the full reservoir interval.
of local stresses that resulted from formation located in Saudi Arabia.16 Historically, this field Porosity and resistivity in the zone of interest
structural complexity. has produced mainly
Oilfield from two major carbonate
Review were fairly uniform; the operator used a CMR
reservoirs.SPRING 13 1980s, two smaller strati-
In the early combinable magnetic resonance logging tool to
Drilling Between the Lines graphic oilStructural Steering
accumulations Fig.discovered.
were 18 The identify the presence of movable oil (above). An
Unconventional resources may require approaches ORSPG
example well was13-STCTSTR
drilled in the18larger of these MDT modular formation dynamics tester con-
for drilling and completions that differ from those two reservoirs. The discovery was further delin- firmed that only a thin layer within the zone had
for conventional reservoirs, but conventional res- eated and tested by several vertical wells. The 16. Al-Suwaidi SH, Lyngra S, Roberts I, Al-Hussain J,
ervoirs can benefit from application of unconven- low-permeability reservoir contains good quality Pasaribu I, Laota AS and Hutabarat S: “Successful
Application of a Novel Mobility Geosteering Technique
tional solutions. Saudi Aramco used the real-time light oil with a relatively high gas/oil ratio. in a Stratified Low-Permeability Carbonate Reservoir,”
structural steering workflow with eXpandBG and In early 2012, Saudi Aramco drilled the first presented at the SPE Saudi Arabia Section Annual
Technical Symposium and Exhibition, Al-Khobar, Saudi
eXpandGST processing to access resources that oth- reservoir development well, deepening an exist- Arabia, May 19–22, 2013.
erwise would have been difficult to produce eco- ing dead producer originally completed in the
Spring 2013 29
Mobility
Formation
Dip
MicroScope
Images
Resistivity
Gamma Ray
Inclination
X,000 Y,000 Z,000
Drilling
Polarity
PeriScope Inversion
Top of reservoir
PeriScope Real-Time
Boundary Inversion
Curtain
Well trajectory
Top of reservoir
Integrated
Curtain
Section
> Well placement and job execution. Within the oil-bearing carbonate zone of interest, engineers confined the target to a narrow permeable streak (bottom,
yellow) bounded by lower-permeability oil-bearing layers (tan). The drilling objective was to guide the well maintaining a constant distance from the
high-resistivity, low-porosity caprock (green) overlying the reservoir. Resistivity (Track 4) and porosity (not shown) data exhibited little variation across the
interval. For guidance, engineers used PeriScope curtain data (Tracks 7 and 8) to maintain the DTB. Geologists also used MicroScope image data (Track 3)
to detect subtle changes in orientation and formation dip (Track 2). Well placement engineers proactively corrected the well trajectory based on polarity
data (Track 6, red indicates drilling up structure, green indicates drilling down structure). Because fluid mobility and permeability were the properties that
differentiated the target interval from the rest of the zone of interest, an FPWD tool was included in the LWD logging suite. Mobility measurements were
acquired at irregular intervals along the well (blue circles, Track 1), but after validating the presence of fluid mobility for approximately 1,700 ft [520 m],
engineers removed the FPWD tool from the string because of concerns about tool sticking. The well placement team steered the well for approximately
2,900 ft [884 m] (bottom, blue) and stayed within the narrow window throughout the interval.
30 Oilfield Review
good mobility and would produce oil. From the FPWD data were acquired for the first 1,700 ft Structural steering involves more tools and
log data, petrophysicists determined that the [520 m] and confirmed that the chosen path was requires more data for analysis than conven-
permeable layer was less than 10 ft [3 m] thick following the high-permeability streak. Each tional drilling; in addition, the costs of structural
and was positioned about 6 ft [1.8 m] below the FPWD mobility test required leaving the drilling steering are higher. But the answers provided by
high-resistivity caprock layer. Log analysts were assembly stationary for 20 min. Significant over- the tools and data to engineers and geologists
uncertain whether the zone with high mobility pulls began to occur after each mobility test, and have the potential to reveal better access to
extended farther out in the reservoir or was sim- the FPWD tool was removed because of opera- more of the reservoir, enhance recovery and pro-
ply a stratigraphic anomaly. tional concerns related to hole conditions and duce more hydrocarbons. Structural steering
Even if the zone extended into the reservoir, sticking. The remainder of the well was then may not be the answer for every well, but the
engineers knew that effectively producing from drilled using only DTB and TST data from opportunity to resolve the complexities of down-
such a small interval in the pilot well would be eXpandBG and eXpandGST processing to deter- hole geology offers operators a tremendous tool
difficult. Thus, they designed a horizontal pilot mine corrections to the wellbore trajectory. for enhancing resource recovery. —TS
producer to more effectively drain the reservoir. Images from the MicroScope tool helped estab-
Challenges included using real-time data from lish the formation dip and were a key input in the
LWD tools to verify the presence of the high- interpretation. The horizontal interval covered
mobility zone and stay within this narrow, high- approximately 2,900 ft [884 m] and remained
permeability window. Porosity and resistivity logs within a 4-ft [1.2-m] sweet spot window for the
provided little help in identifying the zone with entire interval.
the best mobility. The well confirmed that the high-permeabil-
The technical team determined that the best ity streak was not a stratigraphic anomaly and
course of action was to drill the well with a trajec- extended far out into the reservoir. The well was
tory that maintained a constant standoff or dis- tested after completion and produced at a rate of
tance from the overlying caprock. The standoff several thousand bbl/d. Further evaluation is
was based on distance to boundary (DTB) mea- ongoing, but early analysis confirms that because
surements computed from a PeriScope tool. The the well followed the high-permeability path,
team relied on true stratigraphic thickness (TST) resources were accessed that otherwise might
data to maintain a constant position relative to have been difficult to produce economically.
the caprock location. Well placement engineers
with Schlumberger were able to compute TST in Knowledge Is Power
real time using eXpandBG processing of formation At one time, horizontal drilling was an exercise in
dips picked from MicroScope images. Saudi geometry and drilling technology. However, as
Aramco personnel used these interpretations to well placement techniques and practices have
instruct the directional driller in the proper evolved, LWD tools have been introduced that
direction to guide the PowerDrive rotary steer- provide well placement teams with a better grasp
able system. of geologic and subsurface structural conditions.
Based on results from the pilot hole, NMR Integrating downhole data into modeling soft-
data were judged to be insufficient to identify ware provides operators with the ability to visual-
the zone with mobility. Consequently, an FPWD ize subsurface complexities. This knowledge
formation pressure while drilling tool was used gives operators powerful tools to modify drilling
to confirm that the wellbore trajectory remained plans, alter wellbore trajectories and optimize
in the high-mobility streak. To ensure that the completion programs.
wellbore followed subtle changes in dip and Service companies continue to add to the
direction, the geosteering staff used interpreta- assortment of LWD tools that may have been
tions from borehole images acquired with a considered impractical for the drilling environ-
MicroScope tool. ment in the past. Pressure sampling, downhole
Geologists created a 2D structural model seismic acquisition and acoustic logging devices
from pilot well data and forward modeled logging were once considered to be beyond the capabili-
responses for the LWD tools. The well placement ties of tools used while drilling. Just as these
team landed the well near the interval, steering services have been accepted by the industry,
the well stratigraphically upward to attain the high-resolution measurements that image the
required distance to the upper boundary. Once borehole and result in large amounts of data are
the data from eXpandBG processing confirmed now becoming available. Proper interpretation
the required trajectory, the well was drilled main- of these data has the potential to alter the way
taining the proper orientation (previous page). wells are drilled; such drilling is no longer based
primarily on geometry but optimized for down-
hole structural conditions.
Spring 2013 31