See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/204777151

A simple method of determining sand/shale ratios from seismic analysis of

growth faults: An example from upper Oligocene to lower Miocene Niger
Delta deposits

Article  in  AAPG Bulletin · October 2004

DOI: 10.1306/04290403117

CITATIONS READS

33 766

5 authors, including:

Stéphane Pochat Sébastien Castelltort

University of Nantes University of Geneva
60 PUBLICATIONS   604 CITATIONS    139 PUBLICATIONS   2,470 CITATIONS   

SEE PROFILE SEE PROFILE

K. Besnard Charles Gumiaux

. CNRS Orleans Campus
18 PUBLICATIONS   207 CITATIONS    96 PUBLICATIONS   1,690 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Marie Skłodowska-Curie Innovative Training Network on Signal Propagation in Source to Sink for the Future of Earth Resources and Energy (S2S-Future) View
project

Antimony mineralisation: the Variscan case View project

All content following this page was uploaded by Sébastien Castelltort on 12 January 2016.

The user has requested enhancement of the downloaded file.

A simple method of determining AUTHORS
S. Pochat
Géosciences Rennes, UMR
sand/shale ratios from seismic 6118, Université de Rennes 1, Av. du Général
Leclerc, Campus de Beaulieu, 35042 Rennes
analysis of growth faults: An Cedex, France
Stéphane Pochat completed his Ph.D. in sedi-
example from upper Oligocene mentology and structural geology in June 2003
at Rennes University. He addressed the kine-
to lower Miocene Niger matics of growth faults with examples in deep-
water turbiditic environments and their influ-
Delta deposits ences on sedimentary processes and made
paleoclimatic reconstructions based on lacus-
trine sedimentology of Permian lakes. He is
S. Pochat, S. Castelltort, J. Van Den Driessche, currently doing postdoctoral studies in the
K. Besnard, and C. Gumiaux Laboratoire Régional des Ponts et Chaussées
Institute (Lyon, France) on risks assessment.
S. Castelltort
Géosciences Rennes, UMR
ABSTRACT 6118, Université de Rennes 1, Av. du Général
Leclerc, Campus de Beaulieu, 35042 Rennes
A plot of fault throw vs. depth is a simple geometric tool that graph- Cedex, France; Department of Earth Sciences,
ically represents strata thickness variations in growth-fault and growth- Eingeidnössische Technische Hochschule-
fold settings and has previously been used to infer fault kinematics. Zentrum, Sonneggstrasse 5, CH-8092
In this paper, we use it as a prediction tool for lithological change Zürich, Switzerland
employing only seismic data. If growth faulting is a continuous pro- Sébastien Castelltort completed his Ph.D. in
cess, intervals of shale deposition are recorded by unthickened units, June 2003 at Rennes University. He worked
while intervals of sand fill in topographic lows. The throw vs. depth on the origin of high-frequency terrigenous
plot easily allows depiction of unthickened and thickened sedimen- stratigraphic cycles, their relationships with
tary intervals from even rough seismic records and can therefore be growth folds and faults, and the numerical
used to predict sand/shale ratios. The method is here applied to a modeling of fluvial systems. He is now doing
growth fault in the Niger Delta that affects Oligocene to lower postdoctoral studies in the Earth Surface Pro-
Miocene deltaic deposits. Most shale intervals are identified, and the cesses Group at Eingeidnössische Technische
sand/shale ratios are predicted. We suggest that the method can be a Hochschule Zurich (Switzerland), working on
the organization of drainage networks in
valuable tool in oil exploration.
mountain ranges.

J. Van Den Driessche
Géosciences

INTRODUCTION Rennes, UMR 6118, Université de Rennes 1,
Av. du Général Leclerc, Campus de Beaulieu,
The analysis of strata thickness variations is widely employed in 35042 Rennes Cedex, France;
reconstructing the kinematics of growth structures at various de- Jean.Van-Den-Driessche@univ-rennes1.fr
grees of resolution. A commonly used graphical method is called the Jean Van Den Driessche has a Ph.D. from Mont-
‘‘T-Z plot,’’ which consists of plotting for each horizon the vertical pellier University and completed a State thesis
throw T of a stratigraphic marker vs. its depth Z (Tearpock and at Paris 7 University in 1994. He then worked
Bischke, 1991; Bischke, 1994). The obtained curves defined by the on fault-sealing mechanisms in Elf Aquitaine,
data arrays display variations of slope that reflect variations in the and since 1996, he has been a professor at
the University of Rennes. His teaching and
degree of thickening of the strata toward the hanging wall.
research work concerns tectonics, structural
geology, and geomorphology. He has been
chief editor of Geodinamica Acta since 1999.

Copyright #2004. The American Association of Petroleum Geologists. All rights reserved.
Manuscript received October 21, 2003; provisional acceptance January 28, 2004; revised manuscript
received March 25, 2004; final acceptance April 29, 2004.

AAPG Bulletin, v. 88, no. 10 (October 2004), pp. 1357 –1367 1357
 K. Besnard
Géosciences Rennes, UMR If sedimentation can be assumed to always fill fault-induced
6118, Université de Rennes 1, Av. du Général topography to the top (fill-to-the-top model), the T-Z plot can be
Leclerc, Campus de Beaulieu, 35042 Rennes used to constrain the displacement history of growth faults (e.g.,
Cedex, France Mansfield and Cartwright, 1996; Cartwright et al., 1998). Thick-
Katia Besnard works on the numerical mod- ening of strata toward the hanging wall indicates a period of fault
eling of water flows and reactive transport in growth, whereas unthickened intervals are symptomatic of periods
heterogeneous porous media to understand of tectonic quiescence (Figure 1A).
the factors and processes influencing contam- Fault-induced topographies have been widely documented on
inants transport in groundwater flows. She
the present-day sea floor and in ancient deposits through their in-
also worked on the physical and numerical
fluence on sedimentary processes in a range of depositional settings
modeling of tectonic and sedimentation rela-
(e.g., Bornhauser, 1959; Piper and Normark, 1983; Petit and Beau-
tionships. She completed her Ph.D. thesis in
December 2003 at Rennes University, where champ, 1986; Thornburg et al., 1990; Leeder and Jackson, 1993;
she now does postdoctoral studies. Edwards, 1995; Mitchell, 1996; Morris et al., 1998; Kneller and
McCaffrey, 1999; Nelson et al., 1999; Shaw et al., 2004; Soreghan
C. Gumiaux
Géosciences Rennes, UMR et al., 1999; Armentrout et al., 2000; Burgess et al., 2000; Newell,
6118, Université de Rennes 1, Av. du Général 2000; Bouroullec, 2002). Therefore, the assumption of fill-to-the-
Leclerc, Campus de Beaulieu, 35042 Rennes
top sedimentation, which is required to interpret growth strata pat-
Cedex, France
terns in terms of fault kinematics, is not always justified and pre-
Charles Gumiaux is a structural geologist with cludes directly determining fault kinematics from T-Z plots (Childs
particular interests in the application of geo- et al., 1993; Castelltort et al., in press).
statistics. He worked on three-dimensional
In deltaic settings and in most depositional settings, sedimen-
lithospheric deformation modeling of the
tation consists of stratigraphic cycles characterized primarily by al-
Hercynian domain in Brittany (France) based
ternation between more energetically and less energetically sup-
on tomographic, magnetic, and seismic data
analysis. He completed his Ph.D. in April plied materials, such as sand-shale alternations (e.g., Damuth, 1994;
2003 at Rennes University and is currently Edwards, 1995; Cross and Lessenger, 1998; Soreghan et al., 1999).
doing postdoctoral studies at the Vrije Univer- The various processes can react in different ways to differential sub-
siteit, Department of Tectonics (Amsterdam, sidence because of growth faulting: Nondynamic settling of shale
Netherlands). particles (decantation) makes sedimentation homogeneously dis-
tributed across faults (Lowrie, 1986; Cartwright et al., 1998) with-
out being diffused to topographic lows on timescales of tens to
ACKNOWLEDGEMENTS hundreds of thousands of years (Mitchell, 1996; Webb and Jordan,
This study was supported by Total, who funded 2001), whereas dynamic deposition (sands) fills the lows before
Stephane Pochat’s Ph.D. thesis, and by the deposition takes place on the highs.
French Ministry of Research and Education In this way, Castelltort et al. (in press) have proposed that if
for the other authors. We thank in particular fault displacement can be treated as a continuous process, then the
J. L. Montenat and S. Raillard, who gave us slope variations on a T-Z plot may be interpreted as fault-induced
access to the data on the Niger Delta, and variations in topography (Figure 1B). Furthermore, these fault-
S. Raillard for convincing us to publish this induced topographic variations can then be related to lithological
work. Many thanks to K. C. Burke and S. P.
changes as in the case of alternating sand and shale deposition. The
Cumella for their very helpful and enthusiastic
reviews of the manuscript. We also thank T-Z plot method can therefore be applied as a tool to predict litho-
Helen Williams for her corrections of the logical variations on seismic profiles and sand/shale ratios, which
English manuscript. Finally, we are indebted are of fundamental importance in oil exploration and production.
to the Sedimentology Team in Rennes Univer- In their work, Castelltort et al. (in press) presented the theo-
sity for their insightful and mature encour- retical treatment of the interpretation of T-Z plots for the two end
agement that pushed us to further carry our members of fill-to-the-top and variable displacement and topog-
reasoning as independent researchers. raphy models. Although they illustrated only briefly the potential
of the method with an example of growth normal fault in the Niger
Delta, in the present paper, our aim is to develop how we applied
the T-Z plot method to this example to predict lithological changes
and sand/shale ratios from seismic records. This study therefore pres-
ents how to analyze a natural example with the method developed

1358 A Method of Determining Sand/Shale Ratios from Seismic Analysis of Growth Faults
 Figure 1. Significance of T-Z
plot slope variation according
to two end-member models
(Castelltort et al., in press).
(A) Fill-to-the-top model: Each
slope variation characterizes
the ratio between fault displace-
ment rate and sedimentation
rate. A zero-slope interval is
symptomatic of fault inactivity.
A positive-slope interval is
symptomatic of fault activity.
(B) Variable displacement and
topography model: Alternation
of zero-slope segments and
positive-slope segments corre-
sponds to alternation between
topographic creation and topo-
graphic filling (or eroding) in-
tervals. Each pair of topographic
creation and filling-eroding in-
tervals defined on the T-Z plot
makes it possible to trace a lower
curve envelope (dashed line)
that corresponds to a fill-to-the-
top model in this interval. On
each portion of this curve, the
ratio between fault displacement
and the sedimentation rate is
assumed to be constant. How-
ever, every slope variation be-
tween each portion of the curve
is considered to be representa-
tive of the variation in this ratio.

in Castelltort et al. (in press). The case studied is a existing continental slope into the deep sea during the
growth normal fault in the Niger Delta. We recognize late Eocene (Burke, 1972) and is still active today. The
that faults always move in discrete episodes at the stratigraphy of the Niger Delta (Figure 2) can be di-
timescale of seismic events, but as we show in the vided into three major units (Short and Stauble, 1967),
present work, it is justified to treat growth faulting here ranging in age from Eocene to Holocene: (1) the Akata
as a continuous process as long as (1) we work on much Formation (Figure 2), which includes at least 6500 m
longer timescales of deposition, (2) deposition rates (21,400 ft) of marine clays with silty and sandy in-
in the Niger Delta are very rapid, and (3) shale inter- terbeds (Whiteman, 1982), (2) the Agbada Formation
vals in the Niger Delta do not thicken into hanging- (Figure 2), which is characterized by paralic to marine-
wall topographic depressions. coastal and fluvial-marine deposits mainly composed of
sandstones and shale organized into coarsening-upward
off-lap cycles, and (3) the Benin Formation, which con-
GEOLOGICAL SETTING sists of continental and fluvial sands, gravel, and back-
swamp deposits (2500 m [8250 ft] thick) (Figure 2).
Within the Gulf of Guinea, the Niger Delta covers an The continental margin off the Niger Delta is under-
area of about 140,000 km2 (54,000 mi2) and has an going deformation by gravity tectonism (Figures 2, 3)
average sediment thickness of about 12 km (7 mi). This caused by a rapid stepwise seaward sediment progra-
siliciclastic system began to prograde across the pre- dation over a strong thickness of shale deposits (Akata

Pochat et al. 1359

Figure 2. Schematic north-northeast –south-southwest cross section (line AA0 in Figure 3) of the upper extensional part of the
region of the Niger Delta (modified after Cohen and McClay, 1996). The formation names are those formally introduced by Short and
Stauble (1967).

Formation) (Evamy et al., 1978). This formation is through stepwise alluvial progradation facilitated by
overpressured (Merki, 1972) and acts as a mobile sub- large-scale withdrawal and forward movement of the
strate (Figure 2), deforming in response to deltaic pro- underlying shale (Doust, 1990; Doust and Omatsola,
gradation and sediment loading (Doust, 1990; Doust 1990). From the outer shelf (shallow water) to the
and Omatsola, 1990). slope (deep water) of the Niger Delta, the deformation
In the delta top, sedimentation concentrated in is characterized successively by (1) a zone undergoing
numerous arcuate belts (Figure 3) bounded by large- extension (Figure 2), (2) a zone undergoing only trans-
scale regional and counterregional growth faults (Doust, lation featured with shale ridges and diapirs, and
1990; Doust and Omatsola, 1990; Cohen and McClay, (3) a zone of compression, with imbricate toe thrusts
1996) (Figure 2). The activity in each belt has pro- beneath the lower slope and rise (Damuth, 1994; Cohen
gressed in time and space toward the south-southwest and McClay, 1996).

Figure 3. Regional structural

map of the continental part
of the Niger Delta. The studied
area (gray rectangle) is located
on the Greater Ughelli depobelt
region (modified after Bouroul-
lec, 2002). Line AA0 corresponds
to a schematic cross section de-
tailed in Figure 2.

1360 A Method of Determining Sand/Shale Ratios from Seismic Analysis of Growth Faults
 the studied series were deposited between 29.8 and
18 Ma, and the cycles range in duration between 0.3
and 3.5 m.y. (Table 1).

MEASUREMENT METHOD

The first step of the T-Z plot method consists of pick-

ing several markers on a seismic line that are corre-
latable between the two sides of a fault (Figure 5). In
the present work, 78 markers have been correlated
across the studied fault, between the footwall and the
hanging wall (Figure 6) on the seismic profile.
In a second step, we measured the thickness of all
the intervals between successive markers in the hang-
ing wall and the associated vertical throw across the
Figure 4. Map of studied fault (black fill) on horizon 30 (see fault (Figure 7) for each horizon. Following White
Figure 5); depth is in milliseconds two-way traveltime (modified
et al. (1986) and Bischke (1994), the effects of near-
from Bouroullec, 2002). The studied normal fault is oriented
fault deformations and artifacts can be minimized by
northwest-southeast and shows a relatively arched profile. The
location of the cross section is shown by the dashed line BB0 making the measurement at a constant distance from
in Figure 5. The location of wells, Total well 1 and Total well 2, the fault (about 300 m [990 ft] in the present case).
in the footwall and the hanging wall of the fault, respectively, The damage zone (zone of maximum deformation) is
is indicated by the black circles. avoided in this way. This prevents the misinterpreta-
tion of local near-fault folds and artifacts that could
simulate true fault-induced sediment thickness changes.
LOCAL SETTING However, following this procedure may induce an
underestimation of the true amplitude of throw and
Our measurements were carried out on a northwest- thickness variations.
southeast normal fault with a relatively arched profile In a third step, starting from the shallowest un-
(Figure 4), located in the inner shelf of the southeast- faulted interval, we plotted the throw of each horizon
ern part of the Greater Ughelli depobelt (Figure 3). (Y-axis) against their depth (X-axis) down the fault.
This fault affected the deltaic deposits of the Agbada Because of the vertical resolution of our seismic data,
Formation. Four major depositional environments have
been recognized in the Agbada Formation (Vannier Table 1. Approximate Ages (J.-L. Montenat, S. Raillard, 2003,
and Durand, 1994; Schulbaum et al., 1996): (1) the personal communication) of Each of the Maximum Flooding
upper delta plain, (2) the tidal zone, (3) the delta front, Surfaces Deduced from the Mesozoic – Cenozoic Sequence-
and (4) the prodelta. In the study area, the delta pro- Chronostratigraphic Chart of Hardenbol et al. (1998)*
graded in a southward direction from the early Oligo-
Flooding Surface** Age (Ma)
cene (ca. 30 Ma) to late Miocene (ca. 6 Ma).
Our work focuses on a stratigraphic interval com- mfs3d 18
posed of eight regressive-transgressive (R-T) cycles mfs3c 21
(J.-L. Montenat, S. Raillard, 2003, personal communi- mfs4e 22
cation). Each R-T cycle is defined by two maximum mfs4d 23
flooding surfaces that represent the deepest water mfs4b 25
deposits and contain the most argillaceous material. mfs4a 26
According to biostratigraphic data (J.-L. Montenat, mfs5c 27
S. Raillard, 2003, personal communication), and in the mfs5b 28
absence of further constraints, we calibrated the nine mfs5a 29
maximum flooding surfaces of the studied interval using
*These data are presented to give a rough idea of the duration of the R-T cycles
the Mesozoic–Cenozoic sequence-chronostratigraphic considered in this study.
chart of Hardenbol et al. (1998). With this correlation, **mfs = maximum flooding surface.

Pochat et al. 1361

Figure 5. (A) Southwest-northeast seismic line showing northwest-southeast normal fault. (B) Each yellow line corresponds to the
tens number markers used in correlation (see Figure 6). Location of the wells is indicated by the black vertical lines.

Figure 6. Complete line draw-

ing with 78 correlatable markers
on both sides of the normal
fault. The dotted line represents
the fault-constant space axis
chosen for our measurement.
The location of the wells is
indicated by the black vertical
lines. TWT = two-way traveltime.

1362 A Method of Determining Sand/Shale Ratios from Seismic Analysis of Growth Faults
 tical throw compared to the total sediment thickness,
the differential compaction can be considerable (Heg-
arty et al., 1988; Forbes et al., 1991). For instance, in
the case of a fault with 1000 m (3300 ft) of throw,
footwall strata at 1000 m (3300 ft) of depth may be
less compacted than their hanging-wall counterparts
by about 20% (Forbes et al., 1991). In the present case,
the available data (J.-L. Montenat, S. Raillard, 2003,
Figure 7. Method of thickness and vertical throw measure- personal communication) indicate that differential com-
ment for T-Z plot construction. The throw T is defined for each paction between the two fault compartments only
horizon as the vertical distance separating one horizon situated reaches a maximum of about 5%. The impact of such a
on the footwall from the same horizon (of the same age) variation on the sediment thickness and vertical throw
situated in the hanging wall. The depth Z of a given horizon can thus be neglected because it is below our seismic
corresponds to the whole thickness of sediment between the resolution threshold.
shallowest first nonfaulted layer and the considered horizon in
the hanging wall. Horizons were picked at the dashed lines to
avoid near-fault deformation. T-Z PLOT ANALYSIS

Construction of a Synthetic Lithological Column from

vertical throw variations of less than 10 ms two-way T-Z Plot Interpretation
traveltime (TWT) cannot be determined (about 20 m
[70 ft]). Thus, any variation of vertical throw equal to The first observed feature of the obtained T-Z plot
or lower than this value is regarded as negligible. (Figure 8) is the variation of the vertical throw with
In such a deltaic setting, the alternation between depth. The vertical throw encompasses a long-term dim-
sands and shale may induce differential compaction. inution with time (with decreasing depth, i.e., from right
This produces thickness variations between the two to left on the T-Z plot) from 350 to 0 ms TWT over a
fault compartments that do not reflect the initial dep- sediment thickness of 2000 ms TWT (Figure 8). Such a
ositional thicknesses. If the fault has a significant ver- long-term throw-vs.-depth relationship is characteristic

Figure 8. Results of T-Z plot

measurements along the 78
layers. Throw T is reported
on the Y-axis in milliseconds,
depth Z is reported on the X-axis
in milliseconds. The dashed rect-
angle represents the depth cov-
ered by borehole data for both
the hanging wall and the foot-
wall. Note that the major max-
imum flooding surfaces are all
located on the zero-slope por-
tions of the T-Z plot (gray circle).
TWT = two-way traveltime; mfs =
maximum flooding surface.

Pochat et al. 1363

Figure 9. Enlargement of T-Z
plot diagram (line with black
circles) corresponding to the
dashed rectangle in Figure 8.
The slope values for each por-
tion of the T-Z plot are repre-
sented by the black vertical
columns (Y-axis on the right).
We interpret all zero-slope
values as being representative
of topographic creation inter-
vals during shale deposition
(gray rectangle) and all positive-
slope values as being represent-
ative of topographic infilling
intervals during sand deposi-
tion (yellow rectangle). TWT =
two-way traveltime.

of syndepositional faults (Ocamb, 1961). In addition, The idea of this paper is to test the second inter-
given that the youngest marker used horizontally sealed pretation with a simple hypothesis. Indeed, in the case of
the fault, this strongly suggests a long-term damping of a continuous displacement model, one reason for the
the fault displacement rate with time (long-term dim- alternation of topography creation and filling periods is
inution of displacement rate to final fault sealing). the presence of alternately draping and filling sedimentary
The second major feature of this T-Z plot (Figure 8) processes. These may be basically expressed by an alter-
is the presence of alternating positive-slope sectors nation between shale deposited under low-energy con-
and zero-slope sectors (linked to intervals of thickness ditions and higher energy sand depositional conditions.
variations below resolution) all along the curve. This Our available borehole lithological data cover the
feature seems to be another feature of growth faults as whole series but are only well correlated with strati-
it has been described in other works (Bischke, 1994; graphic surfaces on the seismic section in the interval
Mansfield and Cartwright, 1996; Cartwright et al., 1998; of maximum flooding surface 5a to maximum flood-
Bouroullec 2002). ing surface 3d. Therefore, the following discussion will
The existence of these high-frequency T-Z plot slope focus on this approximate 11-m.y.-long interval (dashed
variations is amenable to different interpretations. The box in Figure 8).
two end-member interpretations are as follows: For the interval maximum flooding surface 5a to
maximum flooding surface 3d, the T-Z plot shows 15
1. If sedimentation is assumed as always filling the fault- zero-slope and 14 positive-slope segments (Figure 9).
induced topography, the alternations reflect varia- The positive-slope values are strongly variable with
tions of fault activity with time. Zero-slope segments amplitudes ranging from 0.2 to 0.7. Some positive-
represent times of fault inactivity, and positive-slope slope intervals are affected by slope variations without
segments represent times of renewed displacement. an intervening zero-slope segment (Figure 9).
2. If fault displacement is assumed to be a more or less From Figure 9, a synthetic lithological column
continuous process, without episodic activity, the al- (Figure 10A; with Z-axis in milliseconds TWT) can be
ternations are better explained as alternating times of extracted by assuming that zero-slope intervals repre-
creation (zero-slope intervals) and filling (positive- sent periods of topography creation during shale depo-
slope intervals) of fault-induced topography (Castell- sition. Positive-slope segments reflect filling of topo-
tort et al., in press). graphic depressions by sands.

1364 A Method of Determining Sand/Shale Ratios from Seismic Analysis of Growth Faults
 ing surface 4a and maximum flooding surface 4b, which
are slightly deeper on the synthetic log, probably as a
result of near-fault folding (Figure 6). On the whole,
this suggests a good homogeneity of seismic velocities
with depth.
Our first observation is that all the argillaceous in-
tervals associated with maximum flooding surfaces of
the sequence-stratigraphic interpretation of the borehole
data are correctly predicted by the synthetic column.
Between those maximum flooding surfaces, additional
argillaceous intervals have been found with the T-Z plot
method. Some may be reasonably matched with their
counterparts on the borehole column (such as 2, 3, and 4),
others (such as 1, 5, and 6) are more difficult to unam-
biguously correlate with the borehole shales, and sev-
eral small shales in the borehole have no match in the
synthetic column. This may be attributed to the lower
resolution of the T-Z plot compared to the well-log res-
olution (compare the number of horizons plotted on the
synthetic column Figure 10A to the number of existing
horizons as seen on the borehole data Figure 10B).
Second, we have compared the sand/shale ratios
of the synthetic and real data.
At the scale of the whole, the sand/shale ratios are
very close, with 60:40 on the synthetic log and 55:45
in the borehole.
At the scale of the R-T cycles, the T-Z plot also
Figure 10. Comparison between the synthetic lithological predicts the sand/shale ratios very well for most cycles,
column obtained with the T-Z plot method (column A) and the with the greatest error being 15% for R-T cycle C8. Only
borehole lithological data (column B). The ratios reported on the sand/shale ratios for the two cycles C5 and C6 are
each column represent sand/shale ratios. They are calculated not well predicted by the T-Z plot.
(1) for each major regression-transgression cycle (C1 – C8) and
(2) for the totality of each column. All circles with numbers
inside (1 –6) are shale bodies indicated by the T-Z method, and DISCUSSION
do not correspond to major maximum flooding surfaces in the
borehole sequence stratigraphy. The S body is a thick shale On the whole, the T-Z plot method allows the satis-
body present in borehole data but not well predicted by the T-Z factory prediction of the position of argillaceous in-
plot method. The question marks represent every shale or sand
tervals and sand/shale ratios. Some deviations from
body not predicted by the T-Z plot method. TWT = two-way
borehole data occur, as in R-T cycles C5 and C6.
traveltime; mfs = maximum flooding surface. Borehole litho-
logical data from Total well 2. There, an interval of shale more than 200 m (660 ft)
thick (‘‘S’’ body) has been underestimated (shale
body 5, Figure 10A) and is responsible for the poor
Comparison with Borehole Data estimation of the sand/shale ratios of both cycles.
Three explanations are proposed:
The synthetic lithological column (Figure 10A) has been
rescaled so that its base and top correspond with those 1. The S body was deposited on the footwall but sub-
of the hanging-wall borehole lithological column. Al- sequently eroded (for example, during the succeed-
though the vertical axis units are different, with milli- ing sand deposition), in this way producing an
seconds for the synthetic log and meters for the bore- abnormal hanging-wall thickening of those shales
hole data, the relative positions of stratigraphic surfaces compared to their thickness on the footwall (and
in both columns are similar, except for maximum flood- thus, a positive slope on the T-Z plot).

Pochat et al. 1365

2. If this body corresponds to a time of shale depo- REFERENCES CITED
sition of more than hundreds of thousands of years
(cf. Table 1), the shale could have diffused by creep- Armentrout, J. M., K. A. Kanschat, K. E. Meisling, J. J. Tskama,
ing down from the footwall to the hanging wall L. Antrim, and D. R. McConnell, 2000, Neogene turbidite
(Mitchell 1996) or by redistribution of fine particles systems of the Gulf of Guinea continental margin slope, off-
shore Nigeria, in A. H. Bouma and C. G. Stone, eds., Fine-
because of bottom currents, waves, or tides action,
grained turbidite systems: AAPG Memoir 72, p. 93 – 108.
thus progressively erasing the fault topography in Bischke, R. E., 1994, Interpreting sedimentary growth structures
these shales. from well log and seismic data (with examples): AAPG Bulletin,
v. 78, no. 6, p. 873 – 892.
3. The lithology has been erroneously described on the
Bornhauser, M., 1959, Depositional and structural history of north-
borehole; in fact, the S body corresponds to very west Hartburg field, Newton County, Texas: AAPG Bulletin,
high-frequency sand-shale alternations, producing v. 44, no. 4, p. 458 – 470.
an average argillaceous interval. Bouroullec, R., 2002, Growth faults kinematics: Ph.D. thesis, Impe-
rial College London, England, 396 p.
Burgess, P. M., S. Flint, and S. Johnson, 2000, Sequence strat-
Misfits between the relative positions of some igraphic interpretation of turbiditic strata; an example from
lithological boundaries between the synthetic and the Jurassic strata of the Neuquén basin, Argentina: Geological
Society of America Bulletin, v. 112, no. 11, p. 1650 – 1666.
real data can arise as a result of near-fault deformation
Burke, K. C., 1972, Longshore drift, submarine canyons and sub-
(although we tried to correct this effect by making marine fans in development of the Niger Delta: AAPG Bul-
measurements outside the damage zone) or seismic letin, v. 56, p. 1975 – 1983.
velocity variations (our seismic profile has not been Cartwright, J. A., R. Bouroullec, D. James, and C. S. Johnson, 1998,
Polycyclic motion history of some Gulf Coast growth faults
migrated in depth). Eventually, the absence on the from high resolution displacement analysis: Geology, v. 26,
synthetic results of some small argillaceous intervals p. 819 – 822.
that are present in the borehole data underscores the Castelltort, S., S. Pochat, and J. Van Den Driessche, in press, Using
T-Z plots as a graphical method to infer lithological varia-
fact that the choice of horizons on the seismic profile tions from growth strata: Journal of Structural Geology, v. 26,
(in turn to some extent dependent on the seismic p. 1425 – 1432.
resolution) controls the degree of accuracy of the T-Z Childs, C., S. J. Easton, B. Vendeville, M. P. A. Jackson, S. T. Lin,
J. J. Walsh, and J. Watterson, 1993, Kinematic analysis of
plot method.
faults in a physical model of growth faulting above a viscous
Bischke (1994) already noticed the possibility of salt analogue: Tectonophysics, v. 228, p. 313 – 329.
linking slope variations on T-Z plots to variations of Cohen, H. A., and K. R. McClay, 1996, Sedimentation and shale
tectonics of the northwestern Niger Delta front: Marine and
lithology. In addition, Cartwright et al. (1998) docu-
Petroleum Geology, v. 13, p. 313 – 328.
mented evidence of slope variations on the T-Z plots of Cross, T. A., and M. A. Lessenger, 1998, Sediment volume
several normal growth faults in the Gulf of Mexico. partitioning: rationale for stratigraphic model evaluation and
They were able to interpret those slope variations as high-resolution stratigraphic correlation, in F. M. Gradstein,
K. O. Sandvik, and N. J. Milton, eds., Sequence stratigraphy —
being linked to variations of the fault activity only Concepts and applications: Norwegian Petroleum Society Spe-
because it was possible to rule out significant litholog- cial Publication 8, p. 171 – 195.
ical variations such as sand-shale alternations in their Damuth, J. E., 1994, Neogene gravity tectonics and depositional
processes on the deep Niger Delta continental margin: Marine
study case.
and Petroleum Geology, v. 11, p. 320 – 346.
In conclusion, despite a few anomalies, the charac- Doust, H., 1990, Petroleum geology of the Niger Delta, in J.
terization of alternations between zero-slope segments Brooks, ed., Classic petroleum provinces: Geological Society
and positive-slope segments on the T-Z plot enabled (London) Special Publication 50, p. 365 – 380.
Doust, H., and E. Omatsola, 1990, Niger Delta, in J. D. Edwards,
us to define the alternations between fault-induced and P. A. Santogrossi, eds., Divergent/passive margins basins:
topographic creation periods and topographic filling AAPG Memoir 48, p. 201 – 238.
periods. These were satisfactorily associated with a Edwards, M. B., 1995, Differential subsidence and preservation
potential of a shallow water Tertiary sequences, northern Gulf
simple model of sand-shale alternations that predicts Coast basin, U.S.A.: Special Publication of International
the observed stratigraphy in the borehole data well. Association of Sedimentology, v. 22, p. 265 – 281.
The T-Z plot method needs further testing but can be Evamy, B. D., J. Haremboure, P. Kamerling, W. Knapp, F. A.
Molloy, and P. H. Rowlands, 1978, Hydrocarbon habitat of
a valuable tool to easily detect and quantify sand-shale
Tertiary Niger Delta: AAPG Bulletin, v. 62, p. 1 – 39.
alternations from two-dimensional seismic profiles. Forbes, P. L., P. M. Ungerer, A. B. Kuhfuss, F. Riis, and S. Eggens,
The level of accuracy of the method could be in- 1991, Compositional modeling of petroleum generation and
expulsion; trial application to a local mass balance in the Smor-
creased by comparing different T-Z plots from a greater
bukk Sor field, Haltenbanken area, Norway: AAPG Bulletin,
number of seismic lines on the same growth fault, for v. 75, no. 5, p. 873 – 893.
example. Hardenbol, J., J. Thierry, M. B. Farley, T. Jacquin, P. C.

1366 A Method of Determining Sand/Shale Ratios from Seismic Analysis of Growth Faults
 DeGraciansky, and P. R. Vail, 1998, Mesozoic and Cenozoic of the Gulf Coast Association of Geological Societies, v. 11,
sequence chronostratigraphic framework of European basins, p. 139 – 175.
chart 1, in P. C. DeGraciansky, J. Hardenbol, T. Jacquin, and Petit, J. P., and J. Beauchamp, 1986, Synsedimentary faulting and
P. R. Vail, eds., Mesozoic and Cenozoic sequence stratigraphy palaeocurrent patterns in the Triassic sandstones of the High
of European basins: SEPM Special Publication 60, p. 3 – 13. Atlas (Morocco): Sedimentology, v. 33, p. 817 – 829.
Hegarty, K. A., J. K. Weissel, and J. C. Mutter, 1988, Subsidence Piper, D. J. W., and W. R. Normark, 1983, Turbidite depositional
history of Australia’s southern margin: Constraints on basin patterns and flow characteristics, Navy Submarine Fan,
models: AAPG Bulletin, v. 72, no. 5, p. 615 – 633. California borderland: Sedimentology, v. 30, p. 681 – 694.
Kneller, B. C., and W. D. McCaffrey, 1999, Depositional effects of Schulbaum, L., J. Durand, and D. Rousset, 1996, Traduction des
flow non-uniformity and stratification within turbidity currents surfaces stratigraphiques et des géométries deltaı̈ques lors du
approaching a bounding slope: Deflection, reflection and facies passage de l’échelle puit à l’échelle sismique: Bulletin Société
variations: Journal of Sedimentary Research, v. 69, p. 980 – 991. Géologique de France, v. 167, no. 6, p. 667 – 683.
Leeder, M. R., and J. Jackson, 1993, The interaction between Shaw, J. H., E. Novoa, and C. D. Connors, 2004, Structural controls
normal faulting and drainage in active extensional basins, with on growth stratigraphy in contractional fault-related folds, in K.
examples from the western United States and central Greece: McClay, ed., Thrust tectonics and hydrocarbon systems: AAPG
Basin Research, v. 5, p. 79 – 102. Memoir 82, p. 400 – 412.
Lowrie, A., 1986, Model for fine-scaled movements associated Short, K. C., and A. J. Stauble, 1967, Outline of geology of Niger
with climate and sea-level changes along Louisiana shelf-break Delta: AAPG Bulletin, v. 51, p. 61 – 770.
growth faults: Transactions of the Gulf Coast Association of Soreghan, M. J., C. A. Scholz, and J. T. Wells, 1999, Coarse-grained,
Geological Societies, v. 36, p. 497 – 509. deep-water sedimentation along a border fault margin of Lake
Mansfield, C. S., and J. A. Cartwright, 1996, High resolution dis- Malawi, Africa: Seismic stratigraphic analysis: Journal of Sedi-
placement mapping from 3-D seismic data: Journal of Struc- mentary Research, v. 69, p. 832 – 846.
tural Geology, v. 18, p. 249 – 263. Tearpock, D., and R. E. Bischke, 1991, Applied subsurface geo-
Merki, P., 1972, Structural geology of the Cenozoic Niger Delta, in logical mapping: New York, Prentice-Hall, 780 p.
T. F. J. Dessauvagie and A. J. Whiteman, eds., African geology: Thornburg, T. M., L. D. Kulm, and D. M. Hussong, 1990, Submarine-
Ibadan, Nigeria, Ibadan University Press, p. 635 – 646. fan development in the southern Chile Trench: A dynamic in-
Mitchell, N. C., 1996, Creep in pelagic sediments and potential for terplay of tectonics and sedimentation: Geological Society of
morphologic dating of marine fault scarps: Geophysical Re- America Bulletin, v. 102, p. 1658 – 1680.
search Letters, v. 23, p. 483 – 486. Vannier, F., and J. Durand, 1994, Stratigraphie haute résolution
Morris, S. A., J. Alexander, N. H. Kenyon, and A. F. Limonov, d’un réservoir miocène du delta du Niger: Bulletin Centres de
1998, Turbidites around an active fault scarp on the lower Recherches Exploration Production Elf Aquitaine, v. 18, no. 1,
Valencia fan, northwest Mediterranean: Geo-Marine Letters, p. 1 – 18.
v. 18, p. 165 – 171. Webb, H. F., and T. H. Jordan, 2001, Pelagic sedimentation on
Nelson, C. H., E. B. Karabanov, S. M. Colman, and C. Escutia, rough seafloor topography: 1. Forward model: Journal of Geo-
1999, Tectonic and sediment supply control of deep rift lake physical Research, v. 106, no. B12, p. 30,433 – 30,449.
turbidite systems: Lake Baı̈kal, Russia: Geology, v. 24, no. 2, White, N. J., J. A. Jackson, and D. P. McKenzie, 1986, The
p. 162 – 166. relationship between the geometry of normal faults and that of
Newell, A. J., 2000, Fault activity and sedimentation in a marine the sedimentary layer in the hanging walls: Journal of Struc-
rift basin ( Upper Jurassic, Wessex basin, UK): Journal of the tural Geology, v. 8, p. 897 – 909.
Geological Society (London), v. 157, p. 83 – 92. Whiteman, A., 1982, Nigeria: Its petroleum geology, resources and
Ocamb, R. D., 1961, Growth fault of south Louisiana: Transactions potential: London, Graham and Trotman, v. 2, 394 p.

Pochat et al. 1367

View publication stats