3D MULTICOMPONENT SEISMIC CHARACTERIZATION OF A CLASTIC RESERVOIR IN THE MIDDLE
MAGDALENA VALLEY BASIN, COLOMBIA.
by
c
Copyright by Antonio Jose Velasquez-Espejo, 2012 All Rights Reserved
A thesis submitted to the Faculty and the Board of Trustees of the Colorado School of Mines in partial fulfillment of the requirements for the degree of Master of Science (Geo-physics).
Golden, Colorado Date
Signed:
Antonio Jose Velasquez-Espejo
Signed: Dr. Tom L. Davis Thesis Advisor Golden, Colorado Date Signed: Dr. Terry Young Professor and Head Department of Geophysics
ABSTRACT
The main goal of this research is to characterize the combined structural-stratigraphic trap of the Tenerife Field in the Middle Magdalena Valley Basin (MMVB), Colombia. For the first time in Colombia the structural and quantitative interpretation of modern three-dimensional multicomponent (3D-3C) seismic imaging enables a geometric description, a kinematic interpretation of the structural styles, and the facies distribution of the reservoir. A seismic petrophysics work-flow to better achieve the seismic well-tie. Edited and check-shot calibrated P-wave sonic logs were obtained and coefficients of the Gardner and Castagna equations were calibrated to match the density and shear-wave velocity depth trends for the basin. Seismic modeling was performed to evaluate the PP and PS seismic response of the reservoir interval (Mugrosa Formation).
The structural interpretation methodology involves a 3D fault-correlation and horizon picking for both PP- and PS-PSTM data volumes. Geometric attributes such as coherence and curvature were used to enhance the structural discontinuities. The main unconformity of the Middle Eocene (MEU) was interpreted, and an attribute-assisted interpretation of the reservoir was conducted in detail. While P-wave data provided most of the structural interpretation, converted-wave data provide a better understanding of the faults.
Traditionally, compressive thrust-propagation folds and tectonic inversion have been con-sidered as the main mechanisms controlling the deformation in the MMVB. However, the new interpretation shown in this work provides a different structural concept that involves two major structural styles: 1. Under the MEU the Late Cretaceous and Early Paleocene de-formation, dominated by east-verging thrust and partially inverted Mesozoic normal faults, is preserved. Associated folds exhibit a north-south strike, and their structural development is controlled by a long-lived structural element that dominates the area (the Infantas Paleo-high). 2. North-east striking younger normal faults indicate younger local extension, that
affects the entire Cenozoic sequence. Normal faults are, in fact, the structural heterogeneities that most affect the geometry of the reservoir compartments in Tenerife Field. This normal faulting oriented oblique to the maximum horizontal stress, together with the associated folding, can arise from a left-lateral shear deformation that creates a local trans-tensional regime. Hence, the structure of Tenerife Field at the top of the Oligocene sandstones, can be described as a two-way closure anticline within a negative flower structure. In addition, Upper Eocene - Early Oligocene syn-tectonic deposits are also documented in this work, dating the last episode of deformation associated with the Infantas Paleohigh uplift.
The value of multicomponent data goes beyond the structural interpretation since it provides an independent seismic measure of shear-wave velocities for obtaining VP/VS ratios from interpretation and for performing elastic inversion. From the interpretation of both PP and PS data, the interval VP/VS ratio was computed for the entire Mugrosa Formation. Forward modeling of PS wave response showed that computing VP/VSratio from picking thin intervals may lead to erroneous values since it is not possible to interpret the same seismic events in both PP and PS data. Nonetheless, analysis of the full-waveform (dipole) sonic log together with Gamma Ray measured in the reservoir interval, showed that there is a close correlation between lithology and VP/VS ratio. VP/VS ≈ 1.85 is an effective upper bound to characterize sandstones from fine grained rocks. Further, a model-based elastic inversion of acoustic impedance and VP/VS ratio performed using the PP volume and the sonic logs available, allowed to find stratigraphic features in the Mugrosa and Esmeraldas formations. The attribute extraction from the inverted P-wave amplitude for both acoustic impedance and VP/VS ratio allowed the characterization of stratigraphic features, in particular some channel geometries that are interpreted as part of a meandering fluvial system (point bars and crevasse splays).
The lithological and petrophysical correlation of additional attributes from the elastic inversion and AVO is not reliable since there is no independent density, porosity, resistivity, permeability, etc., measurements to guarantee accurate and stable results; nonetheless VP/VS
analysis using multicomponent seismic in the MMVB shows significant promise. Therefore, the acquisition of critical log data with new well drilling as well as an additional multi-attribute analysis based on AVO and a joint PP-PS inversion are strongly recommended.
TABLE OF CONTENTS
ABSTRACT . . . iii
LIST OF FIGURES . . . ix
LIST OF SYMBOLS . . . xxi
LIST OF ABBREVIATIONS . . . xxii
ACKNOWLEDGMENTS . . . xxiii
DEDICATION . . . xxiv
CHAPTER 1 INTRODUCTION . . . 1
1.1 Location . . . 1
1.2 Antecedents of the Tenerife Field . . . 2
1.3 Research objectives . . . 3
1.4 Data available . . . 3
1.5 Thesis Preface . . . 5
CHAPTER 2 STRATIGRAPHIC AND STRUCTURAL FRAMEWORK . . . 7
2.1 Stratigraphy and basin evolution . . . 7
2.2 Middle Eocene-Pliocene stratigraphic units . . . 11
2.3 Structural styles . . . 17
2.3.1 Stress field . . . 21
2.3.2 Cenozoic syn-tectonic deposits . . . 23
2.4 Field characteristics: Geologic model of the Tenerife Field . . . 23
2.4.2 Reservoir characteristics . . . 25
2.4.3 Previous seismic interpretation . . . 27
CHAPTER 3 PP AND PS SEISMIC DATA . . . 32
3.1 Seismic data acquisition parameters . . . 34
3.2 Seismic data processing . . . 35
3.2.1 Pre-Processing and Post-stack migration . . . 35
3.2.2 PP-PSTM data volume . . . 39
3.2.3 PS-PSTM data volume . . . 42
3.3 Resolution of seismic data . . . 45
3.3.1 Lateral resolution . . . 45
3.3.2 Vertical resolution . . . 46
3.3.3 Wavelet extraction . . . 47
3.3.4 Wedge model and tuning effect . . . 51
3.4 PP vs. PS modes . . . 52
3.4.1 Garotta equation . . . 58
3.4.2 PP-PS registration . . . 58
CHAPTER 4 ROCK PHYSICS AND FORWARD MODELING . . . 63
4.1 Seismic-to-well tie . . . 64
4.1.1 Convolutional model . . . 64
4.1.2 Rock physics: Log editing and density-velocity relationships . . . 65
4.1.3 PP and PS synthetic seismograms and seismic character . . . 72
4.2 Non-zero offset forward modeling of PP and PS data . . . 79
CHAPTER 5 STRUCTURAL AND QUANTITATIVE SEISMIC
INTERPRETATION . . . 87
5.1 Structural seismic interpretation . . . 88
5.2 PP vs PS Structural interpretation . . . 97
5.3 Stress field and deformation kinematics . . . 105
5.4 Syn-tectonic sedimentation . . . 110
5.5 Seismic attributes . . . 113
5.5.1 Geometric attributes: Attribute-assisted structural interpretation . . 115
5.6 Vp/Vs Ratio and lithology discrimination from well-log data . . . 117
CHAPTER 6 SEISMIC INVERSION . . . 127
6.1 Review on seismic inversion theory . . . 128
6.1.1 Fundamentals . . . 129
6.1.2 Zoeppritz equation approximations . . . 131
6.1.3 Model-based inversion . . . 132
6.2 Pre-Stack (elastic) model-based inversion . . . 134
6.2.1 Inversion results . . . 137
6.2.2 Interpretation of seismic-derived stratigraphic features . . . 140
CHAPTER 7 CONCLUSIONS AND RECOMMENDATIONS . . . 153
LIST OF FIGURES
Figure 1.1 Digital Elevation Model of the north-west corner of South America (left) and the location of the MMVB (right) highlighted with a dashed blue line. Color bar depicting heights in meters above the sea level. This is a hinterland basin situated within a modern intermontane
valley. The small blue square is where the Tenerife Field is located. . . 2 Figure 1.2 Summary of the well log data available in the three wells of Tenerife
Field. SP: Spontaneous potential, IND: Induction, DT: P-wave sonic, Dipole: Full-waveform sonic log, GR: Gamma ray log. Notice that there is only dipole sonic and GR in the reservoir interval (Mugrosa
Formation, see description of the reservoir in 2.4.2). . . 4 Figure 2.1 Map of the Colombian Andes depicting regional digital elevation model
and major tectonic-geomorphic provinces. The Middle Magdalena Valley (MMV) is bounded by the Central Cordillera (CC) and the Eastern Cordillera (EC). (From ). Santamarta-Bucaramanga Fault (SBF) and Palestina Fault (PF) are the regional strike-slip systems
affecting the northern part of the MMVB. . . 9 Figure 2.2 Generalized stratigraphic column of Middle Magdalena Valley Basin
(MMVB) illustrating the main petroleum system elements. Highlighted red curved-lines are the Cenozoic unconformities, being the MEU the most important boundary that separated the transition between
environments of marine and continental regimes. . . 10 Figure 2.3 Summary of the MMVB evolution during Mesozoic and Cenozoic (after
). . . 12 Figure 2.4 Left: Schematic cross-section showing the Paleogene basin evolution of
the MMVB. Black arrows represent the distribution of sediment sources based on provenance analysis . Note the eastwards advance of the deformation front of the CC until Late Eocene when the EC emerges. The deposits of La Paz and Esmeraldas formations record the initial uplift of the EC, and pinch out onto the Infantas Paleohigh and on the CC. Right: Rose diagrams of paleocurrent data from Paleogene deposits showing the changes in the principal direction of the sediment supply. . . 13
Figure 2.5 Chronostratigraphic summary chart of Cretaceous and Cenozoic strata along the MMVB from showing the correlation among units with the most used lithostratigraphic nomenclatures. The MEU is in blue
(dashed line), and the projected position of the Tenerife Field in red. . . 14 Figure 2.6 Composite stratigraphic column illustrating main lithologic
characteristics of the MMVB Cenozoic deposits based on field
measurements and descriptions by Gomez et al.. Note the position of La Cira, Mugrosa and Los Corros fossil horizons, and the MEU. Highlighted in light yellow is the Mugrosa Formation, the main
reservoir at Tenerife Field. Modified from Gomez et al.. . . 16 Figure 2.7 Schematic geologic cross section of the MMVB illustrating the main
structural elements and tectonosequences defined by Suarez(1997). See localization of this section in Figure 2.1. Under the MEU boundary, Jurassic and Cretaceous sequences are affected by structures different from those affecting the Tertiary rocks above it. The wedge shape of the Cenozoic basin is controlled by the actual eastward tilt of the MEU. Structurally-projected Tenerife Field is in red as reference. Modified
from Suarez. . . 18 Figure 2.8 Regional structural section across the MMVB parallel to section shown
in Figure 2.7, see Figure 2.1. The EC foothills present a structural style dominated by a thrust-and-fold belt, while within the underlying
Cretaceous sequence in the middle of the basin, the styles are
characterized by faulted blocks and paleohighs. Structurally-projected
Tenerife Field is in red as reference. Modified after Gomez et al.(2005). . 19 Figure 2.9 Structural cross section in the northern MMVB, see Figure 2.1. Here,
unlike the sections of central MMVB (Figure 2.7 and Figure 2.8), the structure underneath the MEU is characterized by positive flower structures. The wrenching is the result of combined influence of the SBF and strike-slip faults in the San Lucas Range. Modified after
Gomez et al.(2005). . . 20 Figure 2.10 Paleostress analysis made by Cortes et al.. Above: Pre-Eocene
deformation phase is associated to E-W to WSW-ENE direction of maximum compression σ1; Below: Post-early Eocene phase is associated to NW-SE to WNW-ESE direction of σ1. The maximum horizontal
Figure 2.11 Seismic data illustrating syn-tectonic sedimentation in the Mugrosa Formation along two parallel seismic sections in the central MMVB. Growth strata can be observed particularly in the upper Mugrosa and Colorado formations. Left: Lisama Anticline. Right: Provincia
Anticline. Modified after Gomez et al.. . . 24 Figure 2.12 Correlation among the Tenerife wells and a reference well log taken
from La Cira - Infantas field, located at 20 Km NE-direction from Tenerife. Correlation among the Tenerife wells and a reference well log taken from La Cira - Infantas field, located at 20 Km NW-direction from Tenerife. Despite the units were identified virtually with the same thickness in all three wells, Mugrosa C looks more than double the
thickness in Tenerife-3. . . 26 Figure 2.13 Typical electrofacies and Vshale in Zone C of of Mugrosa Formation in
La Cira - Infantas Field. After Rojas. Notice that SP log cannot accurately read fining-upward sequences as GR log does. In this particular well the author reports a thickness of about 650 f t and a NTG ratio close to 50%, which contrast with the poor sand content of
the less than 250 f t reported in Tenerife-1 and Tenerife-2 wells. . . 28 Figure 2.14 Composite seismic line showing the structural interpretation of the
Tenerife Field. After Ecopetrol, 2003. Tenerife structure is interpreted here as an asymmetric anticline faulted by a set of south-east-verging thrust faults and associated back thrusts. Neither strike-slip faulting is considered nor growth strata can be identified from this seismic data.
Notice the normal-like faults in red. . . 30 Figure 2.15 Structural map at the top of Mugrosa Formation. After Ecopetrol,
2003. Tenerife is interpreted as an asymmetric anticline faulted by a set of south-east-verging thrust faults and associated back thrusts. Each
well appears to be drilled in independent fault blocks. . . 31 Figure 3.1 Map of Tenerife 3D-3C. The geometry is a regular square with sources
oriented NNE-SSW and receptors run nominally E-W in an orthogonal array. The total area of the survey is about 29 Km2. Notice from elevation model, left corner below, that major elevations (about 140 m
above sea level) are in the northern part of the survey. . . 33 Figure 3.2 Map depicting the distribution of sources and receivers in Tenerife
3D-3C. The sources are in red, and the receivers are in blue. Notice that the shot lines are more irregular than receiver lines due to logistic and environmental issues. The total area of the survey is about 29 Km2. A summary of the geometry acquisition parameters i shown in the right. . 36
Figure 3.3 Map showing shot-receiver lines with bin fold color overlay. . . 37
Figure 3.4 Tenerife 3D-3C: Final stack In-Line 6190. . . 40
Figure 3.5 Tenerife 3D-3C: Final post-stack migration In-Line 6190. . . 40
Figure 3.6 Tenerife 3D-3C: Pre-stack time migration In-Line 6190. . . 41
Figure 3.7 Example of In-Line 6190 comparing the the result after PSTM (a) and PSTM after post-migration processing (b). In c) the PSTM after post-migration processing with the RMS migration velocity overlayed. In d) there is an extracted velocity profile from the location of the X-Line highlighted in red. . . 43
Figure 3.8 Example of X-Line 2320 comparing the the result after PSTM (a) and PSTM after post-migration processing (b). In c) the PSTM after post-migration processing with the RMS migration velocity overlayed. In d) there is an extracted velocity profile from the location of the In-Line highlighted in red. The range of reservoir depths is marked with the red square. . . 44
Figure 3.9 Above: PP seismic data (left) and its power and phase spectra (right). In spite the maximum frequency is about 95 Hz, typically effective frequency achieved is about from 15 to 75 Hz. Below: Vertical resolution computed as either λ/4 or λ/8 using the range of velocities shown above. . . 48
Figure 3.10 Power spectrum of the PP data representative wavelet extracted from actual seismic data. The maximum frequency is about 95 Hz, typically maximum effective frequencies achieved are about 75 Hz; however the dominant frequency is just 25 Hz. . . 50
Figure 3.11 Power spectrum of the PS data representative wavelet extracted from actual seismic data. The maximum frequency is about 50 Hz, however the dominant value is just 13 Hz. . . 51
Figure 3.12 Thin-bed resolution and the classic wedge model of a low-impedance layer encased in a higher-impedance medium. The thickness of the wedge increases from left to right. A synthetic trace is generated at each horizontal position, using an 5/10 − 50/55 Hz Ormsby wavelet. The travel-time dominant period is 20 ms (50 Hz). The peak energy occurs when the wedge thickness is one-quarter of this effective wavelength, i.e., 5 ms (green line). . . 53
Figure 3.13 Tuning effect computation for PP and PS seismic data using the
statistically extracted wavelets. The actual tuning thickness is about 18 ms for P-wave data and 29 ms for PS, and the vertical resolution of
both PP and PS data is virtually the same. . . 54 Figure 3.14 PP and PS reflection and transmission of an incident P-wave on an
elastic boundary. P-waves are represented as green wavelets and
S-waves as purple wavelets. At an interface the P- and S-wave velocities change from VP 1, VS1 to VP 2, VS2, and an incident P-wave is partitioned into four modes, each one with a characteristic wavelength λ = Vf, given a frequency f from the source. Notice that the angle of reflection for the converted wave (PS) is smaller than the angle of incidence θP S ≤ θP. Also, as usually VP ≥ VS, PP-wave may arrive first to the receivers despite the distance traveled by this mode is longer than the
distance traveled by the PS. . . 56 Figure 3.15 Comparison between PS (left) and PP (right) after PSTM displayed in
PP two-way-time. The general structure can be delineated using both PP and PS data. Notice that steep events are imaged better in the PS data than in the PP data. Also notice that in the target zone, between the yellow horizon and the MEU (green horizon) the reflectivity looks
different in both data sets. . . 61 Figure 3.16 Detail of the registration in a In-Line near Tenerife-1 well. This
comparison between PS (left) and PP (right) displayed in PP-time shows positive amplitudes in red and troughs in blue. Notice the
squeeze effect on the PS data and the better resolution of the PP-wave. In spite the lower resolution PS-data allows to interpret high angle events, as reverse faults in the Cretaceous sequence (black arrow below
the MEU). . . 62 Figure 4.1 Cross-plot (right) depicting the correlation between the high frequency
component of the P-wave sonic and conductivity in the offset well. Each depth interval was chosen according to the low-frequency trends of constant slope (left). Notice the direct relationship in the cross-plot, although there is a considerable dispersion of data from the straight line. The linear regression, in the sense of least-squares, allows to
Figure 4.2 Calibration of Gardner’s coefficients using an offset well 40 km south Tenerife Field and calculated pseudo density log. The sonic log edited using the Burch methodology was used to compute density (yellow curve) using the “generalized” Gardner equation (4.6), to be compared with the actual bulk density measured in the well (green curve). The Gardner’s coefficients were changed and tested until converging on an
optimum result. . . 69 Figure 4.3 Processed -verticalized travel times- check-shot obtained from picking
first arrivals of the zero-offset VSP of Tenerife-1. The time-to-depth curve is shown on the left and the vertical interval velocities on the right. In yellow the reservoir interval is highlighted. Notice that the
velocity in the reservoir average a value of 13.000 f t/s. . . 71 Figure 4.4 Edited log data in Tenerife-1 well. From left to right: Check-shot
time-depth curve, SP and GR (acquired in the reservoir interval only), P-wave sonic, computed density using Gardner’s equation for
shaly-sandstones, acoustic impedance and reflection coefficient series. . . . 73 Figure 4.5 P-wave synthetic seismogram and seismic-to-well tie of Tenerife-3. The
right panel shows the synthetic trace in green over the actual PP-data. The left panel shows the time-to-depth curve (blue), the second track is depicting SP (filled in blues) and resistivity (filled in greens).
Calibrated P-wave sonic log corresponds to the red curve in the third track, computed density (pseudo density obtained from the calibrated Gardner equation) is shown as the blue curve in the track labeled as “Track seven”, while the last three curves are: computed acoustic impedance (magenta), reflection coefficient series (green) and the
pseudo-shear sonic (blue). . . 75 Figure 4.6 P-wave synthetic seismogram and seismic-to-well tie of Tenerife-3
zoomed on the reservoir interval. The extracted seismic wavelet used to perform this synthetic is depicted below (see also Figure 3.10). Notice the good correlation between the synthetic seismogram and the actual seismic data. The black arrow indicates the section containing the
reservoir sands within Mugrosa Formation. . . 76 Figure 4.7 Calibration of the Castagna coefficients in Tenerife-1 at the Mugrosa
Formation. On then left, VP versus depth from the dipole sonic, and on the left the green curve is the actual VS also from the full-waveform sonic log, while the red curve is shear-wave velocity computed from the “corrected” mudrock line. Velocities in f t/s. The constants to get this overlap between the actual shear-wave and the predicted shear-wave
Figure 4.8 PS-wave synthetic seismogram and seismic-to-well tie of Tenerife-1. Notice the lower frequency content with respect to the wavelet extracted from PP-wave data. The wavelet used is the same shown in Figure 3.11. . 79 Figure 4.9 PP Synthetic response at different frequencies for the reservoir interval
at Tenerife-1. Colored panels to the left represent the SP, GR and VP/VS ratio. The increasing frequency content from left to right is driven by the selected Ormsby wavelet, while the reflection coefficients are computed from the calibrated P-wave sonic log (red curve). The highest frequency Ormsby wavelet 5/10 − 70/75, right next to the sonic log, has similar frequency content of the extracted wavelet. The right
panel depicts the zero-offset synthetic overlying actual seismic data. . . . 82 Figure 4.10 PP Synthetic response at different frequencies for the reservoir interval
at Tenerife-1, with horizon interpretation. Colored panels to the left represent the SP, GR and VP/VS ratio. The highest frequency Ormsby wavelet 5/10 − 70/75, right next to the sonic log. The right panel depicts the zero-offset synthetic overlying actual seismic data. The interpretation of the main units and sandstone packages within the
reservoir are depicted in color dashed lines. . . 83 Figure 4.11 PS Synthetic response at different frequencies for the reservoir interval
at Tenerife-1, with horizon interpretation. The highest frequency
Ormsby wavelet 5/10 − 40/45. . . 85 Figure 4.12 PS Synthetic response at different frequencies for the reservoir interval
at Tenerife-1. The interpretation of the main units and sandstone
packages within the reservoir are depicted in color dashed lines. . . 86 Figure 5.1 X-Line 2401 with the well-seismic tie at the location of Tenerife-1. . . 89 Figure 5.2 Interpretation of the X-Line 2401 with the well-seismic tie at the
location of Tenerife-1. Notice the set well-defined set of normal faults above the MEU configuring the structural expression of a typical negative flower structure. The main faults compartmentalizing the
reservoir are called TF1 and TF2. . . 91 Figure 5.3 Interpretation of the X-Line 2401 with the geologic formations
highlighted. Notice the different structural styles across the MEU. . . 92 Figure 5.4 In-Line 6204 with well-seismic tie at the location of Tenerife-1 and
Figure 5.5 Interpretation of the In-Line 6204 with well-seismic tie at Tenerife-1 and Tenerife-2. Dips of the normal faults are smaller than in cross line direction. Likewise, deformation affecting Cretaceous rocks seem to be stronger. Notice that, in spite the differentiation of the structural styles across the MEU, the subtle anticline in the Cenozoic sequence follows the regional folding trend of the Cretaceous rocks. Notice also the truncations (red arrows) and onlaps (light green arrows) associated with the MEU. The main fault responsible for the uplift of the Tenerife paleohigh is named as TPF, while the back-thrust associated is called
TPFB. . . 95 Figure 5.6 3D structure map of the MEU in the Tenerife area. Locations of the
three Tenerife wells with a colored P-wave sonic log are depicted as a reference of the seismic well-tie. Notice the position of the Tenerife
Paleohigh. . . 96 Figure 5.7 Time structure map of the MEU in the Tenerife area. Locations of the
three Tenerife wells are displayed in blue. Faults are thick black lines
while the syncline axes are drawn in white. . . 98 Figure 5.8 Time structure map of the Mugrosa Formation in the Tenerife area.
Locations of the three Tenerife wells are displayed as white squares. . . . 99 Figure 5.9 PP (left) vs PS (right) of the In-Line 6190. Notice that the main faults
are well-defined in both volumes but steeper discontinuities are better
imaged in PS data (See for instance the black arrow). . . 101 Figure 5.10 PP (left) vs PS (right) structural interpretation of the In-Line 6190.
Notice that the Esmeraldas Formation reflector has a stronger character on converted-wave than in P-wave data. . . 102 Figure 5.11 Time structure map at the top of the Mugrosa Formation for both PP
(above) and PS data (below). . . 103 Figure 5.12 Isochron map of the Mugrosa Formation for both PP (above) and PS
data (below). . . 104 Figure 5.13 Regional stress field reconstruction based on inversion of fault slip data
(left, ) and Focal mechanisms from University of Harvard, CMT solutions (right, ). In both cases the solutions give a maximum horizontal stress oriented WNW-ESE. However, the plate velocity
Figure 5.14 Strike-slip kinematics after Wilcox et al.. Along a shear shear zone, in this case left-lateral (weight black line), the deformation is controlled by a simple shear mechanism creating a different set of shear planes (R and R0), as well as compressive and extensional structures according to the resulting ellipse of deformation. Notice that minor shear planes
called P and P0 form bounding the pure extension zone. . . 108 Figure 5.15 Time structure map of the Mugrosa Formation in the Tenerife area
depicting the kinematic interpretation of the faults, interpreted as a negative flower structure. The extension is parallel to the direction of the minimum horizontal stress σ3, while faults actually correspond to
shear planes R0 and P0 with both lateral and vertical -normal- offsets. . 109 Figure 5.16 Cenozoic syn-tectonic sedimentation in the Cenozoic sequence
evidenced by growth strata in seismic data. Notice the continuous change of thickness from the crest of the Tenerife paleohigh towards its flanks, as well as the “divergent pattern” of the reflectors in the basal Cenozoic, corresponding to Esmeraldas Formation and Zone C of the Mugrosa Formation (yellow zone). Likewise, in the upper Mugrosa
(green zone) growth strata is observed. . . 111 Figure 5.17 Forward modeling of growth strata in the forelimb of a kink-band fold
(below). Similar growth strata patterns can be found also in trishear folding for ratios of propagation to slip P/S > 1.0. Modified from
http://www.geo.cornell.edu/RWA/trishear/TSworks.html. . . 112 Figure 5.18 Map depicting the P-wave RMS amplitude on the MEU surface. Notice
that both structural and stratigraphic features can be extracted from
the amplitude. . . 115 Figure 5.19 Map depicting the P-wave RMS amplitude at the top of Mugrosa
Formation. . . 116 Figure 5.20 Coherence (variance) calculated over the entire PP data volume (right
panels) time-sliced at the PP travel-time highlighted with the thick blue line (left panels). Above the data without the interpreted faults
and below with the interpretation. . . 118 Figure 5.21 Coherence (variance) calculated over the entire PP data volume (right
panels) time-sliced at the PP travel-time highlighted with the thick blue line (left panels). Above the data without the interpreted faults
Figure 5.22 Cross-plot of VP vs VS along the reservoir interval (Mugrosa Formation) in Tenerife-1. The black lines across the plot are the theoretical
VP/VS = constant lines, which work as reference for VP/VS values. The pink line across the plot is the Catagna’s mud-rock line . The
highlighted yellow zone is an ellipse gathering values of Vsh < 0.25
which corresponds to a sandstone trend. . . 121 Figure 5.23 Cross-plot of VP vs VS along the reservoir interval (Mugrosa Formation)
in Tenerife-2. The pink line across the plot is the Catagna’s mud-rock line. The highlighted yellow zone is an ellipse gathering values of
Vsh < 0.25 that corresponds to a sandstone’s trend. . . 123 Figure 5.24 Cross-plot of VP/Vs vs ZP in the reservoir interval logged within the
Mugrosa Formation in Tenerife-1. The black line across the plot (VP/VS = 1.85) is the upper bound for clean sandstones. The
highlighted yellow zone is an ellipse gathering values of Vsh < 0.25 that
corresponds to a sandstone’s trend. . . 124 Figure 5.25 P-wave modeling of the reservoir interval within Mugrosa Formation
logged with the dipole sonic-GR combo. From left to right the gamma-ray index (v Vsh), VP/VS ratio, Poisson’s ratio and acoustic impedance are depicted. At least five sandstone bodies with
finning-upward electro-facies can be identified from the Vsh curve, and correlated across the other associated logs. Synthetic stacked for a range of offsets up to 6500 f t using Zoeppritz equations is shown in the right panel. . . 126 Figure 6.1 Low frequency acoustic impedance model based on the interpreted
horizons and edited well logs. . . 133 Figure 6.2 Actual acoustic impedance in one of the correlation wells (Tenerife-3),
and the acoustic impedance obtained from inversion, showing the accuracy of the inversion process. The left panel of this display shows an overlay of three impedance curves: the original impedance in blue, the initial guess model in black, and the final inversion result in red, while the right panel shows the synthetic traces calculated from this inversion result compared with the input seismic trace. The error in the third panel is computed between the actual and inverted seismic traces. 136 Figure 6.3 Model-based inverted P-wave data for acoustic impedance constrained
by well control. The acoustic impedance from wells is computed from P-wave sonic and pseudo-density logs computed following the rock physics work flow. Notice the detailed definition of both structural and stratigraphic features. . . 138
Figure 6.4 Model-based inverted P-wave data for shear impedance constrained by well control. The shear impedance from wells is measured only in the reservoir interval (dipole sonic). From Castagna’s equation a pseudo
shear velocity was obtained computed enabling the well-control. . . 139 Figure 6.5 Classical point bar model for a meandering stream, after Slatt. . . 140 Figure 6.6 Data slice at Esmeraldas Formation on the instantaneous frequency of
the acoustic impedance volume. . . 141 Figure 6.7 Data slice at Mugrosa Formation showing the RMS acoustic impedance.
. . . 142 Figure 6.8 In-Line 6202 (above) and X-Line 2307 (below) depicting the elastic
Inversion for VP/VS ratio. Color scale were selected to highlight in yellow the values of VP/VS ratio below the upper-bound limit
established for clean sandstones. Notice the lenticular geometry and poor lateral continuity of the sandstone packages. The reservoir zone can be well-defined as the “yellow interval” below the “no reservoir”
interval at the top of the Mugrosa Formation. . . 144 Figure 6.9 Map depicting the extraction of theVP/VS ratio amplitude on the top of
the reservoir interval within Mugrosa Formation, interpreted as a magenta horizon in the inverted VP/VS (left-lower corner). Notice that values of VP/VS ratio corresponding to sandstone facies are following channel deposits (Ch), point bars and laterally accreted point bars
(PB) and crevasse splay deposits (CS). . . 145 Figure 6.10 Time slice extraction of theVP/VS ratio amplitude at 1400 ms, magenta
line in the in the inverted VP/VS (left-lower corner). Yellow colors represent sandstones. A system of point bar deposits can be interpreted going from south-east to north-west of the seismic survey. . . 147 Figure 6.11 Time slice extraction of theVP/VS ratio amplitude at 1440 ms, magenta
line in the in the inverted VP/VS (left-lower corner). Yellow colors represent sandstones. Notice that the meandering pattern strikes
roughly from east to west at this depth. . . 148 Figure 6.12 Time slice extraction of theVP/VS ratio amplitude at 1480 ms, magenta
line in the in the inverted VP/VS (left-lower corner). Yellow colors
Figure 6.13 Time slice extraction of theVP/VS ratio amplitude at 1520 ms, magenta line in the in the inverted VP/VS (left-lower corner). Yellow colors represent sandstones. Sandstones start being controlled by the presence of the Tenerife Paleohigh. . . 150 Figure 6.14 Time slice extraction of theVP/VS ratio amplitude at 1560 ms, magenta
line in the in the inverted VP/VS (left-lower corner). Yellow colors represent sandstones. Basal sandstones of the Esmeraldas Formation
are totally controlled by the presence of the Tenerife Paleohigh. . . 151 Figure 6.15 Time slice extraction of theVP/VS ratio amplitude at 1600 ms, magenta
line in the in the inverted VP/VS (left-lower corner). Yellow colors represent sandstones. Basal sandstones of the Esmeraldas Formation
LIST OF SYMBOLS
P-wave velocity . . . α,Vp S-wave velocity . . . β,VS Phase angle . . . φ Angular frequency . . . ω Average velocity . . . Vavg Interval velocity . . . Vint RMS velocity . . . VRM S Incidence angle . . . θ Density . . . ρ S-wave Reflection Angle . . . ϕ ≡ θP S Reflection Coefficient . . . R P-wave Reflection Coefficient . . . RP S-wave Reflection Coefficient . . . RS Acoustic impedance . . . ZP Shear impedance . . . ZS
LIST OF ABBREVIATIONS
Middle Magdalena Valley Basin . . . MMVB Eastern Cordillera . . . EC Central Cordillera . . . CC Middle Eocene Unconformity . . . MEU
ACKNOWLEDGMENTS
Firstly, I would like to thank the Department of Geophysics of Colorado School of Mines. Particularly my advisor Dr. Tom Davis for his support, encouragement and guidance. Also thank you for the opportunity to be part of the Reservoir Characterization Project, which has made this research experience possible. Besides my advisor, I would like to thank the rest of my distinguished committee: Dr. Walt Lynn, Dr. Steve Sonnemberg and Dr. Scott MacKay for their guidance and questions. I would also like to thank Bob Benson for his technical assistance in data management. I would like to express my special gratitude to William Agudelo from Ecopetrol S.A. for providing the seismic data sets, well data, and continuous feedback towards my research.
I would like to say thank you to all of my friends and fellows at CSM. Working with everyone has been a enriching experience and I have truly enjoyed sharing both knowledge and fun times with all of you. I would like to acknowledge all of the professors that I have had the opportunity to learn from that have made my journey at Mines not only very challenging but also a rewarding experience.
Finally, specially I would like to thank my lovely and patient girl friend, my family and friends for their help and encouragement during this unforgettable journey.
CHAPTER 1 INTRODUCTION
Exploration in Sub-Andean basins in Colombia has been traditionally challenging due to the complex geology, rough topography and inaccessible areas. Operating in these con-ditions has motivated the use of new technology to reduce exploration risk. Nowadays, the application of new seismic technologies (e.g. new imaging techniques) in these environments beginning to pay dividends. In particular, the research that has been done into advance 3D-3C multicomponent reflection seismology for nearly three decades, utilizing both P-wave and S-wave wavefields, holds promise for application in Colombia. Evaluation of the application of the multicomponent seismic technology is the main focus of my research.
Ecopetrol S.A. acquired a 3D multicomponent seismic data set in the Tenerife Field in the Middle Magdalena Valley Basin (MMVB) (Reyes-Harker et al., 2010). The purpose of this research is to conduct an evaluation of the value and implications of using this new data set in terms of structural and quantitative interpretation in the Tenerife Field, and proposes a new structural model as well as a seismically-derived characterization for the field.
1.1 Location
The Middle Magdalena Valley Basin of Colombia is located in the central part of the country between latitudes 5oand 8onorth and longitudes -73oand -75o west (Figure 1.1). While the total area is about 30,000 km2. The elevations above the sea level range from 50 m to more than 2000 m in the foothills. The MMVB is one the most prolific petroleum basins of Colombia with a long history of hydrocarbon exploration that started with the discovery of a giant oil field called La Cira-Infantas, in 1918. The oilfields in the basin occur mainly as either structural or stratigraphic traps in Tertiary clastic reservoirs.
Figure 1.1: Digital Elevation Model of the north-west corner of South America (left) and the location of the MMVB (right) highlighted with a dashed blue line. Color bar depicting heights in meters above the sea level. This is a hinterland basin situated within a modern intermontane valley. The small blue square is where the Tenerife Field is located.
1.2 Antecedents of the Tenerife Field
The Tenerife Field, operated by Ecopetrol S.A., is located in the central MMVB, about 20 km south-west of the giant La Cira - Infantas Field. This field was discovered in 1971 after positive results of drilling the Tenerife-1 well. The appraisal strategy was followed by the drilling of wells Tenerife-2 and Tenerife-3. The development of the field was stopped later due to failed results in Tenerife-3. Tenerife-1 and Tenerife-2 had production of about 100 STBO/day of 22.8 API crude oil; however they are almost totally depleted today (Ecopetrol, 2003; Sandoval, 2009).
Tenerife Field produces from a combined stratigraphic and structural trap in Oligocene sandstones. The trap consists of an asymmetric, east-verging, faulted anticline located on the western flank of a major anticline structure. The anticline is also affected by some inverse faults and back thrusts (see Section 2.4.3). Thickness of the reservoir interval ranges from 18 to 44 f t, while the depths range from 7200 to 8000 f t (Sandoval, 2009).
1.3 Research objectives
The main focus of this work is to perform a 3D structural interpretation of the multi-component seismic data taking advantage of the improvements in the image given by 3D seismic and the independent source of information provided by the converted-wave data.
A secondary objective is to analyze the spatial distribution of the sand bodies within the reservoir. In fact, the main purpose of the high density design of this 3D-3C data is the mapping of complex patterns associated to fluvial channels. To fulfill the objectives, three complementary approaches were carried out: 1) horizon-derived seismic attributes on both PP and PS interpreted data volumes; 2) rock physics relationships for lithology discrimination; 3) elastic seismic inversion to evaluate the improvements of the quantitative interpretation for identifying stratigraphic features.
1.4 Data available
Well-log data from the three wells Tenerife-1, Tenerife-2 and Tenerife-3, include spon-taneous potential (SP), resistivity and induction logs. Moreover, some well-log data were acquired recently. Indeed, P-wave sonic in all three wells, and a Dipole sonic log with GR (dipole sonic-GR combo) was measured only within the reservoir interval in the wells Tenerife-1 and Tenerife-2. Figure 1.2 shows a summary of the well data available for this project. In addition, in order to tie the well data to seismic volumes PP and PS, two short offset and two long offset VSP’s were recorded. These data allowed for comparison of the corridor stack with reflectors in the wells Tenerife-1 and Tenerife-2.
It is important to mention that due to the scope of this research particularly the lack of density, resistivity and neutron-porosity logs, as well as the lack of complete dipole sonic and GR logs, make the interpretation of seismic-derived attributes and inversion results in terms of lithology and fluid content discrimination more difficult. In order to better deal with this limited information a seismic petrophysics analysis work-flow was implemented (see Section 4.1.2).
Figure 1.2: Summary of the well log data available in the three wells of Tenerife Field. SP: Spontaneous potential, IND: Induction, DT: P-wave sonic, Dipole: Full-waveform sonic log, GR: Gamma ray log. Notice that there is only dipole sonic and GR in the reservoir interval (Mugrosa Formation, see description of the reservoir in 2.4.2).
The 3D-3C seismic data, including stacks, post- and pre-stack migrations and pre-stack gathers of both PP and PS volumes were provided to this research project. This survey is the first one of its type and the densest program ever designed for a deep objective in Colombia. The source parameters were defined using a 6 km long experimental line, while 3-component accelerometers were used to record the data. Details of the design and processing will be covered in Chapter 3.
Ecopetrol S.A. provided the well and seismic data available for this research. 1.5 Thesis Preface
This thesis work was condensed in seven (7) chapters, starting with this introduction (Chapter 1) and ending with some relevant conclusions and recommendations (Chapter 7). Chapter 2 is devoted to careful, although general, treatment of the most important structural, stratigraphic and historic characteristics of the MMVB. After giving an explicit description of the basin evolution and stratigraphic framework, I focus on the structural styles affecting the central part of the basin, the regional stress field, and start the main discussion of this work: What is the most reliable structural interpretation of this type of structures such as Tenerife Field. Whereas most of the chapter is devoted to pure geological descriptions of the basin, I also analyze in the latter part the characteristics of the Tenerife Field field, including the reservoir stratigraphy and the previous interpretation.
In Chapter 3 I focus on the multicomponent seismic data. This chapter starts with the description of the acquisition parameters and PSTM processing of the Tenerife 3D-3C survey. To gain insight into the possibility of interpreting stratigraphic features using both PP-data and converted-wave data, a detailed discussion about seismic resolution is carried out including analysis of amplitude spectrum of the wavelet, wavelet extraction techniques, wedge modeling and tuning effect computation. Finally, the mode conversion theory is briefly introduced, and the PP and PS data volumes are compared in order to evaluate the quality of the image to perform the PP-PS registration.
Chapter 4 is all about the rock physics treatment given to well-log data with two major purposes: 1) to achieve a better seismic-well tie; and 2) to make a rock physics model to correlate the quantitative interpretation of the inversion results (Chapter 6) with actual well-log responses.
Chapter 5 explores the world of the structural seismic interpretation using both PP and converted-wave data sets. A new vision of the structural style is described, and an up-date with a better understanding of the field is discussed. In addition, the new structural framework is analyzed in terms of regional kinematics, and a close relationship between de-formation and sedimentation is identified through syn-tectonic deposits within the reservoir. Although this work is largely focused on the structural interpretation of the multicom-ponent data, the topic of quantitative interpretation, in this case based on seismic attributes and elastic inversion, is also covered in detail in Chapter 6. At the end of this chapter,VP/VS ratio is used to identify stratigraphic features.
CHAPTER 2
STRATIGRAPHIC AND STRUCTURAL FRAMEWORK
The MMVB is an eastward-tilted intermontane basin with a complex history. A series of tectonic events, make the MMVB a poly-historic basin, following the term from the definition stated by Kingston et al. (1983). The Central Cordillera (CC) and Eastern Cordillera (EC) of the northern Andes form the topographic flanks of the north-trending Magdalena Valley Basin (Figure 2.1). The basin’s name is taken from the Magdalena River that flows between these two mountain ranges from the south to the Caribbean sea. The MMVB overlies con-tinental basement composed of granulite-grade metamorphic rocks overlaid by low-pressure meta-sedimentary rocks as stated by Restrepo (1995), which also constitutes the crystalline basement of the CC and EC.
2.1 Stratigraphy and basin evolution
During the Early Mesozoic the area was controlled by an extensional setting evidenced by an intracontinental rifting related either to the break up of Pangea (Pindell & Dewey, 1982; Cediel et al., 2003), or due to a back-arc extension behind a subduction-related magmatic arc (McCourt et al., 1984; Cooper et al., 1995; Sarmiento-Rojas, 2001). Rifting produces syn-rift infill represented by sequences of fluvial and lacustrine sedimentary rocks during the Triassic-Jurassic (Jordan, Giron and Santos formations). These continental deposits consist mainly of red beds and volcanic effusive and pyroclastic deposits, though some marine facies appear locally (Mojica et al., 1996). Syn-rift sedimentation has been documented in several isolated outcrops in the Upper Magdalena Valley Basin and EC (Toussaint, 1995; Mojica et al., 1996; Kammer & Sanchez, 2006; Sarmiento-Rojas, 2001), but not in the MMVB. Recently, syn-rift sequences associated with inverted normal faults have been interpreted from seismic reflection data (e.g. Garavito, 2008; Parra et al., 2012).
Basin sedimentation continued into the Cretaceous in a back-arc setting east of the An-dean subduction zone, dominated by shallow-marine sedimentation (Cooper et al., 1995). By Aptian time the pure extension terminated and a post-rift sedimentation phase began con-trolled by thermal subsidence (Villamil, 1993; Etayo-Serna, 1994; Sarmiento-Rojas, 2001). Thus, syn-rift deposits are overlain by marine limestones and shales deposited during the Early Cretaceous (Rosablanca and Paja formations, respectively). During this phase the following formations were also deposited (Figure 2.2): Tablazo, which consists of limestones interbedded with black shales; Simit´ı, mainly made of shales deposited in an inner to middle shelf; La Luna, that consists of interbedding of limestones and black organic-rich shales; and Umir, that marks a transition into a more clastic environment and it is composed of gray shales and shaly-sandstones with coal intervals. On a broad scale, the marine sedimenta-tion of Cretaceous rocks represent a major transgressive–regressive cycle with a maximum flooding surface close to the Cenomanian–Turonian boundary, corresponding to the maxi-mum Mesozoic eustatic level (Villamil, 1993), producing an excellent regional source rock (La Luna Formation).
The complex evolution during the Cenozoic has been attributed to earliest foreland-basin conditions related to uplift of the Central Cordillera (CC) during the Late Cretaceous–Early Cenozoic (Dengo & Covey, 1993; Cooper et al., 1995; Gomez et al., 2005), and to trans-pression caused by the oblique accretion of the Western Cordillera (WC) until Early Eocene (McCourt et al., 1984; Cooper et al., 1995). Hence, the accretion produced an uplift of the CC causing the first phase of inversion of Mesozoic grabens which continued throughout the Cenozoic (Gomez, 2001). Thus, marine sedimentation slowly shifted into a continen-tal environment. The Lisama Formation, overlies the Maastrichtian shallow marine Umir Formation with a transitional contact and records regressive sedimentation in deltaic and al-luvial plains (Gomez, 2001), and therefore the first unit with continental sediments after the Cretaceous marine-dominated deposits. The major deformation phase of the CC affecting the MMVB marks a regional unconformity called the Middle Eocene Unconformity (MEU).
Figure 2.1: Map of the Colombian Andes depicting regional digital elevation model and major tectonic-geomorphic provinces. The Middle Magdalena Valley (MMV) is bounded by the Central Cordillera (CC) and the Eastern Cordillera (EC). (From Moreno et al., 2011). Santamarta-Bucaramanga Fault (SBF) and Palestina Fault (PF) are the regional strike-slip systems affecting the northern part of the MMVB.
It is the most important boundary within the Cenozoic record and can be observed through-out the basin (Figure 2.2). As inferred from thermochronologic data, the MEU separates a pre-middle Eocene deformation from a post Middle Eocene continuous deformation with thrust-controlled exhumation (Parra et al., 2009; Mora et al., 2010).
Figure 2.2: Generalized stratigraphic column of Middle Magdalena Valley Basin (MMVB) illustrating the main petroleum system elements. Highlighted red curved-lines are the Ceno-zoic unconformities, being the MEU the most important boundary that separated the tran-sition between environments of marine and continental regimes.
Subsequent establishment of a meandering fluvial system is recorded in strata of the La Paz Formation during Middle Eocene. According to Dengo & Covey (1993) and Cooper
et al. (1995) the sedimentation in the MMVB and EC from Eocene to Miocene occurred in a classic foreland basin. Nevertheless, sedimentary record and paleocurrents have shown a transition from a foreland to hinterland basin configuration in the uppermost Eocene strata of the lower Esmeraldas Formation (Moreno et al., 2011) (Figure 2.4). Hence, the alluvial to fluvial deposits of La Paz and Esmeraldas formations record the initial uplift of the EC. Both units show a systematic westward thinning that represents progressive onlap onto the ancient CC. Changes in plate tectonic motions documented in the Late Oligocene to Early Miocene produced the reactivation of the Middle Eocene structures and created an upper Oligocene unconformity (Schamel, 1991). In accordance with this tectonic setting meandering fluvial systems were developed and the Mugrosa and Colorado formations were deposited. In these units growth strata have been well documented (Gomez et al., 2005; Mora et al., 2010; among others), implying syn-tectonic sedimentation associated with an episode of continuous deformation due to the inversion of the EC (Parra et al., 2009; Mora et al., 2010). The Figure 2.3 shows a cross section with the schematic evolution described so far.
Furthermore, a collision of the Panama-Choco island arc with the northwestern margin of South America also occurred during the Middle Miocene. The fluvial to alluvial Real Group represents the last sedimentary record associated with the inversion of the EC. 2.2 Middle Eocene-Pliocene stratigraphic units
This research is primarily focused on the modern structural configuration of the Tenerife Field and the geophysical characterization of Cenozoic reservoirs. Therefore, the Cenozoic formations need to be described in some detail. Cenozoic stratigraphic record changes within the MMVB. For instance, (Figure 2.5)shows a chronostratigraphic correlation of the units across the basin in a north-south cross section.
Various authors, working with different fossil groups, reached conflicting interpretations of the ages of the Middle Eocene to Middle Miocene La Paz, Esmeraldas, Mugrosa, and Colorado formations. The stratigraphy can be constrained based on identification of some
Figure 2.3: Summary of the MMVB evolution during Mesozoic and Cenozoic (after Suarez, 1997).
Figure 2.4: Left: Schematic cross-section showing the Paleogene basin evolution of the MMVB. Black arrows represent the distribution of sediment sources based on provenance analysis (Moreno et al., 2011). Note the eastwards advance of the deformation front of the CC until Late Eocene when the EC emerges. The deposits of La Paz and Esmeraldas formations record the initial uplift of the EC, and pinch out onto the Infantas Paleohigh and on the CC. Right: Rose diagrams of paleocurrent data from Paleogene deposits showing the changes in the principal direction of the sediment supply.
fossil markers, namely: “Los Corros”, “Mugrosa” and “La Cira” fossil horizons, located at the top of the Esmeraldas, Mugrosa and Colorado formations, respectively (Figure 2.6). These horizons consist of thin layers of packstones of bivalves and gastropods within muddy intervals, mainly fresh-water mollusks (Nuttal, 1990 in Gomez et al., 2005). It is important to enhance the fact that the tectonic setting described in Figure 2.4 most likely implies fluvial sands with poor lateral continuity.
Alternatively, Suarez (1997) defined a chrono-stratigraphic framework that allows the correlation of tectono-sequences, bounded by unconformities, instead of litho-stratigraphic units Figure 2.7. Discussing either approach is not the scope of this research; therefore, as the nomenclature described below works fine at the scale of the analyzed seismic data, it shall be used throughout the text.
Figure 2.5: Chronostratigraphic summary chart of Cretaceous and Cenozoic strata along the MMVB from Caballero et al., 2012 showing the correlation among units with the most used lithostratigraphic nomenclatures. The MEU is in blue (dashed line), and the projected position of the Tenerife Field in red.
Gomez et al., 2005 pointed out that most of the MMVB Cenozoic sandstones are litharen-ites and sublitharenlitharen-ites with the exception of the Mugrosa and Colorado, which have pre-dominantly feldspathic sandstones. The components of the Cenozoic sandstones are grains of polycrystalline and monocrystalline quartz, metamorphic fragments (schist, micaceous gneiss, and phyllite), sedimentary clasts (chert and mudstone), feldspar grains, and scarce igneous fragments (Mesa, 1995).
• La Paz Formation (Middle Eocene): This unit overlies the MEU and consists of al-ternating coarse-grained to conglomaratic sandstones and gray claystones. This unit was deposited in alluvial fans in the western part of the basin (Medina et al., 1992 in Suarez, 1997). The major development of sandstones occurred in the eastern flank of the Nuevo Mundo Syncline where La Paz amalgamated litharenites mark the position of the main fluvial channels, which wandered within areas of maximum subsidence of the alluvial plain (Gomez et al., 2005). Polycrystalline quartz and sedimentary grains usually decrease upward into the upper La Paz, and they continue decreasing upward into the sandstones of the Esmeraldas Formation.
• Esmeraldas Formation (Late Eocene): It is composed primarily of thick layers of mud-stones and siltmud-stones, deposited in floodplains of a meandering system (Rubiano, 1995 in Suarez, 1997). Eventually, this unit also includes layers of fine-grained, trough-cross bedded sandstones. Most of the sandstones are litharenites and sublitharenites (Mesa, 1995).
• Mugrosa Formation (Oligocene-lower Miocene): Unconformably overlies the Esmeral-das Formation. Formally defined as the rock interval in between the “Los Corros” and “Mugrosa” fossil horizons, and consists of inter-bedded sandstones and mudstones deposited in meandering fluvial system (Rubiano, 1995 in Suarez, 1997). Mesa (1995) reveals a homogeneous composition of the Mugrosa and Colorado feldspathic sand-stones. Nevertheless, Gomez et al. (2005) described a distinctive facies association of
Figure 2.6: Composite stratigraphic column illustrating main lithologic characteristics of the MMVB Cenozoic deposits based on field measurements and descriptions by Gomez et al. (2005). Note the position of La Cira, Mugrosa and Los Corros fossil horizons, and the MEU. Highlighted in light yellow is the Mugrosa Formation, the main reservoir at Tenerife Field. Modified from Gomez et al. (2005).
inter-bedded fining-upward sandstones and mudstones, interpreted as point bar accre-tions along the inner bends of meandering rivers. Based on net-to-gross ratios reser-voir geologists informally have divided the Mugrosa Formation in two major sandstone packages: Zone B at the top and Zone C at the base. The reservoir description will be covered in more detail in section 2.4.2.
• Colorado Formation (Lower Miocene to lower middle Miocene): Unconformably over-lies the Mugrosa Formation. It is predominantly composed of massive, light-gray to purple-red shales, inter-bedded with fine to coarse-grained, well-sorted fluvial sand-stones (Morales, 1958 in Suarez, 1997), which are difficult to differentiate from the sandstones of the Mugrosa Formation (Mesa, 1995). An increase in sedimentary grains and decrease in feldspar percentage allows one to differentiate the Real Group from the Colorado sandstones (Gomez et al., 2005). Gomez et al. (2005) also pointed out that the facies association of fining-upward layers of cross-bedded sandstones and variegated mudstones indicates decreased channel sinuosity and increased flow energy compared with the Mugrosa Formation.
• Real Group (Upper Miocene): Unconformably overlies the Colorado Formation. The Real Group is a molase composed mainly of conglomerates, conglomeratic sandstones, and gray claystones, deposited in fluvial-alluvial environments (Rubiano, 1995 in Suarez, 1997). Beds of reworked volcanic deposits appear in the upper half of the Real Group. Gomez et al. (2005) interpreted these facies as a braided fluvial system with channel avulsions and annexation to the flood plain.
2.3 Structural styles
Given a complex tectonic history, it is not surprising that MMVB exhibits a complex structural style, characterized by super imposed deformation events. This complexity also makes this province particularly difficult to develop in terms of oil exploration and produc-tion. From the morphological point of view, the MMVB can be described as a monocline
dipping southeastward resembling a “half graben”. It is important to keep in mind that this wedge-like geometry is controlled by the protracted Cenozoic eastward tilting of the CC which, along with the tectonic load of the EC on the eastern side, favored accumulation of a thick Cenozoic sequence in a closed basin (Caballero et al., 2012 in press) (Figure 2.7).
Present-day structural configuration of the MMVB seems to be strongly controlled by the compression represented by the westward advance of the western thrust-and-fold belt of the EC, and somewhat the eastwards advance of the CC. Indeed, Figure 2.1 illustrates a regional NNE-SSW trend of reverse faults limiting both the CC and the EC.
Figure 2.7: Schematic geologic cross section of the MMVB illustrating the main structural elements and tectonosequences defined by Suarez(1997). See localization of this section in Figure 2.1. Under the MEU boundary, Jurassic and Cretaceous sequences are affected by structures different from those affecting the Tertiary rocks above it. The wedge shape of the Cenozoic basin is controlled by the actual eastward tilt of the MEU. Structurally-projected Tenerife Field is in red as reference. Modified from Suarez (1997).
Tenerife Field is located in the central part of the MMVB (Figure 1.1). Structural styles in the eastern margin, against the EC, are dominated by major faults, usually Mesozoic basement-involved inverted normal faults, e.g., La Salina Fault in Figure 2.8. They corre-spond to a compressive deformation developed over a previously rifted basin. The schematic cross-section of Figure 2.8 shows that the Eastern Cordillera foothills consist of
fold-and-thrust belts of Late Oligocene to early middle Miocene, e.g., Lisama Anticline, truncated by out-of-sequence faulting (e.g., La Salina Fault). Moreover, d´ecollement surfaces and foot-wall ramps of the EC thrust-and-fold belt occur at several stratigraphic levels along the MMVB (Gomez et al., 2005). Moreover, the western margin consists mainly of high-angle east verging reverse faults (Figure 2.7).
Since the MEU is an angular unconformity that separates pre-middle Eocene deforma-tion from post Middle Eocene -or Andean- deformadeforma-tion, the structural styles vary across this boundary. Underneath the MEU, the style is dominated by faulted blocks bounded by either normal, inverse, or inverted-normal faults Figure 2.7. These structures are then closely related to the Mesozoic heritage. Further, a significant component of compressional deformation must have occurred to generate important uplift and exposure of Cretaceous rocks leaving some positive local basement features -paleohighs- in the middle of the basin (e.g., Infantas Paleohigh in Figure 2.8), which were active until Early Oligocene (Moreno et al.,2011); see Figure 2.4. Above MEU the style is the result of the Andean deformation, primarily an east-west shortening, interacting with pre-existent structures.
Figure 2.8: Regional structural section across the MMVB parallel to section shown in Figure 2.7, see Figure 2.1. The EC foothills present a structural style dominated by a thrust-and-fold belt, while within the underlying Cretaceous sequence in the middle of the basin, the styles are characterized by faulted blocks and paleohighs. Structurally-projected Tenerife Field is in red as reference. Modified after Gomez et al.(2005).
Nonetheless, the MMVB does not exhibit the same structural style throughout the basin. Firstly, the CC is the site of the modern volcanic arc, but its structural configuration is not
well understood. It may involve a combination of strike-slip and reverse faulting such that the Central Cordillera may represent crustal-scale, thick-skinned, positive flower structures (Bar-rero, 1999 in Gomez et al., 2005). Secondly, in the northern part of the MMVB, especially in the San Lucas Range (Figure 2.1), right-lateral strike-slip faults have been interpreted, being the Palestina Fault the most relevant structure. Likewise, across the valley, a strong lateral displacement component on the structures is inferred from the enormous left-lateral strike-slip fault (Santa Marta - Bucaramanga Fault, SBF in Figure 2.1) that bounds the Santander massif -the northern extension of the EC- (Toro, 1990). Thereby wrenching structures, like positive flower structures, can be observed in the northern MMVB (Figure 2.9).
Furthermore, Gutierrez & Nur (2001) interpret La Cira-Infantas oil field as a group of anticlines arranged in a left-handed, en-´echelon pattern and highly broken by coexisting normal and reverse faults, being part of a small-displacement wrench example. Gutierrez & Nur (2001) also recognize the normal faults as the structural heterogeneities that most affect the external geometry of the reservoir compartments.
Figure 2.9: Structural cross section in the northern MMVB, see Figure 2.1. Here, unlike the sections of central MMVB (Figure 2.7 and Figure 2.8), the structure underneath the MEU is characterized by positive flower structures. The wrenching is the result of combined influence of the SBF and strike-slip faults in the San Lucas Range. Modified after Gomez et al.(2005).
2.3.1 Stress field
The stress field gives us the mechanical frame of modern deformation and rock mechanics conditions. This controls not only the formation of structures but also affects the rheology of the medium, the fracture gradients, and even the seismic anisotropy. Understanding the stress field is as important as the comprehension of the structure itself because the link between the geometrical description (structural styles) and the kinematics of the deformation is given by the applied forces driving the deformation process. In addition, it is important to restore the stress history because the physical characteristics of the rocks, e.g., the wave propagation velocities and geomechanical properties, strongly depend on the stress field.
As shown in previous section the inversion started in Late Cretaceous - Early Cenozoic (Dengo & Covey, 1993; Cooper et al., 1995; Gomez et al., 2005; Parra et al., 2009; Parra et al., 2012; among others). This regional contraction continued throughout the Cenozoic in different episodes creating compressional, transpressional and even locally extensional structures due to the oblique orientation of the regional stress field respect to the pre-existent structures (Kammer, 1999; Taboada et al., 2000; Cortes et al., 2005).
The slightly different stress regimes during the Cenozoic were recognized by Cortes et al. (2005) in the south of the MMVB (Figure 2.10). Inversion of the stress tensor based on fault slip data allowed them to conclude that stress regimes were characterized by a maximum horizontal stress (σ1) oriented E-W to WSW-ENE, that corresponds to an active contraction from Late Cretaceous to Late Paleocene. This direction subsequently changed to NW-SE and finally became WNW-ESE during the Andean deformation phase.
On the other hand, present-day stresses have been described previously (Suarez et al., 1983; Kellogg & Vega, 1995; Taboada et al., 2000; Corredor, 2003; among others). σ1 -direction results from interaction of NW South America with the Nazca and Caribbean plates, and the collision of the Panama block against central Colombia, with a component of northeast-directed motion due to the obliquity of the northern Andes with respect to the boundary of the Nazca and Caribbean plates (Corredor, 2003).
Figure 2.10: Paleostress analysis made by Cortes et al. (2005). Above: Pre-Eocene defor-mation phase is associated to E-W to WSW-ENE direction of maximum compression σ1; Below: Post-early Eocene phase is associated to NW-SE to WNW-ESE direction of σ1. The maximum horizontal stress rotates about 40oclockwise in between these episodes.
2.3.2 Cenozoic syn-tectonic deposits
Growth strata are formed when sedimentation occurs simultaneously with deformation. The differential lateral space of accommodation makes the strata to form a characteristic “fan pattern” divergent from the uplifted area to the zone with more relative subsidence. From Late Oligocene syn-tectonic sedimentation associated with an episode of continuous deformation due to the inversion of the EC has been documented (e.g.Gomez, 2001; Gomez et al., 2005;Parra et al., 2009; Mora et al., 2010).
Figure 2.11 shows two examples of growth strata interpreted on P-wave seismic data in the central part of the MMVB. The first one is the Lisama Anticline which can also be observed as a small structure in the footwall of the la Salina Fault in the regional cross-section shown in Figure 2.8. The thinning of the Oligocene-lower Miocene sequence (upper Mugrosa and Colorado formations), towards the crest of the anticline, indicates that the Lisama Anticline grew at the same time with the deposition of this interval. There is also a regional thinning westwards indicating the growing of a bigger structure in the middle of the basin, perhaps the Infantas paleohigh as pointed out in Figure 2.8. The second example shown in Figure 2.11 is called Provincia Anticline which is represented schematically in cross-section (Figure 2.7). At the eastern flank of the anticline, pre- and post-syn-tectonic sedimentation can be clearly differentiated from the growth strata package in between. There is no change in thickness between lower Mugrosa and MEU, thus the fold can be reconstructed with techniques of thickness preserved, whereas a typical “fan pattern” diverging from the crest of the anticline appears in the reflections corresponding to upper Mugrosa and Colorado formations.
2.4 Field characteristics: Geologic model of the Tenerife Field
Tenerife Field is located in the central part of the MMVB, it corresponds to a combined stratigraphic and structural trap. The reservoir is the Oligocene sandstones of the Mugrosa Formation which are deposited in a meandering fluvial system. Structurally, the trap has been described so far as an asymmetric east verging faulted anticline located in the
south-Figure 2.11: Seismic data illustrating syn-tectonic sedimentation in the Mugrosa Formation along two parallel seismic sections in the central MMVB. Growth strata can be observed particularly in the upper Mugrosa and Colorado formations. Left: Lisama Anticline. Right: Provincia Anticline. Modified after Gomez et al. (2005).
western flank of a major anticline structure (Ecopetrol, 2003). The anticline is apparently controlled and cut by reverse faults and back thrusts (Figure 2.14).
2.4.1 Correlation and well markers
As mentioned in the previous chapter three wells have been drilled in Tenerife Field. Two of them, Tenerife-1 and Tenerife-2, drilled The Cenozoic sequence from the Real Group at the surface to the MEU at bottom hole. Using a reference log taken from the closest oil field, called La Cira - Infantas, the electrofacies and lithologic records were correlated (Figure 2.12).
Correlating based on electro-facies only from SP and induction logs is not very accurate. For instance, the SP log does not capture the bell-shaped electrofacies that represent finning up-ward sequences (Figure 2.13) as GR logs do. However, the formation evaluation logs (FEL) take into account the fossil markers discussed previously (Figure 2.6) for precisely identifying the tops of Colorado, Mugrosa and Esmeraldas formations. Below the MEU the Cretaceous reported is the Umir Formation.
In addition, in Tenerife-3 the thickness of Mugrosa C looks much greater than expected from correlating the other two wells; although the net thickness of the Mugrosa Formation remains more or less the same, which discards a strong effect of the structural dip on the apparent thickness. This reflects an increasing in net sand content toward the north. In fact, according to Rojas (2011) Mugrosa B seems to contain more sandstone in La Cira - Infantas Field (Figure 2.13). We can conclude from the stratigraphic point of view that Mugrosa C in Tenerife-3 is more related to the one found in La Cira - Infantas than to those drilled in Tenerife-1 and Tenerife-2 wells.
2.4.2 Reservoir characteristics
The siliciclastic reservoirs in La Cira-Infantas and Tenerife oil fields have been divided in operational units called: Zone A, within Colorado Formation, Zones B and C that belong to the Mugrosa Formation, and Zone D as the basal deposits of the Esmeraldas Formation. The
Figure 2.12: Correlation among the Tenerife wells and a reference well log taken from La Cira - Infantas field, located at 20 Km NE-direction from Tenerife. Correlation among the Tenerife wells and a reference well log taken from La Cira - Infantas field, located at 20 Km NW-direction from Tenerife. Despite the units were identified virtually with the same thickness in all three wells, Mugrosa C looks more than double the thickness in Tenerife-3.
