You are on page 1of 9

A f r i c a

640 The Leading Edge June 2011


S P E C I A L S E C T I O N : A f r i c a
6HLVPLFLQWHUSUHWDWLRQVWUXFWXUDODQDO\VLVDQGIUDFWDOVWXG\RIWKH
*UHDWHU8JKHOOL'HSREHOW1LJHU'HOWD%DVLQ1LJHULD
A
major feature of the habitat of petroleum in the Niger
Delta Basin is the association of petroleum traps with
growth faults. Because of the signifcant role of these faults in
hydrocarbon accumulation and redistribution in the basin, a
good understanding of the timing of fault motion has now
been shown to be vital for successful exploration of fault-
bounded prospects. In most petroleum habitats, structural
elements such as fault patterns, their kinematics, geometry,
timing, and size of the structures control the distribution
of hydrocarbons in adjacent fault blocks. Te success or
otherwise of an exploration well in such areas depends on
the location of such a well relative to the structural closure
interpreted from the seismic data. Experience has shown that
detailed structural analysis of prospective felds can provide
a reliable kinematic and growth history upon which risks
associated with fault movement, trap integrity and structure,
geometry/size modifcation can be evaluated before deciding
on the drilling location.
Te objectives of this paper were to evaluate the structural
elements that controlled hydrocarbon distribution in three
felds or closures and provide explanations for drilling results
obtained. Tere was the need to provide adequate explana-
tion for the lack of hydrocarbons at predicted levels in an
exploration well and to study the implications of these for
nearby analog felds. A proper understanding of the process
of fault growth by propagation and coalescence, sealing po-
tentials of faults, and, ultimately, improved understanding of
the fault pattern in any feld will lead to a more efective feld
development planning. Te felds studied are situated along
adjacent fault blocks in a part of the Greater Ughelli Depo-
belt of the Niger Delta Basin. Tis area (Figure 1) falls within
the extensional zone of the Niger Delta Basin (Doust and
Omatsola, 1989; Damuth, 1994; Pochat et al., 2004). In an
extensional deltaic setting, the interaction between
sedimentary processes and fault slip is documented
by the growth of faults. Terefore, to understand
the structural control on hydrocarbon distribution
in adjacent fault blocks, it is necessary to carry out
an analysis of the fault kinematics to decipher the
tectonic and stratigraphic history recorded by the
growth faults. Statistical and geometric analyses of
faults from reservoir and top seal intervals of the
various felds help us understand their fractal na-
ture (scaling properties) and to simulate subseismic
faults in the area. Knowledge of the number, size,
and distribution of small or subseismic faults leads
to inferences about their infuence on fuid fow
and top seal leak probability. Tis kind of analysis
is now yielding good results that lead to better feld
development planning and top seal leak assessment
D. K. AMOGU, J. FILBRANDT, K. O. LADIPO, and C. ANOWAI, Shell Petroleum Development Company Ltd.
K. ONUOHA, University of Nigeria
(e.g., Edwards, 1995; Mansfeld and Cartwright, 1996; Cart-
wright et al., 1998; Castelltort et al., 2004).
Structural style and play concept in the study area
Te study area is subdivided into three main fault blocks
by WNW-ESE-trending, basin-ward dipping, and arcu-
ate growth faults (Figure 1). Te three felds or closures (A,
B, and C) under study lie within the same structural trend.
Field A is producing while closure B contains a series of ma-
jor prospects. Closure C is an unappraised discovery (UAD).
A and B are bounded by the same fault (fault 1consisting
of three segments) while closure C is defned by fault 4 which
is antithetic to a main fault (fault 2). Te stratigraphic level
of the objective intervals (F1000K2000) is between 31.3
and 33.3 Ma. Te estimates of the sealing capacity along the
faults at these objective levels in Field A and closure B are
similar. In addition, the hydrocarbon system analyses sug-
gest similar source rock and charge timing as well as struc-
tural confgurations at these levels.
Te predrill prognosis of lithology and depth based on
well 1 in Field A, was confrmed by the postdrill result of
closure B and established the basis and confdence in the cor-
relation (Figure 2). However, while all objectives are saturated
with hydrocarbon in Field A (Figure 2a), they are dry at clo-
sure B (about 4 km from Field A). Fluid inclusion screening,
however, indicated the presence of hydrocarbons at all objec-
tive intervals at closure B. At closure C, about 6 km east of
B, hydrocarbons were encountered only at shallower levels
(C7000 and C8100) than the objective levels.
Two sets of poststack time-migrated refection data cov-
ering about 250 km
2
were used to interpret 170 km
2
over
the study area. A semblance volume was derived from the
refection volume and horizontal semblance used to highlight
Figure 1. Semblance time slice of the study area, showing Field A and closures B
and C. Lines 1-7 are sections used for kinematics analysis. Fault picks correlate
well with fault traces. Te regional geology map is modifed from Onuoha (1999).
Downloaded 16 Nov 2011 to 76.120.23.222. Redistribution subject to SEG license or copyright; see Terms of Use at http://segdl.org/
June 2011 The Leading Edge 641
A f r i c a
developments at the felds. Te adopted workfow included
seismic interpretation, fault pattern analysis, growth estima-
tion and fault kinematics, sequential restoration, and decom-
paction of sections.
Seismic interpretation
Te frst phase of the interpretation involved fault interpreta-
tion. Faults were picked on vertical sections along the direc-
tion of maximum transport and correlated along the fault
trace on semblance slices (Figure 1). Te faults sticks were
correlated with multiple semblance slices at an interval of 0.5
s to constrain fault shape and validate vertical correlation of
faults in a 3D visualization tool.
Subsequently, the fault sticks were modeled to produce
the corresponding fault surfaces. Horizon interpretation con-
stituted the second phase of the interpretation. Te mapped
events were identifed from well correlation and tying well
lateral amplitude variations caused by faults and stratigraphic
discontinuities (Figure 1). Available well tops and logs were
useful in identifying the horizons of interest. Te vertical re-
fection sections (Figure 3) refect stratigraphic and tectonic
Figure 2. (a) Logs of well 1 in Field A, showing some of the hydrocarbon-saturated intervals. (b) Stratigraphic correlation across Field A
with closures B and C.
Figure 3. (a)-(c) Seismic refection sections showing stratigraphic
thicknesses at the closures. (d) Semblance slice.
Figure 4. (a) Horizon interpretation in 3D. (b) Fault-horizon Earth
model. Horizons displayed from Shallow_3 to level K4000 (6920
11,990 ft).
Downloaded 16 Nov 2011 to 76.120.23.222. Redistribution subject to SEG license or copyright; see Terms of Use at http://segdl.org/
642 The Leading Edge June 2011
A f r i c a
tops to seismic events. A 3D visualization method was
preferred for horizon interpretation because of the large data
volume available (i.e., inline, crossline, and time slice used
simultaneously, Figure 4a). Horizons were picked on later-
ally continuous refections, starting from the well locations,
and autotracked at regular pick intervals determined by data
quality. Te fault-horizon Earth model (Figure 4b) resulting
from the seismic interpretation provided the framework for
the structural analysis of the felds using the techniques men-
tioned above.
Fault pattern analysis
Te fault geometries encountered in the area included simple
listric faults, antithetic faults, buried counter-regional faults,
and crossing conjugate faults. Te propagation and the inter-
actions of these faults produced diferent structural features
that impacted the horizons diferently. Te three main faults
(faults 1, 2, and 4) that bound the closures interact within
the vicinity of closure B. Fault 2 terminated behind the clo-
sure while the antithetic fault (fault 4) intersected fault 1 at
the F1000 level (Figure 5). Faults 5 and 1 (Figure 5) formed
a strike or relay ramp close to closure B.
Relay ramps form along arrays of normal faults that con-
tain en-echelon ofsets, or step-overs, (Morley et al., 1990;
Peacock and Sanderson, 1991, 1994). Tey are common in
extensional data sets because they are ubiquitous features in
extensional terrains (Tearpock et al., 1994). On 3D data sets,
the data may become semicoherent, or deteriorate where two
faults overlaps, perhaps because of the fault shadow efect.
If this occurs, then structural horizon mapping may result
in the mapping of two fault surfaces as a single fault surface
(Tearpock and Bischke, 2003). Also, on 2D and 3D seismic
data sets, structural aliasing results in mapping the two over-
lapping faults as one fault. Relay ramps have a limited lateral
zone, therefore, even on good data, en-echelon faults may
be overlooked if the interpretation is based on insufciently
close seismic lines. Te pitfall is that these ramps may or may
not contain a fault that would form a three-way fault closure
(Brenneke, 1995).
Figure 5 shows the F1000 horizon and the throw profles
of fault 1 at F1000. An abrupt drop along a fault throw profle
is typical of a zone of fault coalescence (Filbrandt et al. 2007).
Figure 5. (a) F1000 level showing extracted fault mid line. (b)
Trow profle of fault 1. (c) Semblance slice at F1000 level. Notice
the drop in throw at zone (X) of fault coalescence. (d) and (e) Trow
profles of faults 2 and 3, respectively. Both illustrate the theory of
throw conservation at fault intersection. Fault 2 dropped from about
720 ft to about 320 ft while its branch (fault 3) propagated away
from the intersection with a throw of about 400 ft. (f ) Trow profle
of fault 4.
Figure 6. (a) Te use of the expansion index (E.I.) to express growth-
fault behavior. Section of growth fault with activity in units A and B
expressed in E.I. values >1.0, where E.I. equals ratio of stratigraphic
thickness in hanging wall to footwall (A/A, B/B). Fault activity ceases
at end of unit B, leaving unit C with E.I. of 1.0. (b) Procedure for
generation of T-Z plot. Sections of two growth faults with contrasting
values of E.I. for identical slip rates, U, due to contrasting background
sediment accumulation rates (S1 < S2).
Downloaded 16 Nov 2011 to 76.120.23.222. Redistribution subject to SEG license or copyright; see Terms of Use at http://segdl.org/
June 2011 The Leading Edge 643
A f r i c a
Ad Space
Downloaded 16 Nov 2011 to 76.120.23.222. Redistribution subject to SEG license or copyright; see Terms of Use at http://segdl.org/
644 The Leading Edge June 2011
A f r i c a
Where such drops exist without obvious fault coalescence, it
implies that a fault has been missed out or miscorrelated in
the interpretation. A semblance slice (Figure 5c) taken at the
hanging-wall time equivalent (1984 ms) of F1000 shows
point X as a zone of fault coalescence. At fault intersections,
fault throw must be conserved. If the sum total of the throw
of the faults that emerged from an intersection is not equal
to the throw of the faults that entered the intersection, it im-
plies that a fault has been missed out or miscorrelated. Faults 2
and 3 (Figures 5d and 5e) clearly illustrate the theory of throw
conservation at a fault intersection. Trow in Fault 2 dropped
from about 720 ft to about 320 ft while its branch (fault 3)
propagated away from the intersection with a throw of about
400 ft. Fault 4 (Figure 5f ) has low throw values (5080 ft)
around well 1 in Field 3. Tis implies a high footwall risk for
thick reservoirs (100200 ft) and thin top seals.
Growth and kinematics analysis
To understand the interplay between sedimentation and
fault movement, three methods of kinematic analysis of
growth faults have been proposed (Torsen, 1963; Tearpock
and Bischke, 1991; Bischke, 1994). Te growth or expan-
sion index method (Torsen) is useful in determining peri-
ods of most signifcant growth (Edwards, 1995). However,
this method does not contain any information about abso-
lute displacement since it is simply a ratio (Cartwright et al.,
1998). Hence, local sedimentation rates can produce diferent
growth index values for faults with the same slip magnitude.
For example, consider two faults that accumulate an equal
Figure 7. T-Z plots of seven sections across fault 1. (a) and (b) T-Z trend at Field A and closure B. (c) A transition pattern between Field A
and closure B. (d) Fault grid showing the sections, well tops, and hydrocarbon contacts.
Figure 8. Iscochore maps showing stages of fault propagation and
subregional trapping evolution. Te numbers in the fault polygons
indicate the thickness variation across the faults. Tey suggest periods of
fault activity.
Downloaded 16 Nov 2011 to 76.120.23.222. Redistribution subject to SEG license or copyright; see Terms of Use at http://segdl.org/
June 2011 The Leading Edge 645
A f r i c a
slip in a given time interval (Figure 6, after Cartwright et
al.). Tese two faults are growing in settings with contrasting
sediment accumulation rates (S1 < S2) and are represented by
contrasting expansion indices for the interval, despite their
equal slip rates and amounts.
Te vertical separation versus depth plot (cf., throw/depth
plot; Cartwright et al.; Castelltort et al., 2004; Back, et al.,
2006; Tearpock and Bischke, 1991, 2003) is a high-resolu-
tion method that can detect growth where growth is not obvi-
ous in the data. Te last method is the multiple Bischke plot
analysis (Bischke, 1994; Sanchez et al., 1997). Tis method
helps in identifying stratigraphic and/or tectonic correlation
problems in wells.
In this study, we have adopted the throw-depth (T-
Z) method for the study of fault kinematics. Te measured
throw between hanging wall and corresponding footwall ho-
rizon pairs is plotted against the hanging-wall depth value
of each horizon to produce the T-Z plot. For the horizon
pairs shown in Figure 6b, the fault throw T, with respect
to the B horizon is given by Bd-Bu. A plot of T against
Bd gives a T-Z plot. Plots that have zero slope phase (i.e.,
constant throw) are interpreted as nongrowth or postdeposi-
tional propagation, suggesting that the fault was probably not
emergent or buried at the depositional surface. On the plot,
a sloping phase is indicative of the growth-generating fault-
ing phase (Cartwright et al.; Back et al.). Te incremental
growth or throw of any interval is equivalent to the slope of
that interval.
Figure 3 shows some interpreted horizons correlated
across fault 1. A total of 11 horizon pairs were used for the
T-Z plots in this study. Te T-Z plots (Figure 7) show the
displacement pattern along seven sections across fault 1. Te
number of growth and nongrowth intervals varies along the
strike of the fault. At Field A, the fault has one nongrowth
and two growth phases. At closure B, the fault has three non-
growth and four growth-generating phases. Te plot of the
section between Field A and closure B shows a transitional
pattern between the trends at Field A and closure B.
A similar plot analysis at closure C shows two nongrowth
and three growth phases. Closures B and C show a polycyclic
pattern because they have at least two nongrowth phases. Tis
implies that fault 1 is stable at Field A compared to closure B.
Tis localized contrasting kinematic history along the strike
of fault 1 and the adjacent fault 2 refects the susceptibility of
individual faults to failure under similar vertical stress (i.e.,
sediment load). Te variation in strength of rock materials or
fault integrity observed along fault 1 is probably infuenced
by the interaction of the fault with other faults. Polycyclic
growth behavior has been observed for large growth faults
over long time periods in the Gulf Coast (Edwards, 1995).
In deltaic settings, growth faults are gravitationally driven
tectonic systems. As such, a correlation is expected between
sediment loading and fault activity. Cartwright et al. corre-
lated three phases of growth observed from 17 faults in the
Wadata Fault Zone to three main regressive cycles over ap-
proximately 130 k.y. (130,000 years) during the regressive
phase (i.e., shoreline regression and transgression by sea-level
rise) growth faulting stops. Tis similarity supports the long-
term coupling between sediment loading and fault activity.
At the study area, well results from Field A show that hy-
drocarbons accumulated essentially at the nongrowth phase
(levels F1000K2000). Tis can be correlated to the stabil-
ity of fault 1 and the consequent preservation of fault seal at
Field A. Tis is shown by the hydrocarbon contacts displayed
on the fault surfaces in Figure 7.
Subregional trap history
Subregional trap evolution and fault propagation history of the
study area was evaluated by analyzing isochore maps between
diferent intervals. Te map of K2000H5000 (33.431.3) in
Figure 8 shows that fault 1 had developed at Field A but not at
closure B. Also fault 2 had developed at closure C within this
Figure 9. Oil generation and expulsion curves. (a) Generation curve and (b) expulsion curve.
Downloaded 16 Nov 2011 to 76.120.23.222. Redistribution subject to SEG license or copyright; see Terms of Use at http://segdl.org/
646 The Leading Edge June 2011
A f r i c a
period and fault closures also developed at Field A and against
the main fault at closure C. Figure 8 shows that fault closure ex-
ists at all the objective levels in Field A. At closure B, fault closure
exists only at the shallow objective level. At a deeper level (K),
the fault closure is weak. Closure C has a more robust closure
against fault 2 than fault 4 where well 1 was drilled. Te result
of 1D basin modeling (Figure 9) shows that hydrocarbon gen-
eration commenced at about 27.5 Ma while its expulsion started
at about 19.0 Ma. Both generation and expulsion seem to have
continued until the present. Tis implies that any existing trap
ought to have been flled with hydrocarbons except where such
a trap has been breached or spilled by subsequent fault move-
ment or where no fault seal exists. Te timing of source rock,
reservoirs and faulting/traps are summarized in the petroleum
systems analysis chart (Figure 10).
In 2D view (Figure 11), structural restoration of the
sections through key wells in the felds shows that, at about
the time of inception of hydrocarbon expulsion (19.0 Ma),
closure exists against all the main bounding faults. It also
suggests that the risk of sand-sand juxtaposition is high
at the well 1 location at closure C. Tis is because the an-
tithetic fault has small throw (5080 ft) at the well com-
pared to the thickness of the sands (75280 ft) and shales
encountered by the well.
Conclusions
Traps formed earlier at all objective levels at Field A com-
pared to closures B and C because faulting and closure devel-
opment started and ceased earlier (at about 18 Ma) around
the vicinity of Field A. Relative stability of the fault at Field A
possibly enhanced the retention of trapped hydrocarbons at
Field A compared to closures B and C where faulting ceased
Figure 10. Petroleum systems analysis chart of the Greater Ughelli Depobelt, Niger Delta Basin.
Downloaded 16 Nov 2011 to 76.120.23.222. Redistribution subject to SEG license or copyright; see Terms of Use at http://segdl.org/
June 2011 The Leading Edge 647
A f r i c a
at about 14 Ma. Tis implies that structures in Field A had
been stable and, consequently, had been trapping the ex-
pelled hydrocarbons since 18 Ma to present.
Because faulting and closure development ended at about
14 Ma at closures B and C, it is expected that any trap that
is in place should have been flled with hydrocarbons be-
cause hydrocarbon expulsion in the study area is predicted
to have continued till the present day. However, well results
at the two closures proved otherwise because the wells did
not encounter hydrocarbon accumulation at the objective
levels. Te chances of hydrocarbon accumulation at closure
B was diminished because later faulting caused the closure to
change shape and migrated from west to east. Te relief of the
Figure 11. Sequential decompaction and restoration of sections at the
approximate time of hydrocarbon expulsion to assess structure geometry
and footwall risk. (a) Field A. (b) Closure B. (c) Closure C. Note
possible sand-sand juxtaposition at well 1 in closure C.
Figure 12. Earth model showing strike ramp near closure B.
blank for ad
Downloaded 16 Nov 2011 to 76.120.23.222. Redistribution subject to SEG license or copyright; see Terms of Use at http://segdl.org/
648 The Leading Edge June 2011
A f r i c a
closure at B appears to have been relatively low initially and
was further reduced by fault activity. Today the trap is subtle
and may have only up to 50 ft of closure. Small changes in
interpretation and depth conversion may easily remove the
fault and dip closure at all levels. Also, the zone of coalescence
of faults F1, F2, and F4 behind closure B (Figure 12) may
have created a damaged zone with lower sealing capacity and,
as such, fault seal at closure B may have been breached.
Well 1 in closure B was dry because of the absence of a trap
at the deeper objectives while the risk of fault seal failure is also
high at and around level F1000 because of the small throw and
possibly low shale gouge ratios (SGR). In addition, late fault
movement may have modifed and breached the paleo-closures
(traps) and, as a result, trapped spilled hydrocarbons, leaving a
paleo-amplitude anomaly at closure B. We recommend detailed
fault mapping, analysis of fault geometry, growth patterns, and
subregional trap evaluation as key components of an evaluation
workfow to assess the structural risks associated with prospects
as they advance in the maturation funnel.
References
Back, S., C. Hocker, M. B. Brundiers, and P. A. Kukla, 2006, Tree-
dimensional-seismic coherency signature of Niger Delta growth
faults: integrating sedimentology and tectonics: Basin Research,
18, no. 3, 323337, doi:10.1111/j.1365-2117.2006.00299.x.
Bischke, R. E., 1994, Interpreting sedimentary growth structures
from well log and seismic data (with examples): AAPG Bulletin,
78, 873892.
Brenneke, J. C., 1995, Analysis of fault traps: World Oil, December,
6371.
Cartwright, J. A., R. Bouroullec, D. James, and H. D. Johnson,
1998, Polycyclic motion history of some Gulf Coast growth faults
from high-resolution displacement analysis: Geology, 26, no. 9,
819822, doi:10.1130/0091-7613(1998)026<0819:PMHOSG>2.3
.CO;2.
Castelltort, S., S. Pochat, and J. Van Den Driessche, 2004, Using T-Z
plots as a graphical method to infer lithological variations from
growth strata: Journal of Structural Geology, 26, no. 8, 1425
1432, doi:10.1016/j.jsg.2004.01.002.
Damuth, J. E., 1994, Neogene gravity tectonics and depositional
processes on the deep Niger Delta continental margin: Marine
and Petroleum Geology, 11, no. 3, 320346, doi:10.1016/0264-
8172(94)90053-1.
Doust, H. and E. Omatsola, 1989, Niger Delta: AAPG Memoir, 48,
201238.
Edwards, M. B., 1995, Diferential subsid-
ence and preservation potential of shal-
low-water Tertiary sequences, northern
Gulf Coast Basin, USA, in A.G. Plinth,
ed., Sedimentary facies analysis: A tribute
to the research and teaching of Harold.
G. Reading: International Association of
Sedimentologists, Special Publication 22,
265281.
Filbrandt, J. B., P. D. Richard, and R. Frans-
sen, 2007, Fault growth and coalescence:
insights from numerical modelling and
sandbox experiments: GeoArabia, 12,
1732.
Mansfeld, C. S. and J. A. Cartwright, 1996,
High-resolution fault displacement mapping from three-dimen-
sional seismic data: evidence for dip linkage during fault growth:
Journal of Structural Geology, 18, no. 23, 249263, doi:10.1016/
S0191-8141(96)80048-4.
Morley, C. K., R. A. Nelson, T. L. Patton, and S. G. Munn, 1990,
Transfer zones in the East African rift system and their relevance to
hydrocarbon exploration in rifts: AAPG Bulletin, 74, 12341253.
Onuoha, K. M., 1999, Structural features of Nigerias coastal margin:
an assessment based on age data from wells: Journal of African Earth
Sciences, 29, no. 3, 485499, doi:10.1016/S0899-5362(99)00111-6.
Peacock, D. C. P. and D. J. Sanderson, 1991, Displacement, segment
linkage, and relay ramps in normal fault zone: Journal of Structural
Geology, 13, no. 6, 721733, doi:10.1016/0191-8141(91)90033-F.
Peacock, D. C. P. and D. J. Sanderson, 1994, Geometry and develop-
ment of relay ramps in normal fault systems: AAPG Bulletin, 78,
147165.
Pochat, S., S. Castelltort, J. Van Den Driessche, K. Besnard, and C.
Gumiaux, 2004, A simple method of determining sand/shale ra-
tios from seismic analysis of growth faults: An example from upper
Oligocene to lower Miocene Niger Delta deposits: AAPG Bulletin,
88, no. 10, 13571367, doi:10.1306/04290403117.
Sanchez, R., J.-Y. Chatellier, R. de Sifontes, N. Parra, and P. Munoz,
1997, Multiple Bischke plots analysis, a powerful method to distin-
guish between tectonic or sedimentary complexity and miscorrela-
tions; methodology and examples from Venezuelan oil felds: Me-
morias del Primero Congreso Latinoamericano de Sedimentologia,
Soc. Venezolana de Geologo, Tomo II, Noviembre 1997, 257264.
Tearpock, D. J. and R. E. Bischke, 1991, Applied subsurface geologi-
cal mapping: Prentice-Hall.
Tearpock, D. J. and R. E. Bischke, 2003, Applied subsurface geologi-
cal Mapping second edition: Prentice-Hall, New Jersey.
Tearpock, D. J., R. E. Bischke, and J. L. Brewton, 1994, Quick look
techniques for prospect evaluations: Lafayette, Louisiana Subsur-
face Consultants and Associates.
Torsen, C. E., 1963, Age of growth faulting in Southeast Louisiana:
Transactions - Gulf Coast Association of Geological Societies, 13,
103110.
Acknowledgments: Te authors are grateful to Shell Petroleum
Development Company of Nigeria (SPDC) for providing the data
and other facilities used for this study. We are grateful to Chike
Onyejekwe, exploration manager at the time this work was done,
for approving the study and earlier publication of some results.
Corresponding author: mosto.onuoha@gmail.com
blank for ad
Downloaded 16 Nov 2011 to 76.120.23.222. Redistribution subject to SEG license or copyright; see Terms of Use at http://segdl.org/

You might also like