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

net/publication/229337186

A geometric model of fault zone and fault rock thickness variations

Article  in  Journal of Structural Geology · February 2009

DOI: 10.1016/j.jsg.2008.08.009

CITATIONS READS

355 456

6 authors, including:

Conrad Childs Tom Manzocchi

University College Dublin University College Dublin
116 PUBLICATIONS   5,411 CITATIONS    113 PUBLICATIONS   2,712 CITATIONS   

SEE PROFILE SEE PROFILE

John J. Walsh Andy Nicol

University College Dublin University of Canterbury
242 PUBLICATIONS   12,814 CITATIONS    256 PUBLICATIONS   7,933 CITATIONS   

SEE PROFILE SEE PROFILE

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

SAIGUP View project

Earthquake to millenial-timescale deformation along the Hellenic subduction margin, Eastern Mediterranean View project

All content following this page was uploaded by John J. Walsh on 22 June 2021.

The user has requested enhancement of the downloaded file.

 Journal of Structural Geology 31 (2009) 117–127

Contents lists available at ScienceDirect

Journal of Structural Geology

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

A geometric model of fault zone and fault rock thickness variations

Conrad Childs a, *, Tom Manzocchi a, John J. Walsh a, Christopher G. Bonson b,
Andrew Nicol c, Martin P.J. Schöpfer a
a
Fault Analysis Group, UCD School of Geological Sciences, University College Dublin, Belfield, Dublin 4, Ireland
b
SRK Consulting (UK) Limited, Fifth Floor, Churchill House, Churchill Way, Cardiff, CF10 2HH, UK
c
GNS Sciences, PO Box 30368, Lower Hutt, New Zealand

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

Article history: The thicknesses of fault rock and fault zones and the fault normal separations for breached and intact
Received 5 March 2007 relay zones each show a positive correlation with fault displacement. The displacement to thickness
Received in revised form 25 August 2008 ratio, or average shear strain, varies for the different structures increasing from intact relay zones
Accepted 28 August 2008
(median value ¼ 0.27) to fault rocks (median value ¼ 50). The correlation for fault rocks is widely
Available online 27 September 2008
interpreted as a growth trend controlled by fault rock rheology, but the progression of displacement to
thickness ratios for the different structures suggests an alternative model. In this alternative model
Keywords:
a fault initiates as an array of irregular fault segments. As displacement increases, relay zones separating
Fault zone
Fault rock thickness fault segments are breached and fault surface irregularities are sheared off, to form fault zones con-
Fault damage zone taining lenses of fault-bounded rock. With further displacement these lenses are progressively commi-
Segmentation nuted, and ultimately converted to zones of thickened fault rock. The final fault rock thickness is
therefore influenced strongly by fault structure inherited from the geometry of the initial fault array. The
large scale range on which fault segmentation and irregularities occur provides the basis for application
of this model over a scale range of at least 7 orders of magnitude.
Ó 2008 Elsevier Ltd. All rights reserved.

1. Introduction positive correlation between fault displacement and fault rock

thickness is well established (Robertson, 1983; Hull, 1988; Marrett
Faults are zones of extreme internal complexity and heteroge- and Allmendinger, 1990), with particularly well defined trends
neous strain distribution over a wide range of scales. Although this sometimes being used to infer fault displacements from measure-
complexity does not lend itself to a simple description to which all ments of fault rock thickness (Marrett and Allmendinger, 1990;
faults conform, a simplified and generalised description of faults is Little, 1995). Given that fault rocks are, with rare exceptions, weaker
required to achieve a better understanding of fault evolution and than the surrounding wall rocks, the cause of a continual increase in
for many practical applications, such as the production of oil from fault rock thickness with increasing displacement is not self evident
faulted reservoirs and earthquake hazard assessment. Although the and a number of fault rock widening models have been proposed.
stages in development of a fault are reasonably well established, These widening models typically rely upon shearing and attrition of
the fault geometric components characteristic of these stages (e.g. fault wall rock (Hundley-Goff and Moody, 1980; Robertson, 1982;
relay zones, damage zones, fault zones, fault rocks, etc.) are not Scholz, 1987; Power et al., 1988) or strain hardening of fault rocks
generally combined in outcrop descriptions of fault architecture. In (Hull, 1988; Faulkner et al., 2003). These models are based on the
this paper we collate thickness measurements of fault rocks, fault premise that the positive correlation between fault displacement
zones, relay zones and damage zones and relate these to different and thickness represents a growth trend and that a low displace-
stages of fault growth. We use the collated data to propose a semi- ment fault will progressively widen with increasing displacement
quantitative description of fault structural evolution indexed to (Robertson, 1982; Scholz, 1987; Hull, 1988). With the exception of
ranges of average shear strain intensity associated with these the model of Power et al. (1988), fault rock thickness within these
different geometric components. models is directly controlled by the rheological properties of the
The geometric component of faults to have received most wall rocks and fault rocks without considering the extent to which
attention in the published literature is fault rock thickness. A broad fault geometry may control fault rock development. In this paper
we propose an alternative model in which the thickness and
distribution of fault rock are largely controlled by fault geometry.
* Corresponding author. While fault geometry may to a significant extent be controlled by
E-mail address: conrad@fag.ucd.ie (C. Childs). the lithological and rheological stratigraphy of the wall rock, for

0191-8141/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jsg.2008.08.009
118 C. Childs et al. / Journal of Structural Geology 31 (2009) 117–127

example, due to fault dip changes (refraction) across bedding

planes, we suggest that in other respects the rheological properties
of the fault rock and the wall rock lithologies are of secondary
importance. We argue that internal fault zone geometry and the
distribution of fault rock thickness over the fault surface are
strongly influenced by the locations and dimensions of steps or
bends of the fault surface produced during fault propagation.
Our model for the development of fault rock is underpinned by
a model for the internal geometry of faults and well documented
processes of fault zone evolution. The geometric model adopted
here recognises that faults are not simple planar features but are
highly complex zones within which displacement and strain are
concentrated onto one or several discrete slip surfaces or zones of
intense shearing, enclosing variably strained rock volumes (Wallace
and Morris, 1979; Cox and Scholz, 1988; Childs et al., 1997). This
complexity of structure is largely the result of modification of fault
surface geometric irregularities inherited from the initial propaga-
tion geometry of a fault (e.g., Childs et al., 1996b). When a fault
propagates through a rock volume it does so as an irregular and
segmented surface (Naylor et al., 1986; Mandl, 1987; Cox and Scholz,
1988; Huggins et al., 1995; Marchal et al., 2003; Walsh et al., 2003).
Fault segmentation is observed both in map view and in cross-
section on a wide range of scales (Tchalenko, 1970; Larsen, 1988;
Morley et al., 1990; Peacock and Sanderson, 1991; Walsh et al., 1999)
and for each mode of faulting (Dahlstrom, 1969; Aydin, 1988; Larsen,
1988; Woodcock and Fischer, 1986). The rock volumes between
adjacent kinematically related fault segments are zones of high
strain (Chadwick, 1986; Walsh and Watterson, 1991; Childs et al.,
1995) that are variously referred to as transfer zones, relay ramps,
fault bridges and steps. Here we follow Walsh et al. (1999) and use
the term ‘relay zone’ to refer to any such rock volume between
kinematically related fault segments irrespective of the mode of
faulting or the nature of the strain (contractional, extensional or
constant volume, i.e. neutral).
As displacement increases on a segmented fault, the strains at
relay zones increase, eventually causing failure of the relay zone by
the formation of a linking fault to form a ‘breached relay zone’
(Peacock and Sanderson, 1994; Childs et al., 1995; Ferrill et al., 1999;
Ferrill and Morris, 2001; Soliva and Benedicto, 2004; van der Zee
and Urai, 2005). The site of the breached relay zone will remain as
an irregularity or asperity on the new through-going continuous
fault surface, and both the asperity and the continuous fault may be
bypassed at a later stage. The former relay zone may become
incorporated into the fault, initially as a fault-bounded lens and
ultimately as a zone of thickened fault rock, which may be
entrained along the fault surface in the displacement direction.
Similarly, irregularities or asperities on a continuous surface will be
sheared off, initially forming fault-bounded lenses but ultimately
becoming comminuted to fault rock (Fig. 1). The simple model of

Fig. 1. Outcrop photographs of fault-related structures associated with asperity

removal and fault segmentation. (a) Photograph of a 9 m throw normal fault contained
within the Upper Carboniferous sandstones and shales of Round O Quarry, Lancashire,
UK. The fault dips to the right and the contact between the sandstone and overlying
shale is visible in the fault hanging wall. A fault-bounded triangular wedge of sand-
stone at the centre of the photograph is interpreted to have been derived from the
uppermost part of the footwall sandstone unit by removal of an asperity caused by
a change in fault dip across the sandstone-shale interface (see inset). The motion of the
wedge down the fault has been accompanied by intense fracturing of the component
sandstones and breccia formation where the lens tapers downwards. (b and c)
Segmented faults from the chalk and marl sequences of Flamborough, Yorkshire, UK.
(b) A segmented fault steps across a thin marl unit (above and to the left of the
hammer) giving rise to an irregularity which represents an asperity to fault movement.
(c) A region of intense fracturing adjacent to a bend in a fault. This fault is interpreted
to have had an initial geometry similar to (b) with increased displacement causing
fracturing due to concentration of strain at an initial asperity centred on the offset marl
unit. With further increase in displacement the fault bend may be by passed by
continued movement on the slip-surface arrowed.
 C. Childs et al. / Journal of Structural Geology 31 (2009) 117–127 119

fault architectural evolution presented here is one of progressive in their article is ‘‘the zone where most fault slip has occurred and
strain localisation resulting in a reduction in the area of active fault usually includes the slip surfaces and the band of fault gouge and .
surface by bypassing areas of overlapping fault surfaces and fault cataclasis’’ and, since it does ‘‘not include undeformed blocks
surface irregularities (Ferrill et al., 2001; Walsh et al., 2001). entrained in the fault zone’’, corresponds to the definition of fault
In this paper we collate thickness data for the structures char- rock used here. For the most part we have not discriminated fault
acteristic of the stages of fault zone growth (i.e. fault rock, fault rocks according to the wall rocks from which they are derived, we
zone, relay zone and damage zone). The data, which are from brittle have however, assigned deformation bands, which are character-
faults offsetting a range of rock types with displacements ranging istic of faults in porous sandstones, to a separate fault rock category
over 7 orders of magnitude, provide a semi-quantitative descrip- for reasons discussed below.
tion of fault zone geometry and geometric evolution that is Fault zone thickness measurements, with the exception of data
consistent with systematics of fault rock scaling and distribution. from Wolf (1985) and van der Zee (2002), have been collected by
the authors (Childs et al., 1996a; Childs, 2000; unpublished Fault
2. Data and terminology Analysis Group data) from outcrop and principally from cross-
sections through normal faults. The term fault zone has recently
Study of the evolution of fault thickness has been hampered by been defined as a system of related fault segments that interact and
terminology problems. Terms such as fault rock, fault zone and link and are restricted to a relatively narrow band or volume
damage zone are not uniquely defined and there are no unambiguous (Peacock et al., 2000). Adapting this definition to one which can be
rules for measuring the thicknesses of these features. It has become used as a guide to thickness measurement in the field, we define
common in recent years to describe fault zones as comprising a fault fault zone thickness as the distance between synthetic slip-surfaces
core, containing slip surfaces, gouge, cataclasites and breccias, and (same dip direction and sense of offset) that can be demonstrated at
a damage zone comprising subsidiary minor structures e.g. veins and outcrop to be kinematically related and which each accommodate
small faults (Caine et al., 1996). In this paper we do not use the two at least a few percent of the total offset. Slip-surfaces have been
component core/damage zone description but have extended our judged to be kinematically related when they are connected at
fault descriptions to include other data categories. The correspon- a branchpoint or when they bound regions of elevated strain (e.g.,
dence between the core/damage zone terms and terms used here is fault rock, zones of bed rotation, clusters of antithetic faults). The
shown in Fig. 2. The terms used in this paper are each described in this displacement required to discriminate fault zone bounding slip
section. surfaces from minor faults within a damage zone cannot be rigidly
For the purposes of data collation we have used the term fault defined since it depends on a number of factors such as fault style,
rock to refer to fault gouge, breccia and cataclasite. Field measure- fault density or the extent of outcrop; perhaps the most common
ment of breccia thickness can be subjective as breccias have a broad manifestation of a fault zone is an anastomosing network of
continuum of clast sizes and there is no rigidly defined size cutoff through-going synthetic slip surfaces and associated fault rock.
distinguishing breccia clasts from fault-bounded rock volumes Examples of structures which have been measured as fault zones
(Marrett and Allmendinger, 1990). Problems comparing breccia/ are shown in Fig. 3a and b.
gouge thickness data from different sources are therefore inevitable Relay zone measurements are derived exclusively from normal
(Evans, 1990). We have, where possible, minimised these problems faults and comprise measures of fault displacement and relay zone
by referring to descriptions in the source articles. Data from Knott separation, i.e. the fault normal distance measured between a pair
et al. (1996), for example, have been classified here as fault rock of relay zone bounding faults. The data are derived from the map
thickness, although in the original article, they are referred to as view geometries of relay zones from outcrop studies and inter-
fault zone thickness. The definition provided by Knott et al. (1996) preted 3D seismic reflection surveys, and include both intact relay
zones and breached relay zones in which one, or both, relay
bounding faults has propagated to intersect its partner fault.
Since particular outcrop conditions are required for unambig-
uous identification of a relay zone, many relay zones will instead be
characterised as fault zones. For example, the structures in Fig. 3a
and b may represent cross-sections through relay zones, but this
cannot be established from the available outcrop data; instead the
structures are taken to be fault zones. By contrast, the fault in Fig. 3c
displays a similar internal structure to those in Fig. 3a and b, but
because it is bounded by two slip-surfaces which both tip-out
within the outcrop, the structure can clearly be identified as a relay
zone. The structure shown in Fig. 3c is a contractional relay zone,
i.e. transfer of displacement between the relay zone bounding
faults requires contractional strains. Unlike the structure in Fig 3c, it
is seldom possible to identify relay zones on normal faults in cross-
section so that the available data for contractional and extensional
relay zones are limited (but see Walsh et al., 1999; Kattenhorn and
Pollard, 2001; Kristensen et al., 2008) and these data (other than
Fig. 3c) have not been included in this study.
Measurements of damage zones are derived from the literature.
A damage zone is generally taken as the volume of deformed wall
rocks around a fault surface that results from the initiation, prop-
Fig. 2. Schematic diagram comparing the terms fault rock, fault zone and relay zone agation and build-up of slip along faults (Kim et al., 2004).
used in this paper with the fault core/damage zone description. Fault rock is synon-
Measurement of damage zone thickness, which excludes the core
ymous with fault core sensu Caine et al. (1996). At the centre of the block diagram
(along the dashed line) the bulk of the displacement is accommodated on a single slip- thickness (Caine, et al., 1996), relies on defining a zone within
surface and therefore the fault zone thickness is equal to the thickness of the fault rock/ which fault density is higher than the background density, some-
fault core. Thin lines indicate faults with minor displacements. thing which, depending on the degree of strain heterogeneity, may
120 C. Childs et al. / Journal of Structural Geology 31 (2009) 117–127

Fig. 3. Outcrop photographs of normal faults offsetting a weakly lithified mixed clastic sequence at Taranaki, New Zealand. The main slip-surfaces bounding the fault zones in (a)
and (b) are arrowed and downthrow to the right. The total throw across the fault zone in (a) is 1 m and across (b) is some 10s of metres. In (c) two slip-surfaces bound a zone of bed
rotation and antithetic faulting. The upward decrease in throw on the hanging wall slip-surface and the complementary upward increase in throw on the footwall slip-surface
identify this structure as an intact relay zone. Individual fault segments are shown in black and two bedding horizons are shown with dotted lines. The total throw on the two relay
zone bounding fault segments is 0.4 m.

be difficult to establish. The published measurements of damage seven orders of magnitude. The dataset has a median dis-
zone thickness often include features which, by our scheme would placement:thickness ratio of 50:1 (Fig. 5). The range of fault rock
be classified as fault zones or relay zones, and are predominantly thicknesses measured on a single fault trace is as much as 2.5 orders
for faults within porous sandstones. of magnitude for a given displacement. There is a decrease in the
The fault components defined above do not incorporate the full spread of thickness data from 3 orders of magnitude at low
range of thickness measurements required to define fault geom- displacements to 2 orders of magnitude at high displacements
etry. For example, the width of the zone of normal drag adjacent to (>10 m). This reduction in spread is, at least partly, a sampling
a fault is not included, principally because these data are not yet artefact due to the decrease in the proportion of the fault surface
available in the literature and neither are we aware of any pub- observed in outcrop for large faults compared with small faults.
lished study which investigates the relationship between drag Fault rock data derived from high porosity sandstones are pre-
wavelength and displacement. Another and more important sented separately in Figs. 4b and 5, as these data tend to obscure the
measure in the context of this paper, would be the amplitudes of displacement:thickness correlation for fault rocks developed in
fault surface asperities (i.e. irregularities) and their variation with other lithologies. At small displacements, faults in high porosity
displacement and scale. Although measurements of fault surface sandstones are zones of porosity loss with or without grain size
geometry have been published (Power et al., 1987, 1988; Sagy et al., reduction. Individual deformation bands typically have offsets less
2007), these data are relatively sparse and are not, in any case, than 0.5 cm and are up to 0.5 cm thick. Increase in fault displace-
readily converted to asperity wavelength to displacement ratios ment is preferentially accommodated on newly formed deforma-
(see Section 5). tion bands rather than by reactivation of existing structures. This
It is clear from the above that the definitions of the various fault behaviour has been taken as evidence of strain hardening in
components for which we have collated data are often ambiguous deformation bands (Aydin and Johnson, 1983; Antonellini and
and measurements of their thicknesses can be subjective. Although Aydin, 1995), although laboratory experiments have replicated the
the semi-quantitative model of fault evolution described below development of multiple deformation bands without any evidence
recognises the spectrum of structures which occurs (from intact of strain hardening (Mair et al., 2000). Due to the non-reactivation
relays through to fault gouge), because of the subjective nature of of deformation bands, the displacement:thickness ratios (D:T) for
fault component definition and measurement, the model defines low displacement faults (<10 cm) in high porosity sandstones are
the progression of structures in terms of strain intensity rather than significantly lower than for faults in other rock types, and the
attempting to define precise ranges of strain magnitude for the median D:T ratio is closer to that measured for fault zones than
different fault components. fault rocks (Fig. 5). This difference becomes less marked at higher
displacements when deformation band formation is superseded by
3. Fault component thicknesses the formation of a through-going slip-surface on which subsequent
movement is focused (Jamison and Stearns, 1982; Antonellini and
3.1. Fault rock Aydin, 1995).
Fault rock thickness data are derived from a wide variety of
The fault rock thickness data (Fig. 4a) demonstrate the well sources and include the three modes of faulting in lithologies
established correlation between thickness and displacement over ranging from relatively unconsolidated mixed clastic sequences to
 C. Childs et al. / Journal of Structural Geology 31 (2009) 117–127 121

a b c
Thickness (m)

d e f
Thickness (m)

Displacement (m) Displacement (m) Displacement (m)

Fig. 4. Displacement versus thickness data for the different fault geometric components identified. Vertical bars are thickness ranges measured on individual faults on decametre
scales. The data are derived from the following sources. (a) Fault rock data; Otsuki (1978), Robertson (1983), Hull (1988), Blenkinsop (1989), Marrett and Allmendinger (1990),
Ameen (1995), Little (1995), Caine et al. (1996), Knott et al. (1996), Childs et al. (1996a), Foxford et al. (1998) and Childs et al. (2007). (b) Deformation bands; Knott (1994) and
Manzocchi (1997). (c) Fault zones; Wolf (1985), Childs et al. (1996a), Childs (2000), Faulkner and Rutter (2001) and van der Zee (2002). Relay zones; breached (d) and intact (e),
Imber et al. (2004), Soliva and Benedicto (2004) and Worthington (2006). (f) Damage zones; Caine et al. (1996), Knott et al. (1996), Beach et al. (1999), Fossen and Hesthammer
(2000), Shipton and Cowie (2001). Knott et al. (1996) and Beach et al. (1999) record damage zone thicknesses for one side of the fault only. These have been doubled in this
compilation. The fault rock, fault zone and deformation band data also include previously unpublished data for normal faults from a range of lithologies. Note that the data in Fig. 4b,
and to a lesser extent data shown in Fig. 4a, are influenced by a tendency to round up thickness or displacements less than 1 mm.

granites. The data demonstrate a positive correlation for all fault related to deformation history and conditions, it is perhaps not
modes (Fig. 6a) and for all lithologies, with similar D:T ratios in surprising that different fault types in different host rocks appear to
rocks of very different rheologies (Otsuki, 1978; Robertson, 1983). have similar D:T scaling. Some apparently distinctive characteris-
For example D:T ratios for crystalline rocks fall within the range of tics are however revealed by dividing the dataset into subsets. For
values recorded for poorly lithified sediments (Fig. 6b). Since there example fault rock thicknesses derived from limestones are greater
are many factors which may affect D:T ratios, including those than for clastic sediments (not shown). Previous workers have
conducted detailed studies on limited, more focused datasets, to
distil the effects of individual controls on D:T ratios (Marrett and
Allmendinger, 1990; Knott, 1994; Little, 1995; Childs et al., 2007),
however given the large numbers of factors which can control fault
rock thickness, and the relative paucity of data once they are sub-
divided according to each factor, we believe that it is probably
premature to derive mechanistic conclusions from the global
dataset collated here.

3.2. Fault zone

The fault zone thickness data display a broad correlation with

fault displacement. The median D:T ratio is 2.5 (Figs. 4c and 5) with
a spread in thickness measurements of up to 3.5 orders of magni-
tude at low displacements (<10 cm). Again there is a significant
drop in the data spread at higher displacements (a10 m). The
reduction in the number of measured low D:T ratios (<1) (and
a corresponding increase in the minimum D:T ratio) at higher
displacement is expected, as the slip-surfaces of a wide, large
Fig. 5. Probability distributions of displacement/thickness ratios for different fault displacement fault zone are less likely to be identifiable as part of
geometric components. the same zone at outcrop, than are the slip-surfaces of a thinner and
122 C. Childs et al. / Journal of Structural Geology 31 (2009) 117–127

The lower limit of the relay zone data occurs at D:T ratios w0.01
a (Figs. 4e and 5), because it becomes impossible to demonstrate
displacement transfer between adjacent fault segments at lower
ratios. For example, it would be difficult to demonstrate the pres-
ence of a relay zone between two faults of 1 m displacement with
100 m separation, however if the faults grow to a displacement of
10 m then the presence of a relay zone and related structures (e.g.,
a relay ramp) is more easily recognised.

3.4. Damage zone

Unlike the other categories, the compiled damage zone data do

not define a linear correlation between thickness and displacement
(Fig. 4f). The majority of the data lie within a trend which shows
a relatively subdued increase in thickness (w2 orders of magni-
tude) over the w 4.5 orders of magnitude range of sampled
displacements. The dataset as a whole suggests that damage zone
thickness is established at low displacements but does not grow in
direct proportion to the fault displacement. Such a growth trend
b might be expected, as the majority of the data are derived from
normal faults in porous sandstones and, as for deformation bands
in these rocks, damage zone growth may be retarded or stopped
once a through-going slip-surface is developed. The data of Beach
et al. (1999) define a slope of w0.5 (Fig. 4f) and are consistent with
this interpretation. However, the data from Knott et al. (1996)
define a linear correlation (slope of 1), whilst the data of Shipton
and Cowie (2001), define a linear relationship between displace-
ment and damage zone thickness only for fault displacements
greater than ca. 2 m (Fig. 4f). The reasons for the differences in the
trends defined by the different datasets are unclear. Because
damage zones in porous sandstones are thought to be bypassed
when a through-going slip surface forms at a particular displace-
ment, the correlation between damage zone thickness will break
down above that displacement. Therefore, unlike the other fault
components, damage zones cannot accommodate all, or the
majority, of a fault displacement over the full scale range. For this
reason we do not discuss further damage zones in the model for the
Fig. 6. Plots of fault rock thickness versus displacement for (a) the different modes of evolution of fault zone structure, although we do acknowledge that
faulting and (b) different lithologies. The data in (b) are for normal faults offsetting
poorly lithified clastic sequences (data from Otsuki, 1978; Childs et al., 2007), and for
zones of minor faults and fractures may be important fault
reverse and strike slip faults in crystalline rocks (data from Robertson, 1983; Hull, components, but not necessarily in terms of the shear strain they
1988; Blenkinsop, 1989). accommodate.

lower displacement fault. The individual slip-surfaces of a large 4. Model for fault zone growth
fault are therefore more likely to be recorded as several separate
faults. In addition, the distinction between fault rock and lenses The various fault components plotted in Fig. 4, with the exception
within a fault zone becomes more ambiguous on larger faults so of the fault damage zone data, each display a broadly linear corre-
that large displacement, narrow fault zones are more likely to be lation between fault displacement and thickness, or separation in the
recorded as a single zone of breccia or gouge, as opposed to a fault case of relay zones. While there is significant overlap in the distri-
zone. butions of each component, there is also a clear difference in their
median D:T ratios (Figs. 5 and 7). As the D:T ratio is a proxy for
average shear strain intensity, each of the various fault components
3.3. Relay zones (i.e., fault rock, fault zone and relay zone) can be regarded as repre-
senting structures which occur within a shear strain range. It is
Both the intact (Fig. 4d) and breached (Fig. 4e) relay zone data apparent, both from the difficulty in defining the differences
define positive correlations between displacement and thickness between the various fault components, and the extent of data overlap
(i.e. separation), and have median D:T ratios of 0.27 and 0.46 between them, that there is a continuous spectrum of strain inten-
respectively (Fig. 5). The higher ratio for breached relay zones sities within which the various categories of measurement are
reflects the tendency for relay zones to become breached with defined; difficulties of defining precise cut-offs between categories
increasing shear strain. There is a significant overlap in D:T ratios reflects both the variety of controlling factors and the sampling issues
for intact and breached relay zones as would be expected for discussed above. The continuous spectrum of strain intensities
a dataset collated from a wide range of rock types and sequences reflects the progressive nature of related deformation in which an
and from disparate fault systems. The degree of overlap between individual relay zone evolves through breaching and fault zone
breached and intact relays can be greatly reduced if the data are stages, and with sufficient displacement may ultimately yield a fault
derived from individual areas (e.g. Soliva and Benedicto, 2004), but rock thickness which approximates the initial relay zone separation.
is unlikely to be removed completely because faulted sequences are A conceptual model of fault evolution based on these considerations
inherently heterogeneous. is illustrated in Fig. 8. At the outset (very low displacements at Time
 C. Childs et al. / Journal of Structural Geology 31 (2009) 117–127 123

scheme the thickness of fault rock which will form at a point on the
fault surface is strongly influenced from the outset by the scale of
fault segmentation and the amplitude of fault surface irregularities.
While the separations of individual relay zones are fixed from the
outset, thereby placing an upper limit on fault rock thickness derived
from their breaching and comminution, segmentation can occur on
a wide range of scales so that several fault widening events caused by
relay zone destruction may occur at the same point on a fault (e.g.
structures C and H). This point is illustrated in Fig. 9 which shows an
initial fault array comprising four overlapping fault segments with
variable fault separations. As fault displacement increases, so does
the fault rock thickness measured across the zone as successively
larger zones of overlap are converted to fault rock. This figure is
highly stylised, ignoring fault rock thicknesses other than those
formed in the zones of overlap and assuming that each relay zone will
be incorporated into the fault to become fault rock, but serves to
illustrate how the thickness measured on a fault can increase
progressively although the individual fault geometric components, as
Fig. 7. Thickness versus displacement plot for the four fault geometric components
shown in Fig. 8, do not widen with increasing displacement. Unlike
used in defining the fault evolution model. The large circles show the measurements relay zone separations, asperity amplitudes are not defined solely at
made on the three faults shown in Fig. 3a–c. the propagation stage but asperities may also form as a result of
deformation of a fault surface (e.g. by bedding parallel slip in the wall
rock, Watterson et al., 1998) and by fault segment linkages. For
1) the fault comprises a series of segments each with surface irreg- example, breaching of D at Time 3 (Fig. 8) results in a large asperity (I)
ularities on a range of scales: the latter will include asperities arising in the new through-going fault which is bypassed at Time 4.
from refraction through heterogeneous sequences and rock volumes. Increased displacement beyond Time 4 could lead to preferred
We assume faults to be initially segmented on a wide range of scales displacement of the later footwall splay, with associated fault rock
but the kinematic coherence between fault segments at a given developed along this surface and possibly associated with deforma-
displacement will only be apparent over a limited scale which will tion of the hanging wall fault-bounded blocks. Alternatively, because
increase with increasing displacement. For example at Time 1 only of changes in the footwall geometry it is just as likely that either
the structure at E is easily identified as a relay zone while the fault a new slip surface will be generated or that any one of the hanging
separations for other structures (e.g. A and H) may be too large to wall splays again becomes the principal slip surface, with the footwall
allow them to be identified as components of a single fault. Similarly surface, associated fault rock and immediate hanging wall, becoming
the wavelength of fault surface irregularities may be too large in the new footwall. Repeated movement on slip surfaces and genera-
relation to fault offset and length to constitute a fault asperity; for tion of new slip-surfaces means that the fault zone structure and fault
example, a 100 m wavelength, 1 m amplitude irregularity is unlikely rock content can become very complex despite the operation of
to be the focus for high strain on a fault with a displacement of 10 cm. a single relatively simple process, i.e. asperity removal upon an
As displacement increases all structures experience an increase in already irregular and segmented surface.
average shear strain and develop along growth lines which parallel Generation of new slip surfaces due to bypassing of fault surface
the displacement axis of the idealised displacement versus thickness asperities and movement on new or existing internal slip surfaces
plot in Fig. 8b. Progressively larger relay zones are breached and fault provide a means of locally increasing or decreasing fault rock, or
surface asperities bypassed, each structure becoming incorporated fault zone, thicknesses. Depending on the orientation of the new
into the fault zone and ultimately converted to fault rock which may slip surface, it may completely excise or duplicate the layer of fault
be entrained along the fault in the displacement direction. In this rock so that the preserved fault rock at that point records only the

a b
Fig. 8. (a) Block diagram illustrating the evolution of a fault array with increasing displacement at times 1 to 4. Individual fault irregularities, either asperities or segment
boundaries are labelled A to I. The change in structure with fault displacement (and shear strain) at these irregularities is tracked on the displacement-thickness plot in (b). Vertical
dashed lines in (b) are drawn at displacements at times 1 to 4 and the horizontal lines are growth curves for the individual fault geometric irregularities. Horizontal dashed lines
extend to the left of asperities incorporated into the fault. The areas filled in black in (a) indicate fault rock.
124 C. Childs et al. / Journal of Structural Geology 31 (2009) 117–127

Figure 8b shows separate ranges for each of the fault components

a
although the different ranges have significant overlap (Figs 5 and 7)
and the passage of a point on a fault from, for example, the fault zone
to fault rock field represents a gradual rather than abrupt increase in
the proportion of the incorporated rock volume converted into fault
rock. In accordance with field observation (e.g. Bonson et al., 2007)
Fig. 8a indicates the preferred site for fault rock generation to be fault
branchlines, whether at linkages between segments (A at Time 4) or
at the limits of fault-bounded lenses (B at Time 3).
Fig. 8 illustrates fault evolution at a single scale, although the
processes illustrated will occur simultaneously over a wide range of
scales. At Time 1 the structures labelled in Fig. 8a have a limited
range of D:T ratios, the magnitude of D:T ratios associated with
each structure increasing as the fault grows. However, at scales of
observation smaller than shown in Fig. 8, higher D:T ratios would
be encountered so that fault rock would be seen on the fault
segments at Time 1. The structure shown at Time 4 and the asso-
ciated D:T ratios could therefore represent a portion of one of the
fault segments at Time 1. Therefore at any particular time the range
of D:T ratios associated with a fault may range from infinite (at
a discrete slip-surface) to a lower limit determined by the largest
scale of irregularity/segmentation associated with the fault. The
range of D:T ratios and fault geometric components encountered is
therefore a function of both the displacement and the scale of
observation. Differences in fault geometry and mathematical
schemes for describing faults at different scales of observation have
been reviewed by Ben-Zion and Sammis (2004).
Fig. 3 shows examples of cross-sections through three normal
b
faults offsetting a poorly lithified sequence of sandstones and silt-
stones which serve to illustrate the model. Two thickness
measurements are made for each structure, a fault rock measure-
ment and a fault zone or relay zone measurement and these are
plotted on Fig. 7. The structures of the faults in Fig. 3a and c are very
similar, with two sub-parallel slip surfaces bounding a zone of low
strain accommodated by movement on a small number of antithetic
and, in the case of Fig. 3a, synthetic faults. In Fig. 3c, the upward
termination of the hanging wall bounding fault and the downward
termination of the footwall fault demonstrate that displacement is
transferred between them by an anticlockwise rotation of the relay
zone which is accommodated by movement on antithetic faults;
this structure is identified as an intact relay zone on Fig. 7. The
bounding faults of the structures in Fig. 3a do not tip out within the
outcrop and the appropriate classification is therefore as a fault
zone. The fault in Fig. 3b has the same overall structure and thick-
ness as those in Fig. 3a and c, but has a displacement of some 10s of
metres. The higher shear strain across this structure (a factor of
w50 larger) has resulted in a zone of intense minor faulting, along
antithetic and synthetic faults. The fault rock thickness is w20
times higher than that on the lower displacement structures due to
Fig. 9. (a) Schematic illustration of the increase in fault rock thickness at a point on
both an increase in the number of internal minor faults and thick-
a fault due to the conversion of rock volumes within fault relay zones into fault rock.
Three relay zones with separations x, y and z are converted to fault rock (shaded) with ening of fault rocks associated with the two bounding slip surfaces.
increasing displacement (i to iv). (b) Log-log plot lot of fault rock thickness versus The structure in Fig. 3c is a contractional relay zone while the
displacement showing the change in fault rock thickness measured along a section remaining relay zone data presented, as described above, are for
through M in (a) at displacements i through to iv. This figure is simplified in several neutral relay zones i.e. map view relay zones on normal faults.
ways (see text) for example, the fault rock thickness derived from a relay zone is shown
to be equivalent to the initial relay zone separation.
However the suite of structures in Fig. 3 serve to illustrate the
fundamental feature of the proposed model, namely the evolution
of fault structure by intensification of strain within a zone whose
thickness is established at very low displacement.
displacement on the new slip surface or twice the original fault rock The model of fault evolution described is one of progressive
thickness; in that respect, the preserved fault rock and associated strain localisation resulting ultimately in a continuous fault con-
slip surfaces may represent an aggregate displacement which is taining variable fault rock thickness. The heterogeneity of fault rock
quite different from the displacement across the zone. Detailed thickness variations and fault rock type reflect the fact that both
analysis of fault rock content in otherwise relatively simple normal fault segmentation and fault surface asperities occur on a wide
faults reveals a level of complexity which is best explained by the range of scales. A model of fault architectural evolution based on
repetitive nature of slip surface generation and displacement segmentation and the repetitive nature of asperity removal
(Foxford et al., 1998; Berg and Skar, 2005). provides a basis for explaining not only the continued widening of
 C. Childs et al. / Journal of Structural Geology 31 (2009) 117–127 125

fault rock with increased displacement but also the complexity of here is slightly different to previous asperity models in that the
fault zone content. emphasis is placed on the shearing-off of asperities as blocks of
fault-bounded rock which are subsequently comminuted to fault
5. Discussion rock and entrained along the fault, rather than the progressive
incorporation of fault rock by continuous migration of the wall
5.1. Fault architecture rock/fault rock interface and associated destruction of asperities.
While the end result in terms of volume of fault rock generated may
In recent years it has become common to describe fault internal ultimately be the same, the inclusion of large wall rock blocks
structure in terms of a high strain fault core surrounded by a damage within fault zones is more consistent with field observations (e.g.
zone, in which the fault core is the part of a fault where most of the Faulkner et al., 2003; Bonson et al., 2007).
displacement is accommodated and comprises fault gouge, breccia or Power et al. (1988) present a wear model based on observations
cataclasite (Caine et al., 1996; Rawling et al., 2001; Billi et al., 2003). of power-law scaling of fracture/fault surface roughness over scales
While there are obvious parallels between the core/damage zone from 5 mm to 40 m. In their model, fault rock volume generated by
model and the model described in this paper, with our fault rock slip across the fracture is calculated as the volume of interpene-
being synonymous with fault core, and our other components (fault tration of the two initially mated surfaces. This model does not
zones, breached and intact relay zones and damage zones) all incorporate the shearing off of discrete asperities. Definition of
included in the damage zone component, the differences between a model of fault development from an initial power-law fracture
the two models are not simply semantic since they reflect differences surface, which does incorporate shearing of asperities as lenses of
in implied strain distribution. In the fault core/damage zone relatively low strain rock, requires numerous additional assump-
description, shear strain is a maximum at the core of the fault and tions to define what constitutes a discrete asperity at different
decreases to a background strain at the damage zone margin. In the scales and fault displacements. We have not attempted to devise
model described here, strain is distributed heterogeneously within such a model here as these assumptions cannot, as yet, be sup-
a fault zone and strain distributions on transects across a fault vary ported by field evidence. The measurements of Power et al. (1988)
with location on the fault. Although certain transects across a fault cannot therefore be easily converted to amplitude:displacement
may conform to the core/damage description, highest strains (and data and plotted on Fig. 7 and, to our knowledge, there are no
regions of thickest fault rock) are generally at the margin of the fault published data on asperity amplitudes at different displacements.
zone with relatively little or no strain in the wall rock. These two Furthermore, there is no objective means of determining whether
conceptual models will tend to promote different models for the a fault-bounded lens was formed by asperity removal or fault
evolution of fault internal structure. An increase in core/damage zone segmentation. Therefore, while the model we describe acknowl-
thickness with displacement (Knott et al., 1996; Beach et al., 1999; edges the importance of fault surface irregularity, there is at
Shipton and Cowie, 2001), together with the strain distribution present no means of estimating the relative fault rock volumes
implied in the core/damage zone description, will lead to a model of formed due to segmentation and asperity removal.
fault evolution in which the zone of deformation progressively Strain hardening has previously been invoked as a possible
widens through time (Shipton and Cowie, 2003). Our model envis- explanation for the increase in fault rock thickness with increased
ages a more complex and often punctuated growth scheme, in which displacement (Hull, 1988; Faulkner et al., 2003). However, positive
fault rock thicknesses do not display a monotonic increase but may correlations appear to be observed for all faulted lithologies (e.g.
even show decreases, arising from the repeated activity of cross- Fig. 6b), even though it is relatively rare for fault rocks to be
cutting internal slip surfaces. stronger than their parent wall rocks (one exception is deformation
The term fault damage zone incorporates many different types bands in porous sandstones; Aydin and Johnson, 1983). Strain
of structures including horse-tails at fault tips, relay ramps and hardening does not therefore provide a general explanation for
normal drag: the variety of different structures which can be fault rock widening. Instead we present a geometric, as opposed to
classified as damage zone components has been described by Kim rheologic, model, in which fault rock weakening and progressive
et al. (2004). The scaling properties of these different structures will strain concentration accompany the expansion of fault zones and
not be the same. In our opinion, improved understanding of fault the accretion of fault rock. The fact that the D:T ratios for fault rocks
evolution can be achieved by studying separately the scaling derived from widely different lithologies, e.g. poorly lithified sedi-
properties of each component rather than by grouping all compo- ments and crystalline rocks (Fig. 6b), are relatively similar, even
nents into a single measure. Similarly, the different structures are though the deformation mechanisms and associated fault rock
associated with different three-dimensional arrangements of fault products may be very different, suggests that neither lithology nor
rock (which may have permeabilities either higher or lower than deformation mechanism have, on their own, a pre-eminent control.
both the wall rocks they separate and their parent wall rocks) and Instead, we suggest that geometric effects arising from fault surface
of juxtaposed wall rocks. These arrangements control flow across irregularities and segmentation may be the main factor. Never-
and within faults in either crystalline or clastic reservoirs or aqui- theless, because initial fault geometry is likely to be controlled by
fers. Therefore, as we have demonstrated elsewhere (Manzocchi the mechanical heterogeneity of the wall rock (e.g. bedding;
et al., 2008), we consider that outcrop studies and conceptual Peacock and Sanderson, 1992; Childs et al., 1996a; Wilkins and
models aiming to characterise the permeability structure of faults Gross, 2002; Ferrill and Morris, 2003; Schöpfer et al., 2006), it is
are more useful if the older, more specific, terms are used rather likely that rheology is, at least indirectly, an important factor.
than the generic terms (fault core, damage zone) which have
become prevalent in the last decade. 6. Summary and conclusions

5.2. Fault rock thickness A model for the evolution of fault internal structure has been
presented. In its simplest form, the model is one of progressive
The model of fault evolution described here incorporates the strain concentration in a zone, within which the active fault surface
generation of fault internal complexity both by segmentation progressively approaches, albeit along a potentially complex path,
leading to relay zone failure and by asperity removal. Asperity a more planar geometry. The width of this zone is influenced
models of fault rock widening have been presented previously strongly by the scale of initial fault segmentation and fault surface
(Scholz, 1987; Power et al., 1988 and others). The model described irregularity, but increasing strain intensity with increasing
126 C. Childs et al. / Journal of Structural Geology 31 (2009) 117–127

displacement results in an apparent widening of the various fault Billi, A., Salvini, F., Storti, F., 2003. The damage zone-fault core transition in
carbonate rocks: implications for fault growth, structure and permeability.
components. The by-passing of fault surface irregularities and the
Journal of Structural Geology 25, 1779–1794.
linkage of fault segments on a broad range of scales and displace- Blenkinsop, T.G., 1989. Thickness-displacement relationships for deformation
ments, provides for a model that relates different fault structures to zones: discussion. Journal of Structural Geology 11, 1051–1054.
differing degrees of strain intensity, as measured by increasing Bonson, C.G., Childs, C., Walsh, J.J., Schöpfer, M.P.J., Carboni, V., 2007. Geometric and
kinematic controls on the internal structure of a large normal fault in massive
displacement to thickness ratios. This model provides a basis for the limestones: the Maghlaq Fault, Malta. Journal of Structural Geology 29, 336–354.
well known positive correlation between fault rock thickness and Caine, J.S., Evans, J.P., Forster, C.B., 1996. Fault zone architecture and permeability
fault displacement. The principal features of this model are: structure. Geology 24, 1025–1028.
Chadwick, R.A., 1986. Extension tectonics in the Wessex Basin, southern England.
Journal of the Geological Society 143, 465–488.
 Irregular fault surfaces and fault segmentation arising from the Childs, C., 2000. The structure and hydraulic properties of fault zones. Unpublished
propagation of faults through rock volumes result in fault PhD thesis, University of Liverpool.
Childs, C., Watterson, J., Walsh, J.J., 1995. Fault overlap zones within developing
surface complexities which are bypassed during subsequent normal fault systems. Journal of the Geological Society 152, 535–549.
increases in displacement. Childs, C., Nicol, A., Walsh, J.J., Watterson, J., 1996a. Growth of vertically segmented
 The scale and distribution of initial fault surface complexities normal faults Journal of Structural Geology 18, 1389–1397.
Childs, C., Watterson, J., Walsh, J.J., 1996b. A model for the structure and develop-
provide a critical control on the thickness and distribution of ment of fault zones. Journal of the Geological Society 153, 337–340.
fault rock. Childs, C., Walsh, J.J., Watterson, J., 1997. Complexity in fault zone structure and
 Displacement accumulation causes the removal of asperities implications for fault seal prediction. In: Møller-Pedersen, P., Koestler, A.G.
(Eds.), Hydrocarbon Seals – Importance for Exploration and Production. Special
and the breaching and deformation of relay zones as a fault
Publication, 7. Norwegian Petroleum Society (NPF), pp. 61–72.
becomes progressively more planar with time. Childs, C., Walsh, J.J., Manzocchi, T., Strand, J., Nicol, A., Tomasso, M., Schöpfer, M.P.J.,
 Various fault components, such as relay zones, fault zones and Aplin, A., 2007. Definition of a fault permeability predictor from outcrop studies
fault rock, can be assigned to overlapping shear strain ranges of a faulted turbidite sequence, Taranaki, New Zealand. In: Jolley (Ed.), Struc-
turally Complex Reservoirs. Special Publication, 292. Geological Society,
which together define a continuous spectrum of strain. London, pp. 235–258.
 Fault zone complexity and ultimately fault rock thickness are Cox, S.D.J., Scholz, C.H., 1988. On the formation and growth of faults: an experi-
more directly controlled by fault geometry and segmentation mental study. Journal of Structural Geology 10, 413–430.
Dahlstrom, C.D.A., 1969. Balanced cross sections. Canadian Journal of Earth Sciences
than by the local wall rock and fault rock rheologies. The 6, 743–757.
length-scales and orientations of rheological heterogeneity Evans, J.P., 1990. Thickness-displacement relationships for fault zones. Journal of
(including the layering of the wall rock sequence) may, Structural Geology 12, 1061–1065.
Faulkner, D.R., Rutter, E.H., 2001. Can the maintenance of overpressured fluids in large
however, influence strongly the initial fault surface geometry strike-slip fault zones explain their apparent weakness? Geology 29, 503–506.
and segmentation. Faulkner, D.R., Lewis, A.C., Rutter, E.H., 2003. On the internal architecture and
 This model of fault rock development is consistent with the mechanics of large strike-slip fault zones: field observations of the Carboneras
fault in southeastern Spain. Tectonophysics 367, 235–251.
wide range of fault rock thicknesses on individual faults and Ferrill, D.A., Morris, A.P., 2001. Displacement gradient and deformation in normal
with the lack of a clear lithological dependence, because it fault systems. Journal of Structural Geology 23, 619–638.
acknowledges the complex nature of fault surface irregularities Ferrill, D.A., Morris, A.P., 2003. Dilational normal faults. Journal of Structural
Geology 25, 183–196.
and segmentation, combined with the temporal and spatial
Ferrill, D.A., Stamatakos, J.A., Sims, D., 1999. Normal fault corrugation: implications
variations in fault displacements. for growth and seismicity of active normal faults. Journal of Structural Geology
21, 1027–1038.
Ferrill, D.A., Morris, A.P., Sims, D.W., Stamatakos, J.A., 2001. The geometric strength
of fault systems. http://aapg.confex.com/aapg/de2001/techprogram/paper_
Acknowledgements 8288.htm.
Fossen, H., Hesthammer, J., 2000. Possible absence of small faults in the Gullfaks
Field, northern North Sea: implications for downscaling of faults in some
We thank former members of the Fault Analysis Group, in
porous sandstones. Journal of Structural Geology 22, 851–863.
particular Juan Watterson and Julian Strand, for discussions on fault Foxford, K.A., Walsh, J.J., Watterson, J., Garden, I.R., Guscott, S.C., Burley, S.D., 1998.
zone structure. Alan Morris and David Peacock are thanked for their Structure and content of the Moab Fault Zone, Utah, USA, and its implications
for fault seal prediction. In: Jones, G., Fisher, Q.J., Knipe, R.J. (Eds.), Faulting, Fault
helpful reviews of the manuscript. We thank the following
Sealing and Fluid Flow in Hydrocarbon Reservoirs. Special Publication, 147.
companies and their representatives for supporting our work on Geological Society, pp. 87–103.
faults in Taranaki, New Zealand, carried out as apart of the ITF Huggins, P., Watterson, J., Walsh, J.J., Childs, C., 1995. Relay zone geometry and
brokered FIFT research project: Amerada Hess, BG Energy Holdings, displacement transfer between normal faults recorded in coal-mine plans.
Journal of Structural Geology 17, 1741–1755.
BP Exploration, ConocoPhillips, Kerr-McGee North Sea, Shell, Statoil Hull, J., 1988. Thickness-displacement relationships for deformation zones. Journal
and Total. of Structural Geology 10, 431–435.
Hundley-Goff, E.M., Moody, J.B., 1980. Microscopic characteristics of orthoquartzite
from sliding friction experiments, II. Gouge. Tectonophysics 62, 301–319.
References Imber, J., Tuckwell, G.W., Childs, C., Walsh, J.J., Manzocch, T., Heath, A.E.,
Bonson, C.G., Strand, J., 2004. Three-dimensional distinct element method
Ameen, M.S., 1995. Fracture characterisation in the Chalk and the evolution of the simulations of relay growth and breaching along normal faults. Journal of
Thanet monocline, Kent, southern England. In: Ameen, M.S. (Ed.), Fractography: Structural Geology 26, 1897–1911.
Fracture Topography as a Tool in Fracture Mechanics and Stress Analysis. Special Kattenhorn, S.A., Pollard, D.D., 2001. Integrating 3-D seismic data, field analogs, and
Publication, 92. Geological Society, pp. 149–174. mechanical models in the analysis of segmented normal faults in the Wytch
Antonellini, M., Aydin, A., 1995. Effect of faulting on fluid flow in porous sandstones: Farm oil field, southern England, United Kingdom. American Association of
geometry and spatial distribution. American Association of Petroleum Geolo- Petroleum Geologists Bulletin 85, 1183–1210.
gists Bulletin 79, 642–671. Jamison, W.R., Stearns, D.W., 1982. Tectonic deformation of Wingate sandstone,
Aydin, A., 1988. Discontinuities along Thrust Faults and the Cleavage Duplexes. Colorado National Monument. American Association of Petroleum Geologists
Geological Society of America, Special Paper 222, pp. 223-233. Bulletin 62, 2584–2608.
Aydin, A., Johnson, A.M., 1983. Analysis of faulting in porous sandstones. Journal of Kim, Y.S., Peacock, D.C.P., Sanderson, D.J., 2004. Fault damage zones. Journal of
Structural Geology 5, 19–31. Structural Geology 26, 503–517.
Beach, A., Welbon, A.I., Brockbank, P.J., McCallum, J.E., 1999. Reservoir damage Knott, S.D., 1994. Fault zone thickness versus displacement in the Permo-Triassic
around faults: outcrop examples from the Suez rift. Petroleum Geoscience 5, sandstones of NW England. Journal of the Geological Society 151, 17–25.
109–116. Knott, S.D., Beach, A., Brockbank, P.J., Lawson Brown, J., McCallum, J.E., Weldon, A.I.,
Ben-Zion, Y., Sammis, C.G., 2004. Characterization of fault zones. Pure and Applied 1996. Spatial and mechanical controls on normal fault populations. Journal of
Geophysics 160, 677–715. Structural Geology 18, 359–372.
Berg, S.S., Skar, T., 2005. Controls on damage zone asymmetry of a normal fault Kristensen, M.B., Childs, C.J., Korstgård, J.A., 2008. The 3D geometry of small-scale
zone: outcrop analyses of a segment of the Moab fault, SE Utah. Journal of relay zones between normal faults in soft sediments. Journal of Structural
Structural Geology 27, 1803–1822. Geology 30, 257–272.
 C. Childs et al. / Journal of Structural Geology 31 (2009) 117–127 127

Larsen, P.-H., 1988. Relay structures in a Lower Permian basement-involved Scholz, C., 1987. Wear and gouge formation in brittle faulting. Geology 15, 493–495.
extension system, East Greenland. Journal of Structural Geology 10, 3–8. Schöpfer, M.P.J., Childs, C., Walsh, J.J., 2006. Localisation of normal faults in multi-
Little, T.A., 1995. Brittle deformation adjacent to the Awatere strike-slip fault in New layer sequences. Journal of Structural Geology 28, 816–833.
Zealand: faulting patterns, scaling relationships, and displacement partitioning. Shipton, Z.K., Cowie, P.A., 2001. Damage zone slip-surface evolution over mm to km
Bulletin of the Geological Society of America 107 11, 1255–1271. scales in high-porosity Navajo sandstone, Utah. Journal of Structural Geology
Mair, K., Main, I., Elphick, S., 2000. Sequential growth of deformation bands in the 23, 1825–1844.
laboratory. Journal of Structural Geology 22, 25–42. Shipton, Z.K., Cowie, P.A., 2003. A conceptual model for the origin of fault damage
Mandl, G., 1987. Discontinuous fault zones. Journal of Structural Geology 9, 105–110. zone structures in high-porosity sandstone. Journal of Structural Geology 25,
Manzocchi, T., 1997. Quantification of flow impairment in faulted sandstone 333–344.
reservoirs. Unpublished PhD thesis, Heriot-Watt University. Soliva, R., Benedicto, A., 2004. A linkage criterion for segmented normal faults.
Manzocchi, T., Heath, A.E., Palananthakumar, B., Childs, C., Walsh, J.J., 2008. Faults in Journal of Structural Geology 26, 2251–2267.
conventional flow simulation models: a consideration of representational Tchalenko, J.S., 1970. Similarities between shear zones of different magnitudes.
assumptions and geological uncertainties. Petroleum Geoscience 14, 91–110. Bulletin of the Geological Society of America 81, 1625–1640.
Marchal, D., Guiraud, M., Rives, T., 2003. Geometric and morphological evolution of van der Zee, W., 2002. Dynamics of fault gouge development in layered sand-clay
normal fault planes and traces from 2D to 4D data. Journal of Structural Geology sequences. PhD thesis, RWTH Aachen. Shaker Verlag, ISBN3-8322-0585-3, 155 pp.
2, 135–158. van der Zee, W., Urai, J.L., 2005. Processes of normal fault evolution in a siliciclastic
Marrett, R., Allmendinger, R.W., 1990. Kinematic analysis of fault-slip data. Journal sequence: a case study from Miri, Sarawak, Malaysia. Journal of Structural
of Structural Geology 12, 973–986. Geology 27, 2281–2300.
Morley, C.K., Nelson, R.A., Patton, T.L., Munn, S.G., 1990. Transfer zones in the East Wallace, R.E., Morris, H.T., 1979. Characteristics of faults and shear zones as seen in
African Rift System and their relevance to hydrocarbon exploration in rifts. mines at depths as much as 2.5 km below the surface. United States Geological
American Association of Petroleum Geologist Bulletin 74, 1234–1253. Survey Open File Report 79-1239, 79–100.
Naylor, M.A., Mandl, G., Sijpesteijn, C.H.K., 1986. Fault geometries in basement- Walsh, J.J., Watterson, J., 1991. Geometric and kinematic coherence and scale effects
induced wrench faulting under different initial stress states. Journal of Struc- in normal fault systems. In: Roberts, A.M., Yielding, G., Freeman, B. (Eds.), The
tural Geology 8, 737–752. Geometry of Normal Faults. Special Publication, 56. Geological Society, London,
Otsuki, K., 1978. On the relationship between the width of shear zone and the pp. 193–203.
displacement along fault. Journal of the Geological Society of Japan 84, 661–669. Walsh, J.J., Watterson, J., Bailey, W., Childs, C., 1999. Fault Relays, Bends and Branch-
Peacock, D.C.P., Sanderson, D.J., 1991. Displacements, segment linkage and relay lines. Journal of Structural Geology 21, 1019–1026.
ramps in normal fault zones. Journal of Structural Geology 13, 721–733. Walsh, J.J., Childs, C., Meyer, V., Manzocchi, T., Imber, J., Nicol, A., Tuckwell, G.,
Peacock, D.C.P., Sanderson, D.J., 1992. Effects of layering and anisotropy on fault Bailey, W.R., Bonson, C.G., Watterson, J., Nell, P.A.R., Strand, J.A., 2001.
geometry. Journal of the Geological Society 149, 793–802. Geometrical controls on the evolution of normal fault systems. In:
Peacock, D.C.P., Sanderson, D.J., 1994. Geometry and development of relay ramps in Holdsworth, R.E., Strachan, R.A., Magloughlin, J.F., Knipe, R.J. (Eds.), The Nature
normal fault systems. American Association of Petroleum Geologists Bulletin of the Tectonic Significance of Fault Zone Weakening. Special Publication, 186.
78, 147–165. Geological Society, London, pp. 157–170.
Peacock, D.C.P., Knipe, R.J., Sanderson, D.J., 2000. Glossary of normal faults. Journal Walsh, J.J., Bailey, W.R., Childs, C., Nicol, A., Bonson, C.G., 2003. Formation of
of Structural Geology 22, 291–305. segmented normal faults: a 3-D perspective. Journal of Structural Geology 25,
Power, W.L., Tullis, T.E., Brown, S.R., Boitnott, G.N., Scholz, C.H., 1987. Roughness of 1251–1262.
natural fault surfaces. Geophysical Research Letters 14, 29–32. Watterson, J., Childs, C., Walsh, J.J., 1998. Widening of fault zones by erosion of
Power, W.L., Tullis, T.E., Weeks, J.D., 1988. Roughness and wear during brittle asperities formed by bed-parallel slip. Geology 26, 71–74.
faulting. Journal of Geophysical Research 93 (B12), 15268–15278. Wilkins, S.J., Gross, M.R., 2002. Normal fault growth in layered rocks at Split
Rawling, G.C., Goodwin, L.B., Wilson, J.L., 2001. Internal architecture, permeability Mountain, Utah: influence of mechanical stratigraphy on dip linkage, fault
structure, and hydrologic significance of contrasting fault-zone types. Geology restriction and fault scaling. Journal of Structural Geology 24, 1413–1429.
29, 43–46. Wolf, R., 1985. Tiefentektonik des linksneiderrheinischen Steinkohlengebietes. In:
Robertson, E.C., 1982. Continuous formation of gouge and breccia during fault Drozdzewski, G., Engel, H., Wolf, R., Wrede, V. (Eds.), Beiträge zur Tiefentektonik
displacement. In: Goodman, R.E., Heuze, F.E. (Eds.), Issues in Rock Mechanics, Westdeutscher Steinkolhlenlagerstätten. Geologisches Landesamt Nordrhein-
Proceedings of the 23rd Symposium on Rock Mechanics. American Institute of Westfalen, Krefeld.
Mining Engineers, pp. 379–404. Woodcock, N.H., Fischer, M., 1986. Strike-slip duplexes. Journal of Structural
Robertson, E.C., 1983. Relationship of fault displacement to gouge and breccia Geology 8, 725–735.
thickness. American Institute of Mining Engineers Transactions 35, 1426–1432. Worthington, R., 2006. Geometric and kinematic analysis of Cenozoic faulting
Sagy, A., Brodsky, E.E., Axen, G.J., 2007. Evolution of fault-surface roughness with onshore and offshore west of Ireland. Unpublished PhD thesis, University
slip. Geology 35, 283–286. College Dublin.

View publication stats