Journal of African Earth Sciences 120 (2016) 248e265

Contents lists available at ScienceDirect

Journal of African Earth Sciences

journal homepage: www.elsevier.com/locate/jafrearsci

Investigating fault propagation and segment linkage using throw

distribution analysis within the Agbada formation of Ewan and Oloye
fields, northwestern Niger delta
Abimbola Adewole Durogbitan
Petrofac IES, UK

a r t i c l e i n f o a b s t r a c t

Article history: Throw distribution analysis of the key stratigraphic surfaces (sequence boundaries and maximum
Received 9 April 2015 flooding surfaces) across faults has allowed detailed investigation of the tectonic history within the Ewan
Received in revised form and Oloye fields, northwestern Niger delta. The structure in the studied area is dominated by growth
17 April 2016
fault systems which are listric in cross section and concave to the basin in plan-view. Generally, the faults
Accepted 18 April 2016
are active down to 2000 m depth before they die out or sole into the underlying shale. The hanging-wall
Available online 23 April 2016
blocks of growth faults are deformed into broad rollover anticlines, with some synthetic and antithetic
faults initiated from the anticline crests, and fault splays off major faults, further complicating these
Keywords:
Growth fault
structures. Stratigraphic key surfaces within the syn-faulting succession range in age from 16.7 to
Akata formation 10.35 Ma.
Agbada formation Periods of maximum and minimum throw are established from 2-Dimensional throw distribution on
Listric the growth fault plane. Throw distribution allows analysis of growth fault nucleation, propagation and
Ewan linkage. Each fault nucleated at different and a distinct interval within the stratigraphic section, as a
Oloye result of the paleo-stress distribution between the interacting faults. Nucleation and linkage positions
Rollover anticline can be identified at points of maximum and minimum throw respectively. Following nucleation, faults
Nucleation
propagated radially and linked to form the present geometry. Within the study area, fault propagation
Segment
and segment linkage (lateral and vertical) are important features of the fault system. Understanding of
Linkage
Niger Delta growth fault evolution and linkage has greatly improved prediction of seal potential, trap geometry and
Kink migration. The accurate timing of the segment linkage has helped to evaluate the seal risk.
© 2016 Elsevier Ltd. All rights reserved.

1. Introduction
Early models for normal fault systems were based on the
interpretation of 2-dimensional seismic data (Gibbs, 1984;
Hardman and Booth, 1991; Robert and Moores, 1992) where they
were considered to be relatively simple, single plane structures
with no segmentation. Recent studies on normal fault growth
systems have shown that single/larger fault systems often evolve by
a process of fault segment interaction and linkage (Mansfield and
Cartwright, 1996; Pivnik et al., 2003; Hus et al., 2005, 2006;
Baudon and Cartwright, 2008). Segmentation of normal faults has
been documented by numerous authors (e.g. Peacock and
Sanderson, 1991; Trudgill and Cartwright, 1994; Cartwright et al.,
1996; Walsh et al., 2003b). Segments may initially nucleate as
E-mail address: adewoledurogbitan@yahoo.co.uk.
discrete faults (Cowie et al., 2000), but with increasing displace-
ment and lateral growth they begin to interact and eventually link
(Peacock and Sanderson, 1991; Cartwright et al., 1995). Fault linkage
throw distribution patterns are schematically illustrated in Fig. 1.
The location of the study area is shown in Fig. 1b.
The Niger Delta is characterized by a sedimentation rate that
exceeds the subsidence rate (Doust and Omatsola, 1990), a condi-
tion that preserved the displacement histories of faults as stratal
thickness changes across syn-depositional faults. Understanding
the style of the displacement accumulation (continuous or step-
wise) and the rate of movement of growth faults is important for
the prediction of stratigraphic architecture of oil and gas accumu-
lations and also gives insight into how fault systems develop
through time. Several works on the interaction between fault
displacement and sedimentary processes have been carried out
which show the relationship between tectonics and sedimentation
(Weber and Daukoru, 1975; Doust and Omatsola, 1989; Mansfield

http://dx.doi.org/10.1016/j.jafrearsci.2016.04.014
1464-343X/© 2016 Elsevier Ltd. All rights reserved.
 A.A. Durogbitan / Journal of African Earth Sciences 120 (2016) 248e265 249

Fig. 1. a-d). End members of fault evolution. Arrows indicate the fault tips. e) Showing the location of the study area and age of deltaic sequences in depobelt and relationship to the
broad changes in tectonic style (Doust and Omatsola, 1990).

and Cartwright, 1996; Cartwright et al., 1998; Hodgetts et al., 2001;
Hooper et al., 2002; Imber et al., 2003; Pochat et al., 2004 and Back
et al., 2006). The characteristics of some growth faults in the Niger
Delta interpreted to be active are poorly understood. A better un-
derstanding of these faults will increase our understanding of
previously and presently active geological processes in the Niger
Delta, especially the characteristic growth, interaction and linkage
of faults. In this study, growth fault evolution and associated deltaic
sedimentation in the shallow marine deposits of Ewan and Oloye
fields, northwestern Niger Delta has been investigated using throw
distribution analysis. The main factors determining the throw dis-
tribution within the syn-faulting sequences along isolated syn-
sedimentary faults are the rate of sedimentation and the
displacement rate. Throw distribution analysis for eight growth
fault tips have been analyzed using TrapTester™ software with the
aim of providing quantitative constraints on fault growth models
and the growth fault evolution history. To obtain a detailed record
of normal/growth fault development it must be possible to corre-
late horizons across the fault from footwall to hanging wall, and the
growth strata must contain key surfaces, the ages of which are
250 A.A. Durogbitan / Journal of African Earth Sciences 120 (2016) 248e265
biostratigraphically constrained. Correlation of growth stratigraphy
across the fault was achieved by systematic visualization and
interpretation of key surfaces (sequence boundaries and maximum
flooding surfaces) and depositional patterns (e.g. Durogbitan et al.,
2008, 2010). This has allowed a sedimentologically and structurally
controlled link between the observed footwalls and hanging walls
in the study area.
2. Overview of the structural geometry within the study area
The Niger Delta has a distinctive structural and stratigraphic
zonation. Regional and counter-regional growth faults, developed
in an outer-shelf and upper-slope setting, are linked, via a trans-
lational zone containing shale diapirs, to a contractional zone
defined by a fold-thrust belt that developed in a toe-of-slope
setting (Damuth, 1994). The area studied falls within the outer-
shelf setting characterized by growth faults.
The structure in the area of Ewan and Oloye fields, northwestern
Niger Delta is dominated by several growth faults that strike
eastwest and dip basinward to the south (Fig. 2). These interpreted
growth faults are listric in cross section and concave to the basin in
plan-view (Fig. 3). The observed arcuate geometry of some of these
fault traces (Fig. 3) show significant “kinks” (bends in fault traces)
that could indicate linkages precursor fault segments (e.g.
Cartwright and Trudgill, 1994; Morley et al., 2007). The hanging
wall blocks of growth faults are deformed into broad rollover an-
ticlines, with double plunging hinges parallel to the adjacent
cuspate fault trend. Some synthetic and antithetic faults that
initiated from anticline crests and fault splays off major faults
further complicate these structures. The initiation of these anti-
thetic and synthetic faults may have been caused by the stress
generated as a result of underlying mobile shale (depth below
2500 m) and sediment loading.
3. Offset/isopach maps of stratigraphic surfaces across
growth faults
Depth structure and isopach maps of the interpreted key sur-
faces (the sequence boundaries and maximum flooding surfaces)
reflections provide a record of the depth of structural offset across
major growth fault and the pattern of deformation within the fault
blocks (Fig. 6.8 to 6.26; Durogbitan, 2010). Variable fault throw
distributions were observed which may be due to fault growth and
initiation. This may be linked to irregular stress distribution within
the fields, where strain accommodation is being shifted from one
fault to the other. In some areas within the seismic section, older
surfaces show significantly greater offset across faults than the
younger surfaces (a case of trend of decreasing throw with
decreasing age), thus demonstrating syn-depositional fault
displacement and variability in the throw distribution (Fig. 4).
Areas of greatest throw are linked with the progradation of clastic
wedge and rapid subsidence in progressively more distal area of the
basin over a period of time. This was observed at depth between
1200 and 1600 m within the interpreted faulted section, shown as
black open arrow and red thick line in Fig. 4. In some of the faults,
the older surfaces show a decreased offset across the fault than the
younger surfaces in dip direction (see faults 07 and 08, shown as
black arrow and blue thick line in Fig. 4).
Isopach maps provide an indication of the thickness and loca-
tion of growth strata. Considering the fault throw distribution and
thickness trend between the faults, some of the thickness variation
probably reflects both structural deformation of fault block (Fig. 4)
and change in delta bathymetry (deepening) and or sediment
compaction. The displacement across the faults may have produced
the bathymetric lows in the hanging wall that accumulated
sediment more rapidly and influenced sediment dispersal basin-
ward. The thickness trend mostly reflects incisions, growth fault
and prograding HST clinoform geometry (Durogbitan, 2010). De-
posits in the hanging wall blocks tend to be thicker directly
basinward of areas showing the greatest stratigraphic offset across
the faults and relatively thinner down the basin from areas with
lesser fault throws. In other words, stratigraphic offset is greater at
the centre of the fault segments.
4. Throw distribution across fault array
Several descriptions of throw distribution over the surface of the
normal fault have been carried out (Childs et al., 1995; Nicol et al.,
1996; Rowan et al., 1998; Childs et al., 2003 and Baudon and
Cartwright, 2008). The fault evolution histories of several normal
faults from different tectonic and sedimentary setting have been
discussed with illustration of the main processes that control
displacement distribution (Childs et al., 2003). Gupta et al. (1998)
suggested that the evolution of a fault population has two
distinctive stages; an initial stage characterized by growth of fault
nucleated independently with weak elastic interaction followed by
a state of fault linkage. Whether the growth faults within the study
area grew in this way still remains an open question. The throw
distribution study comprises eight interpreted growth faults. In
order to constrain and evaluate the rates at which faults have
grown and interacted, it is necessary to understand the strati-
graphic framework and delineate key surfaces constrained with
biostratigraphic ages within the sediment package (Durogbitan,
2010). This approach has been used for throw analysis in this
study. Within the interpreted faults and seismic section, eight
stratigraphic surfaces were used. The total age ranges of the
interpreted section are within 10.35e16.7 Ma (Table 1) based on the
regional correlations of Doust and Omatsola (1990), see
(Durogbitan, 2010). An upward decrease in thickness in the
hanging wall of the fault blocks is observed between the isopach of
successive sequence boundaries (Table 1) within the studied area.
This may be due to upward decrease in movement along some
faults tip which in turn reduced the accommodation space.
5. Distribution of throw/throw analysis
The 3D dataset was interpreted to obtain the throw (vertical
separation) on all the interpreted faults and the data from the study
area indicated that the growth fault represent an example of fault
linkages. Some portions of the fault are active and some are inactive
which reflects the complex throw distribution. Since the key
markers (sequence boundaries-channel incisions) in the hanging
wall of the growth faults are traceable, it makes it possible to
evaluate real throw of the fault and to reconstruct the tectonic
history of the fault. Two dimensional throw distribution maps were
developed using TrapTester™ software. The points marked as cut
off values from the seismic section may be subject to some small
errors, namely: uncertainty in the velocities employed for depth
conversion, quality of the seismic data and the precision error i.e.
error due to accurately determining the correct cut off position,
which may be due to some of the growth along the faults surfaces-
horizon drag (Bull et al., 2006).
For some of antithetic and synthetic faults (smaller faults),
throw values are too small to resolve the detailed evolution of fault
displacement. But for larger faults, the detailed evolution has been
resolved. Eight faults and their generated throw distribution maps
are discussed below.
 A.A. Durogbitan / Journal of African Earth Sciences 120 (2016) 248e265
Fig. 2. a). Shows uninterpreted seismic section of inline 10541. b) Interpreted seismic section of the same inline showing the geometry of dominant structure (growth fault) within
the study area. See position in Fig. 3.
6. Fault 3a
There is a large variation of throw along this fault between
horizon 13 (MFS), shown as solid black arrow, and horizon 12 (SB),
shown as open black arrow in Fig. 5. This variation may be due to
bathymetry control as a result of deepening effect causing subsi-
dence, this reflects a stratigraphic control. Fault 3a is initiated
within this interval on at a depth of 1200 m (Fig. 5) and begins to
grow and die off (basinward). Growth occurs within a distance of
about 1.5 km before the strain being accommodated by fault 3a
shifted to fault 3. The greatest throw is within the position of fault
initiation assumed to conform to the hypothesis of Watterson
(1986) and Dawers et al. (1993), who stated that, the highest
displacement is the centre of the fault while the lowest is at the tip.
Thus reflects increase in the throw observed with horizon 13 and
12. The increase in thickness towards the fault as revealed from the
seismic section (Fig. 5) conforms to the evidence of growth within
the interval. Throw distribution attribute map showed and confirm
251
throw variation between these horizons (Fig. 6). The high throw
(position of fault initiation) is shown as blue circle in Fig. 5, also
denoted as throw high (red line) on Fig. 6.
7. Fault 3
The maximum value of the throw is about 180 m, achieved
between depth 1200e1500 m on inline 10,424 (Fig. 7). The large
variation of throw occurred within horizon 11 (mfs), shown as solid
black arrow, and horizon 10 (SB), shown as open black arrow in
Fig. 7. The variations may suggest some evidence of bathymetry
control (deepening) since the highest variation occurred between
sea level falls (H10; Fig 7) and subsequent maximum sea level rise
(H11). This may be as a result of fault movement (subsidence)
causing flooding effect, thus reflects a stratigraphic control. The
strain being accommodated by fault 3a shifted to fault 3 down dip
(Fig. 7). This has allowed more throw to be accommodated on fault
3 within this stratigraphic position (Horizons 10 and 11). The throw
252 A.A. Durogbitan / Journal of African Earth Sciences 120 (2016) 248e265

Fig. 3. Depth structure map of one of the interpreted key surfaces (sequence boundary
4- Horizon 1- Durogbitan, 2010), showing structural geometry, position of seismic
sections and interpreted faults. Distal area is more tectonically faulted. Red small
dotted lines represent position of “kink” e bending in fault traces). (For interpretation
of the references to colour in this figure legend, the reader is referred to the web
version of this article.)

variation down-dip within this fault may reflect seaward pro-

gradation that led to sediment loading. This may have resulted in
the creation of more accommodation within the hanging wall
block. Upward, the progradation is reduced which may be due to
low sediment supply during transgressions and fault growth to-
wards the tip. The high throw (position of fault initiation) is shown
as blue circle in Fig. 7, also denoted as throw high (red line) on Fig. 8.

8. Fault 6
The maximum value of throw is about 300 m, located between
depth 1400e1600 m intervals on inline 10,520. This fault initiated
at stratigraphic horizon 13 (MFS) as confirmed from the highest
throw (Fig. 9), shown as solid black arrow and 12 (SB), shown as
open black arrow on inline 10,472 (Fig. 9). This fault covered the
Fig. 4. Seismic cross section illustrating structural offset of interpreted surfaces across
growth faults and the pattern of deformation within the fault block (Inline 10536).
Areas of greatest throw is shown as black open arrow and red thick line while area of
decreasing offset (fault tip) within the older surfaces is shown as solid black arrow and
blue thick line. (For interpretation of the references to colour in this figure legend, the
reader is referred to the web version of this article.)
 A.A. Durogbitan / Journal of African Earth Sciences 120 (2016) 248e265 253

Table 1
Age estimates and thickness for the depositional sequences within the studied area.
Depositional sequences Thickness (m)
1 480
2 410
3 380
4 370
5 320
6 300?
structure map view (Fig. 3) which may suggest fault linkages i.e.
fault is not originally a single isolated faults but a combination of 2
Age of the base layer (Ma) or more faults that were initially spatially separated (Fig. 10). The
16.7
upper fault segment (segment 2) was assumed to be initiated at
15.5 depth about 1000 m within interpreted horizon 2, shown as solid
13.1 black arrow in Fig. 10. This effect was observed from the throw
12.1 distribution map attribute (Fig. 11) as a variation in throw within
10.6
the fault. In between the intervals of the large throw (initiation
10.35
zone) there are assumed zone of fault linkage (palaeo-fault tips).

Fig. 5. Seismic section showing where fault 3A initiated and (inline 10406). Fault initiation occurred at depth about 1200 m between Horizon 12 (interpreted SB) and Horizon 13
(interpreted MFS).

Fig. 6. Throw distribution attributes map of fault 3A (strike view). Red, yellow and green colours denote position of high throw while blue and purple colours denote position of low
throw. T indicates the position of high throw (fault initiation) circled in white dotted lines. (For interpretation of the references to colour in this figure legend, the reader is referred
to the web version of this article.)

extent of the interpreted survey data. Within the length of the fault, This is mostly observed at the southern part and within the middle
“kink” geometry (bending in fault trace) was observed in depth part of the study area. The throw distribution attribute map (Fig. 11)
254 A.A. Durogbitan / Journal of African Earth Sciences 120 (2016) 248e265

Fig. 7. Seismic section showing where fault 3 initiated and (inline 10424). Fault initiation occurred at depth betweem1200e1500 m between Horizon 10 (interpreted SB) and
Horizon 11 (interpreted MFS).

Fig. 8. Throw distribution attributes map of fault 3 (strike view). Red, yellow and green colours denote position of high throw while blue and purple colours denote position of low
throw. T's-Indicates the position of high throw (fault initiation). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this
article.)

revealed more evidence of vertical segmentation than lateral seg-
mentation in fault 6 (Fig. 12). Below is the schematic illustration of
fault 6 development (Fig. 12). There is initiation of antithetic fault at
the hanging wall block of this fault within the rollover anticline
(Fig. 10).
9. Fault 7
Evidence of fault linkage at the dip and strike direction was also
observed within this fault. From the seismic survey the fault was
observed to initiate at the basal horizon, at about depth of 1600 m
as revealed from the highest throw valve (Figs. 13 and 15) and stop
growing upward around horizon 1 as revealed from the lowest
values of throw (Figs. 14 and 15), shown as solid black arrow in
Fig. 13. Downward, this fault segment is active up to the depth
2100 m before it dies off, as it approaches the underlying mobile
Akata shale or soles into the mobile shale (Fig. 13). On inline 10,480,
there is also evidence of fault initiation within the unconformity 2
as confirmed from the highest throw value (Figs. 14 and 15), shown
as open black arrow and horizon 2, shown as solid black arrow in
Fig. 14. This upper fault segment later joined together with the
down dip segment (Fig. 14), suggesting that fault 7 is made up of
two fault segments linked together along the dip. This may reflect
that strain being accommodated by the lower segment is shifted to
upper fault segment or a consequence of fault growth. The inter-
action and linkages became more significant as fault grew towards
each other with fully linked fault system forming between depth
intervals 1200e1500 m.
The throw distribution attribute map displayed these observa-
tions (Fig. 15) that from the north the fault initiates at the basal
horizon as shown by the increase in throw upward. Towards the
middle of the fault (towards the south), the strain shifted more to
 A.A. Durogbitan / Journal of African Earth Sciences 120 (2016) 248e265 255

Fig. 9. Seismic section showing where fault 6 initiated and (inline 10472). Fault initiation occurred at depth betweem1400e1700 m between Horizon 12 (interpreted SB) and
Horizon 13 (interpreted MFS).

Fig. 10. Seismic section of fault 6 at (inline 10545), showing the position of antithetic fault that developed at the hanging wall block within the rollover anticline of the fault. See
position in Fig. 3 as “kink”- bending of fault trace.

the upper segments. The upper fault segment remains active
further to south. Fig. 16 show a schematic illustration of fault 7
development as observed from the throw distribution attributes
map. This zone corresponds to a strong kink in the mapped trace of
Fault 7 (Fig. 3), which supports the interpretation of linkage of
previously isolated fault segments at this location. This fault
showed both lateral and vertical propagation (Fig. 16a), but with
more evidence of lateral segmentation than vertical segmentation
as revealed from the throw distribution attribute map (Figs. 15 and
16), and schematic illustration in Fig. 1d.
10. Fault 8
The maximum value of throw is about 350 m, located at depth
interval 1000e1400 m on inline 10,464 (Fig. 17), the fault initiated
at the back of fault 7. The lower segment assumed to be initiated at
horizon 10, shown as open black arrow while the upper segment
initiated at horizon 9, shown as solid black arrow in Fig. 17. This
fault later merged with fault 7 as one single isolated fault on inline
10,600 (Fig. 18). From the throw distribution attribute map (Fig. 19),
the strain was shifted/localized more towards the upper segment.
Towards the south (seaward), there is a decrease in the throw
which may be due to interaction between fault 7 and 8 and/or shale
tectonics (uplift) before the two faults merge together. Below is the
schematic illustration of fault 8 development based on the throw
distribution attributes map (Fig. 20). Observations from throw
distribution attribute maps show more evidence of vertical seg-
mentation than lateral segmentation. There is also evidence of
“kink” as observed from the fault trace of this fault (Fig. 3), that
supports the interpretation of fault linkage. This two fault segments
probably initiated at the same time as revealed from throw distri-
bution attributes map, but the upper segment accommodated more
strain (Fig. 19). Lower segment of fault 8 dies off, as it approaches, or
soles out into the mobile Akata shale.
11. Fault 9
The maximum throw on fault 9 is about 250 m. From the
northern part, the fault initiated at stratigraphic surface denoted as
Horizon 2, shown as black arrow in Fig. 21. This initiation may be
256 A.A. Durogbitan / Journal of African Earth Sciences 120 (2016) 248e265

Fig. 11. Throw attributes map of fault 6 (strike view). Red, yellow and green colours denote position of high throw while blue and purple colours denote position of low throw. Fault
06 showed more evidence of vertical segmentation than lateral segmentation. (For interpretation of the references to colour in this figure legend, the reader is referred to the web
version of this article.)

Fig. 12. Schematic illustration of fault 6 development as observed from the attribute map Fig. 11. (a) Showing the interpretation of fault segment as revealed from the displacement
attributes along the strike. (b) Showing the fault linkage of the segments along the dip while (c) final isolated developed fault. Fault 06 showed more evidence of vertical seg-
mentation than lateral segmentation.

Fig. 13. Seismic section showing where fault 7 initiated (Inline 10472). Fault initiation occurred at depth about 1600 m between Horizon 12 (interpreted SB) and Horizon 13
(interpreted MFS).
 A.A. Durogbitan / Journal of African Earth Sciences 120 (2016) 248e265 257

Fig. 14. Seismic section showing where fault 7 initiated (Inline 10480), showing initiation position of upper fault segment (segment 2) at depth of about 1100 m (between Horizon 2
(interpreted SB) and Unconformity 2 (interpreted SB). The two fault segments linked together to form an isolated fault (fault 7).

Fig. 15. Throw distribution attributes map of fault 7 (strike view). Red, yellow and green colours denote position of high throw while blue and purple colours denote position of low
throw. Fault 07 showed more evidence of lateral segmentation than vertical segmentation. (For interpretation of the references to colour in this figure legend, the reader is referred
to the web version of this article.)

due to rapid progadation and sediment loading. Towards the
middle of the interpreted fault another segment initiated at Hori-
zon 1, shown as open black arrow in Fig. 21. Towards the south
(seaward), there is another initiation within the horizon 11, shown
as open red arrow in Fig. 21. It was observed that there is reduction
in throw at the basal horizon which may reflect fault tip or possibly
effect of shale tectonics. From the throw attribute map (Fig. 22),
fault 9 showed more evidence of lateral segmentation than vertical
segmentation (Fig. 1). Below is the schematic illustration of fault 9
development within the study area (Fig. 23). This fault is made up
at least 3 segments as observed from the seismic section inline
10,400 to 10,640 (Fig. 21; inline 10,528), throw distribution attri-
butes map (Fig. 22) and depth structure map show significant
“kinks” (Fig. 3).
12. Fault 11
The maximum value of throw is about 300 m. This fault is active
within the depth 1200e2000 m interval (Fig. 24; inline 10,432).
There is variation in throw within this depth interval along the fault
which relates to stratigraphic control and/or fault segmentation.
The fault is more active at some interval than others within the
study area as revealed from the attribute map that show inter-
mittent of several high and low throws intervals (Fig. 25). This may
reflect sediment input, lithology type and differential compaction,
subsidence as local controls. At the middle of the throw distribution
attribute map (Fig. 25), there is low value of throw in the strike
direction, shown as open black arrow. This low may be due to
initiation of antithetic fault in the footwall of fault 11 which reflect
some evidence of lateral segmentation. This antithetic fault initi-
ated at inline 10,456 (Fig. 26) and dies off at inline 10,520 then
reactivated or becomes active again at inline 10,568. The antithetic
fault later dies off at inline 10,624. As soon as the antithetic fault
dies off fault 11 is activated and continues to grow seaward again.
The related low throw shown as open black arrow on Fig. 25 reflects
the position of a kink on the depth structure map (Fig. 3), which
could suggest lateral fault segment linkage. The thin linear lows
intercalated with linear high area, shown as solid black arrow on
Fig. 25, is be attributed to the effects of differential compaction of
lithologies (interpreted horizons) as local control. As observed from
258 A.A. Durogbitan / Journal of African Earth Sciences 120 (2016) 248e265

Fig. 16. Schematic illustration of fault 7 development as observed from the attributes map. (a) showing both lateral and vertical propagation that leads to linkage. See Morley et al.
(2007) (b) showing the interpretation of fault segment as revealed from the displacement attributes along the strike. (c) showing the fault linkage of the segment 1 and 2 along the
dip while (d) is the final isolated developed fault, the two fault segments linked together to form an isolated fault (fault 7). Fault 07 showed more evidence of lateral segmentation
than vertical segmentation.

Fig. 17. Seismic section showing where fault 8 initiated and the synthetic relationship with fault 7 (Inline 10464). The fault consist of two fault segments assumed at the same time
(segment 1 at horizon 10- interpreted SB; depth 1200 m) and (segment 2 at horizon 9-interpreted MFS; depth 1700 m).

the throw distribution attribute map (Fig. 25), this fault showed
evidence in both lateral and vertical segmentation.
13. Fault 12
The maximum value of throw is about 320 m, and occurs at a
depth of 1500e2000 m on inline 10,528 (Fig. 27). Throw on this
fault increases with depth (Fig. 27). The increase of throw at the
basal horizons may be linked to structural deformation (rollover
anticline) and change in stress (Morley et al., 2007). Evidence of a
rollover anticline is observed within the hanging wall block of fault
12, which may reflect effects of rapid sedimentation, fault
displacement and/or shale tectonics (uplift) below the interpreted
section (Fig. 27). This effect caused the strata at the hanging wall
block to be tilted upward (curved-rollover anticline) which appar-
ently cause increase in throw. This effect is illustrated in a sche-
matic diagram in Fig. 29. It is also observed from the throw
distribution attribute map (Fig. 28), this fault showed evidence in
both lateral and vertical segmentation.
14. Geometry and evolution of fault system
The structure in the area of Ewan and Oloye fields, northwestern
Niger Delta is dominated by a growth fault system that strikes east-
west and dips basinward (to the south). The growth faults are
characterized by broad convex-basinward plan-view geometry,
 A.A. Durogbitan / Journal of African Earth Sciences 120 (2016) 248e265 259

Fig. 18. Seismic section showing where fault 8 merged with fault 7 to become a single fault (Inline 10600).

Fig. 19. Throw distribution attributes map of fault 8 (strike view). Red, yellow and green colours denote position of high throw while blue and purple colours denote position of low
throw. Fault 08 showed more evidence of vertical segmentation than lateral segmentation. (For interpretation of the references to colour in this figure legend, the reader is referred
to the web version of this article.)

Fig. 20. Schematic illustration of fault 8 development as observed from the attributes map. (a), showing the interpretation of fault segment as revealed from the displacement
attributes along the strike. (b) showing the fault linkage of the segment 1 and 2 along the dip (c) final isolated developed fault 8, the two fault segments linked together to form an
isolated fault (fault 08). Fault 08 showed more evidence of vertical segmentation than lateral segmentation.
260 A.A. Durogbitan / Journal of African Earth Sciences 120 (2016) 248e265

Fig. 21. Seismic section showing where segments that developed into single isolated fault 9 initiated (Inline 10,528). The fault consist of three fault segments assumed to be
initiated at Horizon 2 (upper segment-sb; c1000 m), Horizon1 1 (middle segment-sb; 1400 m) and horizon 11 (lower segment, interpreted mfs; depth c1800 m) and (segment 2 at
horizon 9-interpreted MFS; depth 1700 m).

Fig. 22. Throw distribution attributes map of fault 9 (strike view). Red, yellow and green colours denote position of high throw while blue and purple colours denote position of low
throw. Fault 09 showed more evidence of lateral segmentation than vertical segmentation. Solid white arrow show positions of lateral and vertical linkages. (For interpretation of
the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 23. Schematic illustration of fault 9 development as observed from the attributes map and seismic sections. (a), showing the interpretation of fault segment as revealed from
the displacement attributes along the strike. (b) showing the fault linkage of the segment 1, 2 and 3 along the dip (c) final isolated developed fault 9, three fault segments linked
together to form an isolated fault (fault 9). Fault 09 showed more evidence of lateral segmentation than vertical segmentation.
 A.A. Durogbitan / Journal of African Earth Sciences 120 (2016) 248e265 261

Fig. 24. Seismic section showing geometry and position of single isolated fault 11 (Inline 10432). The fault shows variation in throw within depth interval (1200e2200 m which
may relate to stratigraphic control or/and fault segmentation.

Fig. 25. Throw distribution attributes map of fault 11 (strike view). Red, yellow and green colours denote position of high displacement while blue and purple colours denote
position of low throw. Fault 11 showed evidence in both lateral and vertical segmentation. (For interpretation of the references to colour in this figure legend, the reader is referred
to the web version of this article.)

Fig. 26. Seismic section showing geometry and position of single isolated fault 11 and associated antithetic fault at the back of fault 11 (Inline 10456). This antithetic fault initiated
on this inline (10456), dies off on inline 10520, reactivated on inline 10568 and later dies off at inline 10624 within the interpreted section.
262 A.A. Durogbitan / Journal of African Earth Sciences 120 (2016) 248e265

Fig. 27. Seismic section showing geometry and position of single isolated fault 12 (inline 10528). The fault shows variation in throw within depth interval which also may relate to
stratigraphic control or/and fault segmentation. Development of this fault is similar to fault 11.

Fig. 28. Throw displacement attributes map of fault 12 (strike view). Red, yellow and green colours denote position of high displacement while blue and purple colours denote
position of low throw. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 29. Schematic structural interpretation of seismic section of inline 10528, illustrating evidence of tilting of stratal (rollover anticline) on the interpreted key surfaces across the
growth fault (Not to scale).

decreasing in dip with depth. The hanging wall blocks of growth
faults are deformed into broad rollover anticlines, with some syn-
thetic and antithetic faults initiated from the anticline crests and
splays off major faults further complicate these structures. The
initiation of antithetic and synthetic faults complicated the throw
distributions of the growth faults, which reflects the evolution of
the fault array.
Throw distribution analysis of growth faults has allowed
investigation of the tectonic history within the syn-faulting suc-
cession of the study area. Periods of maximum throw were estab-
lished from seismic sections and 2-dimensional throw distribution
attribute maps. Throw for individual faults are generally variable
due to the fault initiation, segment linkage and shale tectonic effect.
Within the growth interval, fault propagation and segment linkage
is an important feature of these fault systems. Throw variations
seen within the segmented faults indicate some general features
 A.A. Durogbitan / Journal of African Earth Sciences 120 (2016) 248e265 263

such as: - (i) throw typically decreases upwards, suggesting that the
interpreted segmented faults are behaving as a kinematically
coherent system, (ii) variation in the throw distribution occurs over
time i.e. there is considerable variation in the throw distribution at
different time interval (horizon interval), and (iii) linkages between
fault segments are achieved by tip-to-fault linkage. Generally, the
faults are mostly active above 2000 m below which they sole into
the underlying mobile Akata shale which serves as a decollment
surface for the clastic deposition. The fault analysis of the study area
revealed fault interactions which may be due to constructive and
destructive interference pattern of stress field distribution (Back
et al., 2006).
Although, throw distributions in a growth fault usually increase
with depth, there is a decrease in fault throw from the top to base in
most of the analyzed faults. The reduction of the throw downward
to the basal horizons may be because the faults observed in the
study area were originally segments that have since linked, and the
variation in throw with depth reflects the location of the tip line of
two interacting faults. The complicating effect of shale diapirism
may also modify the fault throw with depth. This study has shown
that most of the faults within the interpreted section did not
originally develop as a single isolated fault; rather initially isolated
segments which subsequently linked together to form single
through-going faults. This agrees with the fault segment linkage
models of e.g. Gupta et al. (1998); Childs et al. (2003), suggesting
that the faults within the study area followed two distinctive
stages; an initial stage characterized by fault nucleation followed by
a stage of fault linkage (Fig. 30). Most of the complexities seen in
the syn-sedimentary faults within the study area are due to these
processes; fault initiation, high sedimentation rate and linkage.
Observations from the sequence analysis show that throughout
sequence development within the study area, the shelf margin has
also been controlled by growth fault which has migrated basinward
through time (Appendices 6A, 6C and 6D; Durogbitan, 2010). In this
case, one of the processes of transporting sand into the basin is Fig. 30. Schematic diagram illustrating the evolution of Fault 6 on Sequence 1 (Fig. 31),
suggesting that the faults within the study area followed two distinctive stages; an
shelf collapse, with resultant slumping. There is a significant
initial stage characterized by fault nucleation followed by a stage of fault linkage. On
reduction in the number of growth faults, from the deepest horizon Fault 6, the depocentres is related to fault segmentation, with a breached relay ramp.
to the shallowest. This may be due to changes in sediment depo-
sition rates (rapid sediment loading) causing minor faults to
become buried and truncated; fault segmentation and differential observed in the fault traces in map view (Fig. 3). There is agreement
displacement on adjacent segments. Sedimentation patterns sug- with the work of Childs et al. (2003) and Gupta et al. (1998), who
gest that faulting, local incision and HST clinoform geometry proposed that evolution of a fault population has two distinctive
dominate the formation of depocentres across the shelf stages; initial stage characterized by fault nucleation followed by a
(Durogbitan, 2010). stage of fault linkage. Most of the complexities seen in the syn-
sedimentary faults within the study area are due to these pro-
15. Synthesis and discussions cesses; fault initiation, high sedimentation rate, linkages and shale
tectonics. It is evident that the fault segment geometry controls the
The results of the fault throw analysis have shed more light on distribution of throw and this is variable on different time (hori-
the variability and complexity of interplay between tectonic and zons). The variation of the throw at the basal horizons may be
sedimentation within the study area. Integration with sequence linked to structural deformation (shale tectonics) or the effect of
stratigraphy has been used to unravel the detail of growth fault fault movement, causing the sedimentary strata to be tilted thereby
evolution and segmentation. The fault analysis revealed fault in- developing a rollover anticlinal structure. Change in stress may
teractions which may be due to constructive and destructive have caused most throw variations (Morley et al., 2007). This
interference pattern of stress field distribution and thus allow variation and complexities may also be attributed to the activation
detailed reconstruction of the fault geometry and tectonic history and nucleation nature of fault with the surrounding which caused
within the study area. Offset of horizons by the faults shows that the strain to be shifted to the fault that is active at that particular
the vast majority of the interpreted faults have greatest offset be- time. Decrease in the vertical offset of stratigraphic surfaces across
tween depth 1200 and 1800 m of about 3 km within the study area. faults basinward were also observed in most cases, and this reflects
Passing vertically downward the faults partially or completely lose decrease in accommodation rate as progressively sediment by-
displacement towards the top of the mobile Akata shale. passed to more distal basin area.
Faults within the study area have been interpreted not to have However, the throw variation seen within the segmented faults
originally evolved as single isolated faults. They are mostly may indicate some general features such as: (a) The complex sys-
segmented faults that linked together through time to form the tems of the segments maintain a fairly continuous throw variation,
present through-going fault geometry. This segmentation has suggesting that the interpreted segmented faults are behaving as
resulted into throw variability and significant kinks that are kinematically coherent system. (b) Variation in the throw
264 A.A. Durogbitan / Journal of African Earth Sciences 120 (2016) 248e265
Fig. 31. Thickness map between Sequence Boundary 1 and Sequence Boundary 2. The isopach map shows thicks as multi-lobate to lunate bodies but generally more lobate.
Generally, T's are the position of thicks, white lines are the fault traces and contour interval is 20 m (Durogbitan, 2010).
distribution occurs overtime. (c) There is considerable variation in
the throw distribution at different time interval (horizon interval).
(d) Linkages between fault segments are achieved by tip-to-fault
linkage.
Based on the fault throw distribution analyses, three tectonic
episodes are proposed for the fault evolution within the study area
(middle Miocene).
1. Early middle Miocene active extension that produce thickening
of the lower part of the sequences
2. Late middle Miocene event during which most of the original
extension was tilted to form extensional rollover anticlines. This
may be caused by shale tectonics, growth faulting, rapid sedi-
mentation and some fault initiation surrounding the major
faults. These phenomena caused active tilting of the hanging
wall of the fault block.
3. Early Pliocene event: normal reactivation of fault initiated as a
result of decrease in accommodation and increase in sediment
supply and loci of the sedimentation that have shifted basin-
ward. This tectonic activity has provided most of the present day
extensional displacement distribution.
4. Faults in the study area still have extensional displacement to
the present day.
Fault segmentation and development of rollover anticlines in
the growth structures have implications for the petroleum system.
It has been shown that the throw distribution are variable across
the growth faults within the studied area with the assumption from
the previous works that the highest throw occurring where the
fault initiated i.e. the centre of the fault and dies towards the fault
tip. This variability in throw distribution within the studied area
has implication for the development of the elements of the petro-
leum systems (seal, trap, reservoir and source rocks) on the fields,
with the continuity of the reservoir units mainly affected due to the
offset of the stratal surfaces. This will in turn have an effect on the
fields’ performance during production. The model of fault evolution
described above has given insight into the position and time of fault
activity which may have direct implication on hydrocarbon
migration and sealing of fault within the reservoirs (McClay, 1990).
The accurate timing of the segment linkage would also be invalu-
able for an evaluation of seal risk (Cartwright et al., 2007).
It is worth noting that, within the study area and generally
within the Niger Delta, most of the faults are commonly longer than
the extent of the 3D dataset (Higgins et al., 2007; Weber and
Daukoru, 1975). Therefore, there is a paucity of completely
imaged faults. By and large, all the faults observed within the study
section are segmented, interact with one another and detached at
some point above the mobile shale.
16. Conclusions
It has been shown that the interpretation of a 3D seismic
reflection data from the study area has allowed detailed analysis of
throw history of faults during the middle Miocene. The fault pop-
ulation is composed of segmented faults that are kinematically
linked both laterally and vertically. This implies that growth faults
within the interpreted section did not originally develop as single
isolated faults but as segmented faults that linked together over
time to form single through-going faults. Vertical offset of the
stratigraphic surfaces across faults are variable, and the variability
may be due to the fault initiation, structural deformation (shale
tectonics) caused by the mobile underlying Akata shale and fault
movement. Interpretation of throw analysis presented here can be
directly used to establish repeatable correlations of differential
growth strata in a wide range setting characterized by syn-tectonic
sedimentation. Results of throw distribution from this study shows
that the deposition of deltaic growth strata can be significantly
influenced by a combination of syn-sedimentary movement along
growth faults and the development of fault topography as a result
of these linkages. The identification of matching key surfaces (most
especially incised valleys) on both sides of some of syn-
sedimentary normal faults provides good links between corre-
sponding stratigraphic levels for this study.
Many authors have concluded that fault growth processes
generally involve tip line propagation and linkage of fault (e.g.
Peacock and Sanderson, 1991; Trudgill and Cartwright, 1994;
Cartwright et al., 1996; Mansfield and Cartwright, 1996; Pivnik
et al., 2003; Walsh et al., 2003b; Hus et al., 2005, 2006 and
Baudon and Cartwright, 2008). Evidence of linkage processes have
been observed in the seismic data sets. The work shows that linked
(vertically and laterally) and interacting faults are present in the
study area. In summary, the main findings are: (1) There are evi-
dence for lateral and vertical linkage of fault segments. (2) Seg-
ments show throw minima where there are fault intersections and
interaction with other structures and adjacent fault segments. (3)
There are throw variation along each of the fault segments with
maximum values occurring where the fault initiated.
References
Back, S., Hocker, C., Brundiers, M.B., Kukla, P.A., 2006. Three-dimensional-seismic
coherency signature of Niger Delta growth faults: integrating sedimentology
and tectonics. Basin Res. 18, 323e337.
Baudon, C., Cartwright, J.A., 2008. The kinematics of reactivation of normal faults
using high resolution throw mapping. J. Struct. Geol. 30, 1072e1084.
Bull, J.M., Barnes, P.M., Lamarche, G., Sanderson, D.J., Cowie, P.A., Taylor, S.K.,
Dix, J.K., 2006. High-resolution record of displacement accumulation on an
active normal fault: implication for models of slip accumulation during
repeated earthquakes. J. Struct. Geol. 28, 1146e1166.
Cartwright, J.A., Trudgill, B.D., 1994. Relay-ramp forms and normal-fault linkages,
 A.A. Durogbitan / Journal of African Earth Sciences 120 (2016) 248e265
Canyonlands national park, Utah. Geol. Soc. Am. Bull. 106, 1143e1157.
Cartwright, J.A., Trudgill, B.D., Mansfield, C.S., 1995. Fault growth by segment link-
age: an explanation for scatter in maximum displacement and trace length data
from the Canyonlands grabens of SE Utah. J. Struct. Geol. 17, 1319e1326.
Cartwright, J.A., Mansfield, C., Trudgill, B., 1996. The growth of normal faults by
segment linkage. In: Buchanan, P.G., Nieuwland, D.A. (Eds.), Modern De-
velopments in Structural Geology, Interpretation, Validation and Modelling
Geological Society Special Publication 99, pp. 163e177.
Cartwright, J.A., Bouroullec, R., James, D., Johnson, H.D., 1998. Polycyclic motion
history of Gulf Coast Growth Faults from high resolution kinematic analysis.
Geology 26, 819e822.
Cartwright, J.A., Huuse, M., Aplin, A., 2007. Seal bypass systems. Am. Assoc. Pet.
Geol. Bull. 91 (8), 1141e1166.
Childs, C., Watterson, J., Walsh, J.J., 1995. Fault overlap zones within developing
normal fault systems. J. Geol. Soc. Lond. 152, 535e549.
Childs, C., Nicol, A., Walsh, J.J., Watterson, J., 2003. The growth and propagation of
synsedimentary faults. J. Struct. Geol. 25, 633e648.
Cowie, P.A., Gupta, S., Dawers, N.H., 2000. Implications of fault array evolution for
active synrift depocentres development: insights from a numerical growth
model. Basin Res. 12, 241e261.
Damuth, J.E., 1994. Neogene gravity tectonics and depositional processes on the
deep Niger Delta continental margin. Mar. Pet. Geol. 11 (3), 320e346.
Dawers, N.H., Anders, M.H., Scholz, C.H., 1993. Growth of normal faults: displace-
ment e length scaling. Geology 21, 1107e1110.
Doust, H., Omatsola, E., 1989. Niger delta. Am. Assoc. Pet. Geol. Mem. 48, 210e238.
Doust, H., Omatsola, E., 1990. Niger delta. In: Edwards, J.D., Santogrossi, P.A. (Eds.),
Divergent/Passive Margin Basin AAPG, vol. 48, pp. 201e238.
Durogbitan, A.A., Gawthorpe, R.L., Redfern, J., 2008a. Sequence stratigraphic anal-
ysis of middle Miocene paralic deposit (Agabda formation), Ewan and Oloye
fields, Central Niger delta. In: Proceeding of AAPG Annual Convention and
Exhibition 2008.
Durogbitan, A.A., Gawthorpe, R.L., Redfern, J., 2008b. Seismic straigraphy and
development of deep fluvial incisions (Paleaovalleys?) in the middle Miocene
deposits (Agbada formation), Ewan and Oloye fields, northwestern Niger delta.
In: Proceeding of NAPE Annual Convention and Exhibition 2008.
Durogbitan, A.A., 2010. Seismic, Sequence Stratigraphic and Structural Analysis of
Ewan and Oloye Fields (Middle Miocene), Northwestern Niger Delta. PhD Thesis
unpubl. Earth Science Department, The University of Manchester, p. 337.
Gibbs, A.D., 1984. Structural evolution of extensional basin margins. J. Geol. Soc.
Lond. 141, 609e620.
Gupta, S., Cowie, P.A., Dawers, N.H., Underhill, J.R., 1998. A mechanism to explain
rift-basin subsidence and stratigraphic patterns through fault array evolution.
Geology 26, 595e598.
Hardman, R.P.F., Booth, J.E., 1991. The significance of normal faults in the exploration
and production of North Sea hydrocarbons. In: Roberts, A.M., Yielding, G.,
Freeman, B. (Eds.), The Geometry of Normal Faults, Geological Society Special
Publication, vol. 56, pp. 1e13.
Higgins, S., Davies, R.J., Clarke, B., 2007. Antithetic fault linkages in deep water fold
and thrust belt. J. Struct. Geol. 29, 1900e1914.
Hodgetts, D., Imber, J., Childs, C., Flint, S., Howell, J., Kavanagh, J., Nell, P., Walsh, J.,
265
2001. Sequence stratigraphic responses to shoreline-perpendicular growth
faulting in shallow marine reservoirs of the champion field, offshore Brunei
Darussalam, South China Sea. AAPG Bull. 85, 433e457.
Hooper, R.J., Fitzsimmons, R.J., Grant, N., Vendeville, B.C., 2002. The role of defor-
mation in controlling depositional patterns in the south central Niger Delta,
West Africa. J. Struct. Geol. 24, 847e859.
Hus, R., Acocella, V., Funiciello, R., De Batist, M., 2005. Sandbox models of relay ramp
structure and evolution. J. Struct. Geol. 27, 459e473.
Hus, R., De Batist, M., Klerkx, J., Matton, C., 2006. Fault linkage in continental rifts:
structure and evolution of a large relay ramp in Zavarotny; Lake Baikal (Russia).
J. Struct. Geol. 28, 1338e1351.
Imber, J., Childs, C., Nell, P.A.R., Walsh, J.J., Hodgetts, D., Flint, S., 2003. Hanging wall
fault kinematics and footwall collapse in listric growth fault systems. J. Struct.
Geol. 25, 197e208.
Mansfield, C.S., Cartwright, J.A., 1996. High resolution fault displacement mapping
from three-dimensional seismic data; evidence for dip linkage during fault
growth. J. Struct. Geol. 18, 249e263.
McClay, K.R., 1990. Extensional fault systems in sedimentary basins: a review of
analogue model studies. Mar. Pet. Geol. 7 (3), 206e233.
Morley, C.K., Gabdi, S., Seusutthiya, K., 2007. Fault superimposition and linkage
resulting from stress changes during rifting: examples from 3D seismic data,
Phitsanulok Basin, Thailand. J. Struct. Geol. 29, 646e663.
Nicol, A., Watterson, J., Walsh, J.J., Childs, C., 1996a. The shapes, major axis orien-
tations and displacement patterns of fault surfaces. J. Struct. Geol. 18, 235e248.
Nicol, A., Walsh, J.J., Watterson, J., Gillespie, P.A., 1996b. Fault size distributionsdare
they really power e law? J. Struct. Geol. 18, 191e197.
Peacock, D.C.P., Sanderson, D.J., 1991. Displacement, segment linkage and relay
ramps in normal fault zones. J. Struct. Geol. 13, 721e733.
Pivnik, D.A., Ramzy, M., Steer, B.L., Thorseth, J., El Sisi, Z., Gaafar, I., Garing, J.D.,
Tucker, R.S., 2003. Episodic growth of normal faults as recorded by syntectonic
sediments, July oil field, Suez Rift, Egypt. AAPG Bull. 87, 1015e1030.
Pochat, S., Castelltort, S., Van Den Driessche, J., Besnard, K., Gumiaux, C., 2004.
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. AAPG Bull. 88, 1357e1367.
Robert, J.T., Moores, E.M., 1992. In: W.H. Freeman and Company (Ed.), Structural
Geology, p. 532.
Rowan, M.G., Hart, B.S., Nelson, S., Flemings, P.B., Trudgill, B.D., 1998. Three
dimensional geometry and evolution of a salt related growth fault array:
Eugene Island 330 field, offshore Louisiana, Gulf of Mexico. Mar. Pet. Geol. 15,
309e328.
Trudgill, B.D., Cartwright, J.A., 1994. Relay ramp forms and normal fault linkages:
Canyonlands National Park, Utah. Geol. Soc. Am. Bull. 106, 1143e1157.
Walsh, J.J., Bailey, W.R., Childs, C., Nicol, A., Bonson, C.G., 2003. Formation of
segmented normal faults: a 3-D perspective. J. Struct. Geol. 25, 1251e1262.
Watterson, J., 1986. Fault dimensions, displacements and growth. Pure Appl. Geo-
phys. 124, 365e373.
Weber, K.J., Daukoru, E.M., 1975. Petroleum geology of the Niger delta. Proc. Ninth
World Pet. Congr. 2, 209e221.