Sunteți pe pagina 1din 13

AAPG Bul l et in , v. 85, n o . 4 ( Apr il 2001) , pp.

587599 5 8 7
Displacement distributions
on extensional faults:
Implicationsfor fault stretch,
linkage, and seal
S . A . S te w a rt
A B S T R A C T
Most extensional faults are characterized by displacement distri-
butions that range from zero at the fault tips to a maximum value
at some point along the length of the fault. Regardless of the pattern
of displacement distribution, a geometrical requirement resulting
from the displacement variation along the fault is that beds must
stretch parallel with the strike of the fault. It has been suggested
that minor faults and fractures evolve perpendicular to the main
fault to accommodate this stretch. The amount of stretch that this
subpopulation of faults must accommodate is quantied here using
several alternative displacement prole models. The choice of pro-
le model is much less signicant than the ratio of maximum
displacement to fault length (D
max
/L). The amount of stretch rap-
idly increases from about 0.7% at D
max
/L of 0.05 to about 3% at
the typical upper limit for D
max
/L of 0.1. These relationships pre-
dict tens to hundreds of meters of fault-parallel stretch associated
with kilometer-scale extensional faults. The size and spatial distri-
bution of stretch accommodation faults should be incorporated in
three-dimensional (3-D) fault zone models, and there are several
implications for fault linkage and seal. One possible strain accom-
modation scenario is that a small number of faults that have dis-
placements signicant enough to generate reservoir-juxtaposition
leak points could exist. Alternatively, a subseismic fault and fracture
population could either signicantly degrade horizontal permeabil-
ity parallel with the major fault or, if open, constitute a target
for high-angle exploitation drilling. The intersections between the
stretch accommodation faults and the main fault could provide con-
duits for fault valving or low-entry pressure leak points. The most
commonly observed candidate stretch accommodation faults occur
in fault overlap zones in segmented fault systems, coincident with
maximum displacement gradients on the fault planes. In these
cases, cutoff stretch accommodation could be an integral factor in
fault segment linkage.
Copyri ght 2001. The Am eri can Associ ati on of Petrol eum Geol ogi sts. Al l ri ghts reserved.
M anuscri pt recei ved Apri l 19, 1999; revi sed m anuscri pt recei ved M ay 8, 2000; nal acceptance June 30,
2000.
A U T H O R
S. A. St ewar t UpstreamTechnology
Group, BP, BurnsideRoad, FarburnIndustrial
Estate, Dyce, AberdeenAB217PB, United
Kingdom; stewartsa1@bp.com
Simon Stewart completed a Ph.D. at Imperial
College, London, on the structural geology
and paleomagnetismof the Carboniferous
thrust belt in north Spain. Since then he has
spent eight years at Amerada Hess on various
North Sea teams, workingmainly on three-
dimensional (3-D) seismic interpretation. In
2000 he joined BP as a structural geologist,
workingon the application of integrated
drilling, geomechanical, and seismic data. His
special interests includethe relationship
between mechanostratigraphyand structural
style and development of 3-Dseismic
interpretation work ows for fast, accurate
mapping.
A C K N O W L E D G E M E N T S
This article beneted fromreviews by R.
Schlischeand G. Yieldingand comments from
J. Cartwright and J. Underhill. I thank T. Wynn
for assistingwith programming. The mapping
shown in Figure 6a is by J. Argent. I thank
Veritas DGCfor permission to showseismic
data. The views expressed here are solely my
own and do not necessarilyrepresent those of
Amerada Hess Ltd. or BP.
5 8 8 Di spl acem ent Di stri buti ons and Stretch of Extensi onal Faul ts
fault tip
cutoff stretch accommodation
faults
m
a
in
fa
u
lt
p
la
n
e
F i g u re 2 . Three-dimensional sketchdiagramshowingthedis-
placement on a major fault decreasing to zero at the fault tip.
Cutoff stretch accommodation faults that accommodate LL
areshown (Destro, 1995). Variationindisplacement vector (ar-
rows on fault plane), an additional possible effect of cutoff
stretch, is also shown (Roberts, 1996).
L
fault tip fault tip
D
max
L
F i g u re 1 . Fault planeseparationsketch(Allandiagram) show-
ingnal proleof ahanging-wall beddingcutoff alongthelength
of thefault (L) compared with theunfaulted bedlengthL. The
stretch of the beds duringfaultingis LL; D
max
is referenced
to a horizontal linein this sketchthis assumes that thereis no
footwall uplift of the correspondingbed.
I N T R O D U C T I O N
It is well known that in the case of idealized, spatially
isolated extensional faults, the displacement of strata
across the fault plane ranges from zero at the fault tip
to a maximum value (D
max
) near the middle of the
map-view length (L) of the fault (e.g., Walsh and Wat-
terson, 1987; Dawers et al., 1993; Schlische, 1995; Ni-
col et al., 1996a). Regardless of the manner in which
the fault has grown, a simple diagram of the separation
of corresponding hanging-wall and footwall beds pro-
jected onto the plane of the fault shows that the beds
must stretch in the plane parallel with the fault, be-
cause the beds have a nal length L in this line of
section compared with their initial length L (Figure 1).
Stretch of hanging wall and footwall has also been pre-
dicted in three-dimensional (3-D) numerical models of
extensional faults (Ma and Kusznir, 1993, gure 5).
The intersection of bedding with a fault plane is
commonly referred to as a structural cutoff line, and I
employ the nomenclature cutoff stretch here to refer
to stretch of bedding immediately adjacent to the fault
plane. The only scenario in which this stretch would
not occur is if shortening occurred parallel with the
extensional faults during rifting. In this article I assume
that such shortening does not generally occur and that
mapped fault tips can be regarded as pins that can be
connected by a straight line to dene the unfaulted bed
length L. I assume this because most fault systems are
spatially isolated, linear trends and if sections are taken
parallel with these trends but sufciently offset on,
say, the footwall side to be removed from any
displacement-gradient related folds, there does not
tend to be any evidence for regional shortening (e.g.,
Schlische, 1995).
Various phenomena have been attributed to the
accommodation of cutoff stretch during extensional
fault evolution (Figure 2). Destro (1995) suggested
that a subpopulation of extensional faults trending at
high angles to and terminating at the intersections with
the master fault accommodate cutoff stretch. Destro
(1995) presented eld and seismic examples of these
faults and termed them release faults. In this article
I use the more self-explanatory, if unwieldy, term cut-
off stretch accommodation faults. Roberts (1996)
mapped systematic variations in displacement direc-
tion along the length of extensional fault planes in the
eld and noted that the variations could be accounted
for by differential stretch of the hanging-wall beds rela-
tive to the footwall. Roberts (1996) also noted the
presence of cutoff stretch accommodation faults in
these eld examples. Schlische (1995) illustrated a
seismic-based example of transverse faults accommo-
dating bed-length extension during faulting (gure 12
in Schlische [ 1995] ). In this article I argue that the
numerous published examples of transfer faults that
breach relay ramps to link overlapping fault segments
(e.g., Peacock and Sanderson, 1994; Cartwright et al.,
1996; Ferrill et al., 1999; Walsh et al., 1999) represent
the most common occurrence of cutoff stretch accom-
modation faults.
The purpose of this article is to quantify cutoff
stretch in relation to the ratio of maximum fault dis-
placement to fault length (D
max
/L) and to discuss the
effects of cutoff stretch accommodation mechanisms
on fault linkage and seal. Many of the extensional fault
systems in prospective basins are more complex than
the idealized faults analyzed here, in terms of fault sys-
tem segmentation and fault intersections; however,
such systems are also characterized by displacement
St ewar t 5 8 9
variations, and the issues discussed here are generally
relevant.
D I S P L A C E M E N T D I S T R I B U T I O N M O D E L S
The idealized faults considered here are isolated from
interaction with other faults or heterogeneities that
might affect the shape and displacement prole. Such
faults were termed unrestricted by Nicol et al.
(1996a). Restricted faults, however, become geomet-
rically modied from their idealized form by interac-
tion with other faults, the free surface, or mechanical
layering of the stratigraphy. The latter effect in partic-
ular causes layer-parallel elongation of the fault planes
and may consequently affect D
max
/L ratios (Nicol et
al., 1996a). The discussions presented in followingsec-
tions implicitly consider unrestricted faults, but the re-
sults and discussion are applicable to any fault where
D
max
/L is known and the ends of the faults are tips
rather than branches that have other faults. The fol-
lowing calculations of cutoff stretch are conned to the
end product of fault evolution as observed today in the
eld or on seismic and characterized by the D
max
/L
ratio. The alternative mechanisms for fault growth
(e.g., lengthening of L through time vs. constant L and
increasing D
max
/L) are not considered in this section;
however, the impact of cutoff stretch on the related
issue of fault linkage is discussed in a following section.
Various best-t displacement proles have been
advocated for extensional faults. Reviews to date agree
that a linear increase in displacement from the fault
tips to D
max
ts most eld data (e.g., Dawers et al.,
1993; Clark and Cox, 1996; Peacock and Sanderson,
1996; Cowie and Shipton, 1998), although there is
considerable scatter in most of the available data and
other types of prole could be tted in many cases
(e.g., gure 13 of Nicol et al. [ 1996a] ). A 3-D seismic
example of a well-imaged array of restricted exten-
sional faults illustrates the difculty in selecting a sin-
gle, characteristic displacement prole (Figure 3). The
faults in this example are restricted by interaction with
one another and constraining interfaces below (a salt
detachment) and above (the free surface). Rather than
enter into the debate as to which, if any, prole is the
most appropriate, I consider here a range of idealized
proles representing most published models. Cutoff
stretch is calculated for each, enabling denition of a
range of stretch for a given D
max
/L ratio that encom-
passes the range of published displacement proles.
The alternative displacement proles are introduced in
the following sections with some brief comments and
are illustrated in Figure 4. In each case it is assumed
that D
max
occurs in the middle of the fault.
Li near
The linear type of prole (Figure 4a) was noted in some
of the earliest eld measurements of displacement pro-
le (C-type prole of Muraoka and Kamata [ 1983] )
and is currently popular as the best t to most mea-
sured proles (e.g., Cowie and Shipton, 1998). Cowie
and Scholz (1992a) raised doubts as to whether the
linear displacement could be maintained to the fault
tip itself; however, the linear prole down to the res-
olution limit of eld and seismic mapping appears to
be accommodated by minor structures in many cases
(McGrath and Davison, 1995; Cowie and Shipton,
1998). A separate issue arising from a linear prole in
its ideal form is the angular, chevron-style of the
hanging-wall syncline in the fault-parallel prole (Fig-
ure 4a). The very small radius of curvature at the gra-
dient reversal is not easily accommodated by most fold-
ing mechanisms. This fold style would show up as
V-shaped stratum contour patterns dening the
hanging-wall and footwall closures, but the contours
on structural maps of extensional faults invariably
swing smoothly, dening exural, or buckle-style,
rather than chevron-style folds (e.g., Gibson et al.,
1989; Harding and Tuminas, 1989).
Ar c of a Ci r cl e
To avoid the problem of lack of evidence for chevron-
style folding of the cutoffs, one could simply t the arc
of a circle (Figure 4b) through the fault tips and point
of maximum displacement. The curvature of the arc is
constant, but a requirement for ductile deformation at
the fault tips remains.
Catenar y
A catenary curve (Figure 4c) can be dened as that
formed by a exible, homogeneous cord hanging be-
tween two points of support, acted on by no force
other than gravity. If gravity were important in deter-
mining the evolution of a fault, as is the case with very
large faults (e.g., Ma and Kusznir, 1995), the catenary
might be a more appropriate alternative to the linear
prole than the arc of a circle. As seen in following
sections, however, the difference in computed cutoff
stretch between a catenary curve and the arc of a circle
F i g u re 3 . Oblique viewof 3-Dseismic cube showing an extensional fault array. The top surface is clipped to a prominent, picked
marker. The faults within this array appear to showa range of displacement proles: in some cases, a linear prole gives the best
visual t; in others, sinusoid or circular proles appear more suitable. Some suggestions are annotated. Real fault systems like this
defy application of single, representative proles and illustrate that it is worth calculatingcutoff stretch for a range of prole types.
Dataexampleis fromTriassic of thesouthernNorthSea. Thefaults inthisarrayareJurassicinageandwererestrictedbyinteraction
withoneanother andwithasalt detachment belowandthefreesurfaceabove. TheD
max
/Lratiosinthisfault systemarelow, ranging
between 0.01 and 0.02. Note also the variation between faults in partitioningof displacement between footwall uplift and hanging-
wall subsidence.
(a) (b) (c)
(d) (e) (f)
-0.5 0 0.5
0.4
fault length L
displacement D
-0.5 0 0.5
0.4
fault length L
displacement D
-0.5 0 0.5
0.4
fault length L
displacement D
-0.5 0 0.5
0.4
fault length L
displacement D
-0.5 0 0.5
0.4
fault length L
displacement D
LINEAR CIRCULAR CATENARY
ELLIPTICAL SINUSOID
normalized distance 1 0
0
1
n
o
r
m
a
l
i
z
e
d

d
i
s
p
l
a
c
e
m
e
n
t
lin
e
a
r
s
i
n
u
s
o
i
d
c
a
t
e
n
a
r
y
c
irc
u
la
r
&
e
llip
t
i
c
a
l
F i g u re 4 . Parts (a) through (e) showvarious displacement models for extensional faults (discussed in text). Vertical horizontal
scale. In each example, curves are shown for D
max
/L ratios of 0.05, 0.1, 0.2, and 0.3. In practice, extensional faults that have ratios
greater than 0.1areuncommon (seeFigure4b), but theyareillustratedhereto emphasizethegeometrical contrasts. (f) Comparison
of the displacement proles shown in (a) through (e), with displacement and distance fromD
max
to one of the fault tips both
normalized.
St ewar t 5 9 1
is negligible at the D
max
/L ratios that characterize ex-
tensional faults.
Hal f- Per i meter of an El l i pse
The displacement distribution on a crack in a perfectly
elastic material has an elliptical form and has been em-
ployed in numerical models of fault evolution (Wil-
lemse et al., 1996). Elliptical displacement proles
(Figure 4d) can represent the at-topped displacement
proles that have been noted on relatively isolated, un-
restricted faults (Dawers et al., 1993; Schlische et al.,
1996) and also on laterally interacting fault elements
(Peacock and Sanderson, 1996).
Si nusoi d
One of the rst fault displacement models was char-
acterized by an approximately sinusoidal form (Figure
4e), having no angular apex at D
max
and a gradually
tapering displacement gradient toward the fault tips
(Walsh and Watterson, 1987). More recently proposed
bell-shaped proles have broadly the same form
(Cowie and Scholz, 1992a), although few real exam-
ples of faults having this type of prole have been
found (Cowie and Shipton, 1998; cf. Contreras et al.,
2000).
The alternative proles are compared on a nor-
malized displacement/fault half-length plot in Figure
4f; note that the elliptical prole appears circular on
this type of display. The proles chosen here to rep-
resent the alternative possibilities for displacement dis-
tributions on real faults are free from the superposed,
short-wavelength variations that characterize many
measured displacement proles (e.g., Nicol et al.,
1996a; Peacock and Sanderson, 1996; Cartwright and
Manseld, 1998). These variations from idealized pro-
les may indeed be due to cutoff stretch accommo-
dation structures, or measurement error, or some fault
evolution mechanism (e.g., growth by segment link-
age) that leads to more irregular nal cutoff patterns
than those shown in Figure 4. Nevertheless, the range
of proles chosen here should be adequate to dene
the range of minimum cutoff stretch in terms of
D
max
/L.
Q U A N T I F I C A T I O N O F C U T O F F S T R E T C H
The percentage stretch of L is calculated here as a func-
tion of D
max
/L for each of the displacement distribu-
tion models introduced previously, and the results are
shown in Figure 5a. The relevant equations are pro-
vided in the Appendix. For completeness, the amount
of stretch represented by a boxcar, or rectangular dis-
placement prole, is also shown (displacement
D
max
along the entire fault), but this prole has innite
displacement gradients at each end of the fault and
cannot occur on an isolated fault (Maerten et al.,
1999).
Ignoring the rectangular prole, the minimum
stretch is unsurprisingly given by the linear model and
the maximum by the elliptical prole. The stretch es-
timates lie within a range of 1% at D
max
/L less than
0.05. A rapid increase exists in all of the stretch esti-
mates greater than 0.05, having some divergence be-
tween the minimum and maximum estimates at a
given D
max
/L. The range of stretch estimates shown in
Figure 5a can be averaged (excluding the rectangular
prole) and represented by a best t trend line. This is
a power function and is shown in Figure 5b; its equa-
tion is
1.85
D
max
% cutoff stretch 202 (1)

L
This type of relationship also approximates the decay
in amount of stretch parallel with the fault plane, hav-
ing distance away from the fault plane. The precise
nature of this decay depends on the geometry of the
hanging-wall fold in the dip direction, which is in turn
related to the geometry of the fault (e.g., Gibson et al.,
1989; Kerr and White, 1994; Ma and Kusznir, 1995).
In these calculations I assume that the displace-
ment is partitioned entirely into hanging-wall subsi-
dence or footwall upliftin other words that displace-
ment D
max
is measured from a horizontal line
connecting the fault tips. Displacement partitioning,
however, appears to vary with fault scale and setting.
Displacement across kilometer-scale normal faults is
commonly partitioned on the order of 85:15% be-
tween hanging-wall subsidence and footwall uplift
relative to regional datums (J. A. Jackson et al., 1988;
Gibson et al., 1989). Exceptions at this scale occur
where there is a density inversion in the faulted lithol-
ogies, for example, salt migrates into the footwalls of
detached faults, where it can assist in partitioning the
majority of extensional fault displacement into foot-
wall uplift (e.g., M. P. A. Jackson and Talbot, 1994).
In such cases, the relevant proportion of D
max
would
have to be used instead of D
max
itself in Figure 5a to
5 9 2 Di spl acem ent Di stri buti ons and Stretch of Extensi onal Faul ts
F i g u re 5 . (a) Percentage of cutoff stretch plotted against
D
max
/L for various displacement prole models. Labels corre-
spond to the models shown in Figure 4 (a linear, b
circular, c catenary, d elliptical, e sinusoid); R
rectangular or boxcar prole (not shown in Figure 4). The rec-
tangular prole cannot occur in practice without transfer faults
at the ends of the main fault strand. (b) Range dened by the
elliptical and linear end members. Best t trend line (dashed)
tted to average values of the functions shown in part (a); r
2
0.9998. ShadedzonecorrespondstoD
max
/Lrangeidentied
inpart (c). (c) D
max
/Ldatacompiledfromvarioussources. Each
polygonrepresentstheeldoccupiedbyapreviouslypublished
dataset; most of theseconsistof between10and30datapoints.
Most dataarefromCowieandScholz (1992b), excludingstrike-
slip fault data sets. Additional data fromNicol et al. (1996a),
Schlischeet al. (1996), and Wibberley et al. (1999).
10
-4
10
-3
10
-2
10
-1
10
0
10
1
10
2
10
3
10
-3
10
4
10
3
10
2
10
1
10
0
10
-1
10
-2
m
a
x
i
m
u
m

d
i
s
p
l
a
c
e
m
e
n
t

D
m
a
x

(
m
)
fault length L (km)
0
.
0
0
1
0
.
0
1
0
.
1
1
D
m
a
x

/
L
(c)
(b)
(a)
0
1
2
3
4
6
5
7
8
9
10
0 0.05 0.1 0.15 0.2
D
max
/L
c
u
t
o
f
f

s
t
r
e
t
c
h

%
nominal lower limit
for defining commercially
significant structures
R d b
c
e
a
0
1
2
3
4
6
5
7
8
9
10
0 0.05 0.1 0.15 0.2
D
max
/L
c
u
t
o
f
f

s
t
r
e
t
c
h

%
0.01
estimate cutoff stretch. Smaller scale, outcrop and
centimeter-size faults tend to be depicted as lens-
shaped, having displacement partitioned approxi-
mately equally between hanging wall and footwall
(e.g., Peacock and Sanderson, 1994; Schlische et al.,
1996).
A review of published D
max
/L data from various
geological settings, encompassing a range of rock types
and scales of fault, is compiled in Figure 5c. Dening
the size range of fault that is of most commercial sig-
nicance, in terms of dening or segmenting a hydro-
carbon eld, is difcult, but above a nominal lower cut-
off of approximately 50 m D
max
, larger faults
consistently show D
max
/L ratios between 0.01 and 0.1
(Figure 5b, c). This in turn denes a range of stretch
likely to be encountered in kilometer-scale extensional
faults (i.e., at 0.01 D
max
/L 0.1) ranging from neg-
ligible to about 3%of the map-view length of the fault.
S I Z E A N D S P A T I A L D I S T R I B U T I O N O F
C U T O F F S T R E T C H A C C O M M O D A T I O N
S T R U C T U R E S
The results in Figure 5a suggest that strata adjacent to
an isolated extensional fault 10 km long in map view
that has D
max
of 1 km would have to accommodate
approximately 300 m of stretch. If the cutoff stretch
in this case were accommodated by a single fault dip-
ping at 45, that fault would have a throw of 300 m.
At the other extreme, the stretch could be accommo-
dated by a large number of millimeter-scale fractures,
perhaps evenly distributed along the length of the
fault. Field examples of extensional faults that have a
St ewar t 5 9 3
genetically related but mutually perpendicular joint set
were described by Kattenhorn et al. (2000). Rather
than one of the two extremes mentioned previously, it
is probable that a cutoff stretch accommodation struc-
ture population would consist of a mixture of large
(potentially seismic scale) and relatively small faults
and fractures (e.g., Nicol et al., 1996b; Pickering et al.,
1996).
Prediction of the spatial distribution of a fault
population is not straightforward; the depiction of
fairly evenly spaced stretch accommodation faults in
Figure 2 is schematic. One approach to the prediction
of subseismic faults and fractures is relating their den-
sity to the curvature of the folded surface (Antonellini
and Aydin, 1995; Stewart and Podolski, 1998). With
this in mind, the proles shown in Figure 4 indicate
various locations, dependent on type of prole, where
relatively small radius of curvature preferentially
causes faults to occur (note the assumption that the
fault plane itself is not curved). For example, the si-
nusoid prole has the smallest radii of curvature (i.e.,
maximum values of curvature) near the fault tips and
D
max
. In contrast, the circular, catenary, and elliptical
proles have sharp inections at the fault tips them-
selves but are otherwise relatively smooth proles.
Price and Cosgrove (1990, p. 189190) alluded to
fault-parallel stretch in terms of curvature of the beds
in the plane parallel with the fault. In that analysis,
however, Price and Cosgrove (1990) used the curva-
ture to directly quantify the bending strain in strong
beds of a given thickness, assuming that those beds de-
formed by tangential longitudinal strain. This approach
is probably not generally valid, as it does not consider
cutoff stretch and implies that the fault length de-
creases to allow the neutral surface in each strong bed
to remain the same length in the plane parallel with
the fault.
A seismic example from the North Sea is shown
in Figure 6. In this case the main fault has a D
max
/
L of approximately 0.05, and less than 1% of cutoff
stretch is predicted (see Figure 5). Nonetheless, a
hanging-wall strike line shows minor faults that trend
at a high angle to the main fault structure (Figure
6c). These faults are not obviously linked to under-
lying basement structures, unlike the faults that are
seen in the dip section (Figure 6b), and they are in-
terpreted here as cutoff stretch accommodation
faults, albeit minor ones. Some previously published
examples of minor faults are similar in appearance
and context to those shown in Figure 6 (e.g., gure
12 in Schlische [ 1995] ; gure 5 in Clausen and
Korstgard [ 1996] ). These published examples might
also represent instances of cutoff stretch accommo-
dation structures.
If cutoff stretch, and the requirement for a popu-
lation of structures to accommodate it, were as fun-
damental a property of extensional faults as argued
here, it might seem odd that few examples of these
structures are discussed in the literature (except Des-
tro [ 1995] and Roberts [ 1996] ). Abundant literature
exists, however, on minor faults that trend at high
angles to overlapping extensional faults within the
zones of overlap (relay zones). These are discussed in
relation to cutoff stretch accommodation in the fol-
lowing section.
D O E S C U T O F F S T R E T C H
A C C O M M O D A T I O N C O N T R O L F A U L T
S E G M E N T L I N K A G E ?
An alternative, but related, method to cutoff curva-
ture for identifying the location of cutoff stretch ac-
commodation structures is displacement gradient.
Zones of highest displacement gradient on a fault have
the highest rate of change in cutoff length relative to
a straight reference line connecting the fault tips, so
these are the zones where cutoff stretch accommo-
dation structures might be expected. Many major
fault systems are segmented, and the individual faults
are separated by zones of overlap, or relay ramps (e.g.,
Peacock and Sanderson, 1994; Trudgill and Cart-
wright, 1994). In extensional fault systems, parts of
faults within overlap zones commonly show the high-
est displacement gradients (Walsh et al., 1999; Con-
treras et al., 2000). Several eld studies have dem-
onstrated fractures and fault systems cutting across
overlap zones (e.g., Trudgill and Cartwright, 1994;
Mack and Seager, 1995; Nicol et al., 1996a), and
some studies have shown that these structures accom-
modate fault-parallel extension (Ferrill et al., 1999).
Faults that cross overlap zones linking the overlapping
faults are sometimes known as transfer faults and are
recognized as a key stage in the evolution of a fault
system from being segmented to linked in a single,
throughgoing fault (e.g., Peacock and Sanderson,
1994; Childs et al., 1995; Crider and Pollard, 1998).
Accommodation of cutoff stretch is offered here as a
mechanism that promotes the propagation of frac-
tures across overlap zones (requiring the transfer
faults to be nonvertical), in other words, as a key fac-
tor in the process of fault linkage.
5 9 4 Di spl acem ent Di stri buti ons and Stretch of Extensi onal Faul ts
St ewar t 5 9 5
F i g u re 6 . Seismicillustrationof minor faultsthat areinterpretedhereascutoff stretchaccommodationfaults. (a) Locationof seismic
panels on Top Triassic surface, Inner MorayFirth Basin, North Sea. This surfacewas mappedonakilometer-spaced2-Dseismicgrid.
The main fault is the Smith Bank fault. The D
max
/L ratio of this fault at the level of this surface is approximately 0.05. (b) Dip line
illustrating the Smith Bank fault and showing the location of the hanging-wall strike line. (c) Strike line along the hanging wall,
approximately 2 kmfromthe fault scarp. Minor faults that trend at a high angle to the Smith Bank fault are interpreted as cutoff
stretch accommodation structures (labeled). The thickening of the Cretaceous section away fromthe D
max
location is due to the
depositional architectureof the megasequence. Note that the vertical exaggeration of this seismic panel is approximately15 .
I M P L I C A T I O N S O F C U T O F F S T R E T C H
A C C O M M O D A T I O N S T R U C T U R E S F O R
F A U L T S E A L
Dip-closed structures in the footwalls of extensional
faults or hanging-wall rollovers of listric faults rely on
fault seal for upside volumetrics. Fault seal can fail in
several ways (reviews by Watts [ 1987] and Losh et al.
[ 1999] ). Two main types of fault seal failure occur.
First, reservoir juxtaposition leak points allow leakage
directly across the fault plane from one side to the
other (Allan, 1989; Knipe et al., 1998). Second, the
fault plane can act as a conduit to permeable layers at
shallower levels, with overpressured uids entering the
fault zone during episodes of fault zone dilation caused
by tectonic reactivation or the overpressure itself (Hip-
pler, 1993; Losh et al., 1999; Sibson, 2000). The rele-
vance of cutoff stretch accommodation structures to
both of these processes is discussed in the following
sections.
Reser voi r Juxtaposi ti on Leak Poi nts
Juxtaposition of reservoir and permeable layers di-
rectly across a fault is routinely mapped using fault-
plane separation (Allan) diagrams (Allan, 1989). These
diagrams are more accurate where using 3-D seismic
data, but where only two-dimensional (2-D) data are
available, it is possible to construct cutoff proles that
do not reect the actual strata geometries adjacent to
5 9 6 Di spl acem ent Di stri buti ons and Stretch of Extensi onal Faul ts
(a)
(b)
(c)
2
-
D

s
e
is
m
ic

lin
e
s
upper reservoir
in footwall
upper reservoir
in hanging wall
lower reservoir
in footwall
lower reservoir
in hanging wall
mapped D
max
no mapped juxtaposition leak point of
lower reservoir in footwall
possible juxtaposition leak point near
crest of lower reservoir in footwall
intersection of footwall cutoff
stretch accommodation fault
with main fault
intersection of hanging wall cutoff
stretch accommodation fault
with main fault
cutoff stretch
accommodation faults
F i g u re 7 . Illustration of the possible effect of cutoff stretch
accommodationfaultsonpoorlyconstrainedAllandiagrams.(a)
Map view showing the heave of a fault and minor stretch ac-
commodation structures. The location of an arbitrarily located
2-Dseismic grid is overlain. (b) The separation of two notional
reservoir horizonsasmight beconstructedonanAllandiagram
usingthe2-Dseismic grid shownin(a). Thefootwall closurein
thelower reservoir appearstobeoffsetfromtheupper reservoir
in the hanging wall. (c) Actual fault plane separation including
stretch accommodation faults, showing that there may be res-
ervoir juxtapositions near the crest of the closure in the lower
footwall reservoir. Thispossibilitycouldbedealt withbyadding
error bars to the initial Allan diagram.
the fault plane owing to omission of cutoff stretch ac-
commodation faults (Figure 7).
Fault seal analysis based on Allan diagrams can be
supplemented with estimates of fault rock attributes
such as permeability, derived from knowledge of dis-
placement, hanging-wall and footwall lithologies, de-
formation mechanisms, and local fault clustering char-
acteristics (Yielding et al., 1997; Knipe et al., 1998).
The fault zone models employed in these fault seal
analyses tend not to consider minor faults and fractures
trending at high angles to the main fault plane, such as
a population of cutoff stretch accommodation struc-
tures. For example, cutoff stretch accommodated by
subseismic faults in a footwall closure could result in a
population of fractures trending at high angles to the
main fault (Figure 8). A fracture set of this nature has
been described by Kattenhorn et al. (2000). These
fractures could be zones of relatively high or low per-
meability (Mattai et al., 1998) and, in the low-fracture
permeability case illustrated in Figure 8, act as uid
ow barriers, creating a poorly drained zone extending
some distance away from the leading edge of the fault
block. Employing the relationships between density of
fractures and bed curvature or fault displacement gra-
dient discussed in previous sections, a curvature prole
parallel with the main fault or a displacement map
might give an indication of the spatial variation in res-
ervoir degradation.
Faul t Pl ane Condui ts
Many dry holes have demonstrated that faults that sig-
nicantly offset reservoirs, juxtaposing them against
impermeable lithologies along most of their length,
have nevertheless leaked (e.g., Jones et al., 1999). The
timing of this leakage can commonly be linked to ep-
structural
contours
fractures accommodating
cutoff stretch
poorly drained
zone
K
x
vs.K
y
F i g u re 8 . Case where cutoff stretch is accommodated bynu-
merous fractures rather than a few larger faults. This fracture
set can impart signicant bedding-parallel permeability aniso-
tropytoasiliciclasticreservoir (K
x
vs. K
y
), inhibitingowparallel
with the fault. The depicted orientation of thesefractures is for
illustrative purposes only; trends in a real fracture set formed
in this circumstance would probably vary along the length of
the fault (Roberts, 1996). Fractures parallel with themainfault,
for example, associated with the dip-line prole through the
damage zone, are not shown.
St ewar t 5 9 7
isodes of fault reactivation (Sibson, 1992, 2000; Jones
et al., 1999), but predicting locations at which leakage
might occur along the length of the fault is not straight
forward. Mineral exploration geologists have long been
aware that uid ow along faults can be localized to
specic points (Newhouse, 1942), and several studies
have suggested that uid ow is preferentially localized
to fault or fracture intersections (e.g., Ren et al., 1995;
Sibson, 2000). Fewer data are available from petroli-
ferous basins; however, Losh et al. (1999) found lateral
variations in uid ow within a fault zone in the Gulf
of Mexico, noting that uid ow was concentrated in
the vicinity of an intersection of a minor fault and the
main fault. Strain concentration in the more extreme
case of intersecting faults that cut one another is well
documented (e.g., Watterson et al., 1998; Maerten et
al., 1999). The lines of intersection between cutoff
stretch accommodation faults and the associated main
fault zone are steeply dipping (e.g., Figures 2, 7c) and
so are ideal conduits for uids to leave deeply buried
hanging-wall reservoirs and travel to shallower per-
meable layers. The difculties in predicting the loca-
tion of cutoff stretch accommodation faults were dis-
cussed in a previous section, and a possible tendency
for cutoff stretch accommodation in segmented fault
systems to be concentrated in relay ramps was noted.
Field examples presented by Garden et al. (1997) and
Miller (1999) showed that upward uid ow parallel
with a fault system occurred near the tips of individual
faults rather than near the D
max
areas.
C O N C L U S I O N S
Extensional faults typically show D
max
/L ratios ranging
from 0.01 to 0.1. The calculations presented here show
that planar faults at the upper end of the D
max
/L range
experience up to 3% of stretch parallel with the trend
of the main fault itself, depending on the partitioning
of extension between hanging wall and footwall. This
cutoff stretch is accommodated by a population of
faults and fractures. The numerical and spatial prop-
erties of this population of stretch accommodation
structures have not been addressed in detail here, but
they should ultimately be included in 3-D models of
fault zones (e.g., Knipe et al., 1998) and structure pop-
ulations that accommodate the internal strain in fault
blocks (Walsh et al., 1996; Hesthammer and Fossen,
1998). Fractures and faults that link overlapping fault
segments during growth of a fault system may be ge-
netically related to cutoff stretch. In terms of fault seal,
cutoff stretch accommodation faults may complicate
displacement proles on Allan diagrams and give lo-
calized conduits for vertical leakage up fault zones. The
nature of the cutoff stretch accommodation structure
population affects reservoirs within fault blocks: per-
vasive fracture sets cause bedding-parallel permeability
anisotropy adjacent to the main fault, whereas a few
larger faults could dene pressure cell boundaries at a
eld scale.
A P P E N D I X : L I N E L E N G T H E N I N G
E X P R E S S I O N S F O R V A R I O U S F A U L T
D I S P L A C E M E N T M O D E L S
Fault trace length in map view (strike dimension) L
Maximum displacement D
max
Stretched cutoff length (100% displacement partitioned into hang-
ing wall) L
Li near Di spl acement Pr ol e (I soscel es Tr i angl e)
2
L
2
L 2 D (2)
max

4
Ar c of a Ci r cl e
2
L L
1
L D sin (3)
max
2
4D
max L
D
max
4D
max
Hal f of the Per i meter of an El l i pse
Half of the perimeter of an ellipse requires a complex numerical
solution. Empirical approximations are inaccurate at low D
max
/L.
The following computer code gives a reasonable numerical
approximation.
a1 0.44325141463
a2 0.06260601220
a3 0.04757383546
a4 0.01736506451
b1 0.24998368310
b2 0.09200180037
b3 0.04069697526
b4 0.00526449639
m(2 (D
max
/L)) 2
L L ((1 m (a1 m (a2 m (a3 m a4))))
m (b1 m (b2 m (b3 m b4))) log(1/
m))
Catenar y Ar c
Find a numerically using
L
a D a cosh (4)
max
2a
5 9 8 Di spl acem ent Di stri buti ons and Stretch of Extensi onal Faul ts
and use a in
L
L 2a sinh (5)

2a
Si nusoi d
2p 2
L pD
max
L cos(x) 1 dx (6)

0 2p L
R E F E R E N C E S C I T E D
Allan, U. S., 1989, Model for hydrocarbon migration and entrapment
within faulted structures: AAPG Bulletin, v. 73, p. 803
811.
Antonellini, M., and A. Aydin, 1995, Effect of faulting on uid ow
in porous sandstones: geometry and spatial distribution: AAPG
Bulletin, v. 79, p. 642671.
Cartwright, J. A., and C. S. Manseld, 1998, Lateral displacement
variation and lateral tip geometry of normal faults in the Can-
yonlands National Park, Utah: Journal of Structural Geology,
v. 20, p. 319.
Cartwright, J. A., C. Manseld, and B. Trudgill, 1996, The growth
of normal faults by segment linkage, in P. G. Buchanan and
D. A. Nieuwland, eds., Modern developments in structural in-
terpretation, validation and modelling: Geological Society Spe-
cial Publication, v. 99, p. 163177.
Childs, C., J. Watterson, and J. J. Walsh, 1995, Fault overlap zones
within developing normal fault systems: Journal of the Geolog-
ical Society, v. 152, p. 535549.
Clark, R. M., and S. J. D. Cox, 1996, A modern regression approach
to determining fault displacement-length scaling relationships:
Journal of Structural Geology, v. 18, p. 147152.
Clausen, O. R., and J. A. Korstgard, 1996, Planar detaching faults in
the southern Horn Graben, Danish North Sea: Marine and Pe-
troleum Geology, v. 13, p. 537548.
Contreras, J., M. H. Anders, and C. H. Scholz, 2000, Growth of a
normal fault system: observations from the Lake Malawi basin
of the east African rift: Journal of Structural Geology, v. 22,
p. 159168.
Cowie, P. A., and C. H. Scholz, 1992a, Physical explanation for the
displacement-length relationship of faults using a post-yield
fracture mechanics model: Journal of Structural Geology, v. 14,
p. 11331148.
Cowie, P. A., and C. H. Scholz, 1992b, Displacement-length scaling
relationship for faults: data synthesis and discussion: Journal of
Structural Geology, v. 14, p. 11491156.
Cowie, P. A., and Z. K. Shipton, 1998, Fault tip gradients and process
zone dimensions: Journal of Structural Geology, v. 20, p. 983
997.
Crider, J. G., and D. D. Pollard, 1998, Fault linkage: Three-
dimensional mechanical interaction between echelon normal
faults: Journal of Geophysical Research, v. 103, p. 24373
24391.
Dawers, N. H., M. H. Anders, and C. H. Scholz, 1993, Growth of
normal faults: displacement-length scaling: Geology, v. 21,
p. 11071110.
Destro, N., 1995, Release fault: a variety of cross fault in linked
extensional fault systems, in the Sergipe-Alagoas basin, NEBra-
zil: Journal of Structural Geology, v. 17, p. 615629.
Ferrill, D. A., J. A. Stamatakos, and D. Sims, 1999, Normal fault
corrugation: implications for growth and seismicity of active
normal faults: Journal of Structural Geology, v. 21, p. 1027
1038.
Garden, R., S. Guscott, K. A. Foxford, S. Burley, J. J. Walsh, and J.
Watterson, 1997, An exhumed ll and spill hydrocarbon fair-
way in the Entrada sandstone of the Moab anticline, Utah
(abs.), in J. Hendry, P. Carey, J. Parnell, A. Ruffell, and R.
Worden, eds., Second international conference on uid evolu-
tion, migration and interaction in sedimentary basins and oro-
genic belts: extended abstracts volume, p. 287290.
Gibson, J. R., J. J. Walsh, and J. Watterson, 1989, Modelling of bed
contours and cross-sections adjacent to planar normal faults:
Journal of Structural Geology, v. 11, p. 317328.
Harding, T. P., and A. C. Tuminas, 1989, Structural interpretation
of hydrocarbon traps sealed by basement normal block faults at
stable ank of foredeep basins and at rift basins: AAPG Bulletin,
v. 73, p. 812840.
Hesthammer, J., and H. Fossen, 1998, The use of dipmeter data to
constrain the structural geology of the Gullfaks eld, northern
North Sea: Marine and Petroleum Geology, v. 15, p. 549573.
Hippler, S. J., 1993, Deformation microstructures and diagenesis in
sandstone adjacent to an extensional fault: implications for the
ow and entrapment of hydrocarbons: AAPG Bulletin, v. 77,
p. 625637.
Jackson, J. A., N. J. White, Z. Garfunkel, and H. Anderson, 1988,
Relations between normal-fault geometry, tilting and vertical
motions in extensional terrains: an example from the southern
Gulf of Suez: Journal of Structural Geology, v. 10, p. 155170.
Jackson, M. P. A., and C. J. Talbot, 1994, Advances in salt tectonics,
in P. L. Hancock, ed., Continental deformation: Oxford, Per-
gamon Press, p. 159179.
Jones, G., P. Rorison, R. Frost, R. Knipe, and J. Colleran, 1999,
Tectono-stratigraphic development of the southern part of
UKCS Quadrant 15 (eastern Witch Ground Graben): impli-
cations for the MesozoicTertiary evolution of the Central
North Sea Basin, in A. J. Fleet and S. A. R. Boldy, eds., Petro-
leum geology of northwest Europe: Proceedings of the 5th con-
ference, Geological Society, London, p. 133151.
Kattenhorn, S. A., A. Aydin, and D. D. Pollard, 2000, Joints at high
angles to normal fault strike: an explanation using 3-D numer-
ical models of fault-perturbed stress elds: Journal of Structural
Geology, v. 22, p. 123.
Kerr, H. G., and N. White, 1994, Application of an automatic
method for determining normal fault geometries: Journal of
Structural Geology, v. 16, p. 16911709.
Knipe, R. J., G. Jones, and Q. J., Fisher, 1998, Faulting, fault sealing
and uid ow in hydrocarbon reservoirs: an introduction, inG.
Jones, Q. J. Fisher, and R. J. Knipe, eds., Faulting, fault sealing
and uid ow in hydrocarbon reservoirs: Geological Society
Special Publication, v. 147, p. viixxi.
Losh, S., L. Eglinton, M. Schoell, and J. Wood, 1999, Vertical and
lateral uid ow related to a large growth fault, south Eugene
Island Block 330 eld, offshore Louisiana: AAPG Bulletin,
v. 83, p. 244276.
Ma, X. Q., and N. J. Kusznir, 1993, Modelling of near-eld subsur-
face displacements for generalized faults and fault arrays: Jour-
nal of Structural Geology, v. 15, p. 14711484.
Ma, X. Q., and N. J. Kusznir, 1995, Coseismic and postseismic sub-
surface displacements and strains for a dip-slip normal fault in
a three-layer elastic-gravitational medium: Journal of Geo-
physical Research, v. 100, p. 1281312828.
Mack, G. H., and W. R. Seager, 1995, Transfer zones in the southern
Rio Grande rift: Journal of the Geological Society, v. 152,
p. 551560.
Maerten, L., E. J. M. Willemse, D. D. Pollard, and K. Rawnsley,
1999, Slip distributions on intersecting normal faults: Journal
of Structural Geology, v. 21, p. 259271.
Mattai, S. K., A. Aydin, D. D. Pollard, and S. G. Roberts, 1998,
Numerical simulation of departures from radial drawdown in a
St ewar t 5 9 9
faulted sandstone reservoir with joints and deformation bands,
in G. Jones, Q. J. Fisher, and R. J. Knipe, eds., Faulting, fault
sealing and uid ow in hydrocarbon reservoirs: Geological So-
ciety Special Publication, v. 147, p. 157191.
McGrath, A. G., and I. Davison, 1995, Damage zone geometry
around fault tips: Journal of Structural Geology, v. 17, p. 1011
1024.
Miller, M. G., 1999, Active breachingof a geometric segment bound-
ary in the Sawatch Range normal fault: Colorado, USA: Journal
of Structural Geology, v. 21, p. 769776.
Muraoka, H., and H. Kamata, 1983, Displacement distribution along
minor fault traces: Journal of Structural Geology, v. 5, p. 483
495.
Newhouse, W. H., 1942, Ore deposits as related to structural fea-
tures: Princeton, New Jersey, Princeton University Press, 280 p.
Nicol, A., J. Watterson, J. J. Walsh, and C. Childs, 1996a, The
shapes, major axis orientations and displacement patterns of
fault surfaces: Journal of Structural Geology, v. 18, p. 235248.
Nicol, A., J. J. Walsh, J. Watterson, and P. A. Gillespie, 1996b, Fault
size distributionsare they really power law?: Journal of Struc-
tural Geology, v. 18, p. 191197.
Peacock, D. C. P., and D. J. Sanderson, 1994, Geometry and devel-
opment of relay ramps in normal fault systems: AAPG Bulletin,
v. 78, p. 147165.
Peacock, D. C. P., and D. J. Sanderson, 1996, Effects of propagation
rate on displacement variations along faults: Journal of Struc-
tural Geology, v. 18, p. 311320.
Pickering, G., J. M. Bull, and D. J. Sanderson, 1996, Scaling of fault
displacements and implications for the estimation of sub-
seismic strain, in P. G. Buchanan, and D. A Nieuwland, eds.,
Modern developments in structural interpretation, validation
and modelling: Geological Society Special Publication, v. 99,
p. 1126.
Price, N. J., and J. W. Cosgrove, 1990, Analysis of geological struc-
tures: Cambridge, England, Cambridge University Press, 502 p.
Ren, S. K., J. L. Walshe, R. G. Paterson, R. A. Both, and A. Andrew,
1995, Magmatic and hydrothermal history of the porphyry-
style deposits of the Ardlethan tin eld, New South Wales, Aus-
tralia: Economic Geology, v. 90, p. 16201645.
Roberts, G. P., 1996, Variation in fault-slip directions along active
and segmented fault systems: Journal of Structural Geology,
v. 18, p. 835845.
Schlische, R. W., 1995, Geometry and origin of fault-related folds
in extensional settings: AAPG Bulletin, v. 79, p. 16611678.
Schlische, R. W., S. S. Young, R. V. Ackermann, and A. Gupta,
1996, Geometry and scaling relations of a population of very
small rift-related normal faults: Geology, v. 24, p. 683686.
Sibson, R. H., 1992, Implications of fault-valve behaviour for rupture
nucleation and recurrence: Tectonophysics, v. 211, p. 283293.
Sibson, R. H., 2000, Fluid involvement in normal faulting: Journal
of Geodynamics, v. 29, p. 469499.
Stewart, S. A., and R. Podolski, 1998, Curvature analysis of gridded
geological surfaces, in M. P. Coward, T. S. Daltaban, and H.
Johnson, eds., Structural geology in reservoir characterisation:
Geological Society Special Publication, v. 127, p. 133147.
Trudgill, B., and J. Cartwright, 1994, Relay-ramp forms and normal-
fault linkages, Canyonlands National Park, Utah: Geological So-
ciety of America Bulletin, v. 106, p. 11431157.
Walsh, J. J., and J. Watterson, 1987, Distribution of cumulative dis-
placement and seismic slip on a single normal fault surface:
Journal of Structural Geology, v. 9, p. 10391046.
Walsh, J. J., J. Watterson, C. Childs, and A. Nicol, 1996, Ductile
strain effects in the analysis of seismic interpretations of normal
fault systems, in P. G. Buchanan, and D. A Nieuwland, eds.,
Modern developments in structural interpretation, validation
and modelling: Geological Society Special Publication, v. 99,
p. 2740.
Walsh, J. J., J. Watterson, W. R. Bailey, and C. Childs, 1999, Fault
relays, bends and branch-lines: Journal of Structural Geology,
v. 21, p. 10191026.
Watterson, J., A. Nicol, J. J. Walsh, and D. Meier, 1998, Strains at
the intersections of synchronous conjugate normal faults: Jour-
nal of Structural Geology, v. 20, p. 363370.
Watts, N. L., 1987, Theoretical aspects of cap-rock and fault seals
for single- and two-phase hydrocarbon columns: Marine and
Petroleum Geology, v. 4, p. 274307.
Wibberley, C. A. J., J.-P. Petit, and T. Rives, 1999, Mechanics of
high displacement gradient faulting prior to lithication: Jour-
nal of Structural Geology, v. 21, p. 251257.
Willemse, E. J. M., D. D. Pollard, and A. Aydin, 1996, Three-di-
mensional analyses of slip distributions on normal fault arrays
with consequences for fault scaling: Journal of Structural Ge-
ology, v. 18, p. 295309.
Yielding, G., B. Freeman, and D. T. Needham, 1997, Quantitative
fault seal prediction: AAPG Bulletin, v. 81, p. 897917.

S-ar putea să vă placă și