Downloaded from http://sp.lyellcollection.

org/ at University of Birmingham on June 10, 2015

Identification and spatial distribution of fractures in

porous, siliciclastic sediments

R O Y H. G A B R I E L S E N 1, R A N D I - K R I S T I N A A R L A N D 1'2'3
& EINAR ALSAKER 1

1Geological Institute, University o f Bergen, All@aten 41, N-5007 Bergen, Norway

2 Norsk Hydro Research Centre, Sandsliveien 90, N-5020 Bergen, Norway
3 Present address." E S S O Harge a.s., Grenseveien 6, N-4033 Forus, Norway

Abstract: Spatial and size distribution data of fractures provide essential information in
statistical bulk strain estimates and in predictive fracture scaling models for sediments
which have undergone brittle deformation. Most fracture populations in naturally deformed
sediments include structures which originated at different stages, and in response to different
conditions during the deformation history of the rock. Hence, the final fracture population
may contain fractures which genetically are related to gravitational (near-surface) insta-
bilities, burial, uplift and unroofing, thermal expansion and contraction, and regional
tectonism. Furthermore, grain size, bed thickness and different rheological properties may
strongly influence the fracture style and frequency.
Particularly when only well cores are available, identification of fractures of different
genetic origin is difficult, and fracture classification and predictive modelling can only be
accomplished with data acquired from careful fracture logging. In such logging, information
on position of the well relative to major structures, the geometry of nearby faults, and the
general geological (sedimentary and tectonic) environment need to be included. Also, frac-
ture frequencies should be compared to, and normalized according to, tithology and bed
thickness, and the fracture frequency diagrams should be corrected for eventual well
deviation before predictive fracture frequency modelling is performed.

Flow in hydrocarbon reservoirs is commonly different fracture sets in a fracture population is

influenced by fracture systems, which may either likely to be influenced to different degrees by
enhance (Narr & Currie 1982; Nelson 1981; Till- parameters such as history of burial and uplift,
man 1983; Tillman & Barnes 1983; Watts 1983) bed thickness, fluid pressure, grain size and near-
or depress (Stearns & Friedman 1972; Dunn ness and position relative to major structures
et al. 1973; Pittman 1981; Aydin & Johnson like faults and folds, implying that fractures
1983; Wilke et al. 1985; Gabrielsen & Koestler produced in different settings and at different
1987; Antonellini & Aydin 1994) permeability. stages of the geological history of the deforming
This emphasizes the need to predict fracture fre- rock may appear with distinctly different fracture
quencies and spatial distribution of fractures, frequencies and spatial distributions.
and to evaluate permeability characteristics of In a number of recent studies, valuable attempts
fractures at all scales. The prediction of fracture have been made quantitatively to characterize
frequencies is, however, hampered by several fracture populations.These works have treated
methodological and practical restrictions. Firstly, the relation between fault dimensions and fault
a scale problem exists in fracture studies of hydro- displacement (Watterson 1986; Barnett et al.
carbon reservoirs, where fracture analysis is 1987; Walsh & Watterson 1987, 1988, 1989;
frequently based primarily on reflection seismic Marrett & Allmendinger 1991; Cowie & Scholz
data, wireline logs and cores. The problem 1992a, b; Gillespie et al. 1992; Jackson & Sander-
occurs because little information is available on son 1992), and between fracture frequencies and
fractures at the scale between faults with throws bulk strain (Childs et al. 1990; Walsh et al. 1991;
above the resolution limit in reflection seismic Scholz & Cowie 1990). Hopefully, such studies
data, which at the very best is in the order of 10 will provide tools which will help to make correct
to 20 m, and fractures mapped in cores which predictions of fracture distribution and con-
typically display offsets of centimetres and occa- nectivity in the future, but to obtain this, high
sionally decimetres. Secondly, fracture popula- quality fracture databases must be available.
tions in naturally deformed sediments are It is the scope of the present paper to focus on
generally inhomogeneous in the context that the quality of fracture data bases used in future
they include fractures that are not necessarily studies, and to propose criteria which should be
genetically related. The distribution of genetically applied in acquiring such data. Hence, we try

GABRIELSEN,R. H., AARLAND,R.-K. & ALSAKER,E. 1998. Identification and spatial distribution of fractures in
porous, siliciclastic sediments. In: COWARD,M. P., DALTABAN,T. S. 8~; JOHNSON,H. (eds) Structural Geology in
Reservoir Characterization. Geological Society, London, Special Publications, 127, 49-64.
 Downloaded from http://sp.lyellcollection.org/ at University of Birmingham on June 10, 2015

50 R.H. GABRIELSEN ET AL.

to characterize fracture systems of different Description of natural fracture populations

genesis, and then to evaluate some general
parameters that are of importance in fracture The term 'fracture' is frequently used in a casual
development. manner, and particular nomenclature problems
are related to the scale of such features; the dis-
tinction between joints and faults is closely
linked to the scale at which such features are
The inhomogeneity of natural fracture studied (e.g. Ramsay & Huber 1987, p. 505). In
populations our opinion, this problem sometimes influences
the exactness in descriptions of fractures, and
The idea that individual fractures grow regularly this is of particular importance in studies which
and concentrically from a point of nucleation by include classification and statistical handling of
stable or simple incremental processes (Walsh & fracture sets.
Watterson 1989), and that fracture swarms and To avoid exclusion of certain fracture types,
faults develop from single fractures into larger we propose to include all planar or subplanar
composite structures and networks with secondary inhomogeneities (in a rock) imposed
increasing deformation (Aydin 1978; Aydin & by outer or inner stresses in the definition of
Johnson 1978; Gabrielsen & Koestler 1987) is fractures (Gabrielsen & Aarland 1995). This defi-
intuitively feasible. Detailed strain analyses of nition generally follows that of Bates & Jackson
single structures, however, suggest that these (1987) and Nystuen (1989), and includes all
relations may be more complicated for single types of mode I fractures such as joints and
fractures as well as for complete fracture open and mineral-filled fissures (veins), mode II
systems, and that they are influenced to a con- (shear) fractures and mode III (hybrid) fractures
siderable extent by deformation mechanisms (in the therminology of Hancock (1985)). We
and physical characteristics of the individual particularly emphasize that 'deformation bands'
fracture (Gillespie et al. 1992; Wojtal 1994; (Aydin 1978; Aydin & Johnson 1978), which in
Bfirgmann et al. 1994). early stages of development are identified only
Westaway (1994) suggested that although self- as zones of reduced porosity, are included in
similarity in size distribution of fault popula- this definition. The term 'deformation band' is
tions is not essential for the result of strain esti- synonymous to granulation seam (Heald 1956),
mates, and that several methods for sampling shear fracture (Dunn et al. 1973) and band
may be successful, such estimates are sensitive fault (Cruikshank et al. 1991).
to the actual numerical relation between the To obtain a good base for analysis of fracture
fracture sub-populations. Also Wojtal (1994), populations, it is necessary to categorize frac-
studying contractional faults, demonstrated tures according to their genesis, in order to
that mixing of faults of different scales and dif- prevent mixing of fracture populations in statisti-
ferent genesis or path of development may cal modelling. Fractures related to burial and
result in faceted plots, which in turn suggests uplift/unroofing respond to the stresses defined
that the fault population is basically multifractal by the reference state and the thermal expansion
or pseudofractal. This will be the case if popula- and contraction (Engelder 1993). Tectonic frac-
tions of 'bounded' faults ('large' faults that tures, on the other hand, may be subdivided
delineate larger structures such as fault blocks into fractures related to 'contemporary tectonic
or duplexes) and 'unbounded' faults ('small' stress', which is defined on a plate-wide scale,
faults within larger structures) are mixed in the and which is presumably related to plate
analysis. Hence, in fracture analysis and in margin interactions (Sbar & Sykes 1973), and
the application of scaled fracture frequencies in 'local tectonic stress', which can be ascribed to
bulk strain estimates, the identification of the local topographic effects or stress fluctuations
hierarchical status of the fault, as well as the due to inhomogeneities associated with, for
processes of deformation associated with each example, fault plane geometries. Residual
fracture set, is essential. To obtain meaningful stresses may contribute to both tectonic- and
fracture statistics, it is important to acknowl- burial-related fracturing, but is assumed to be
edge the effects of such properties as grain of minor practical significance in fracture log-
size, bed thickness and different rheological ging.
characteristics associated with changing litholo- The identification of small-scale fractures
gies and diagenesis. In the study of cores, struc- related to these principal deformation types is
tures associated with drilling and core handling not trivial, since characteristics such as morphol-
also need to be identified and corrected for ogy and texture may be similar for fractures that
(Kulander et al. 1990). have been initiated and developed under different
 Downloaded from http://sp.lyellcollection.org/ at University of Birmingham on June 10, 2015

FRACTURES IN SILICICLASTIC SEDIMENTS 51

geological conditions. Therefore, the total geo- According to some authors, fracture popula-
logical environment of the fracture population tions generated by compaction and decompaction
needs to be taken into consideration in fracture are dominated by tensile (mode I) fractures.
analysis. To do this, however, one needs to char- Engelder (1985) used the Voight & St.Pierre
acterize the fracture populations, which occur in (1974) relation, which defines the influence of
different types of deformation environments, in rock mechanical properties and the thermal effect
general terms. on the state of stress at depth in the uniaxial refer-
In the following discussion, we distinguish ence state, to substantiate that sediments that have
between the different environments of fracturing, undergone burial and lithification may reach
namely those associated with burial and unroof- tensile stresses which are larger than the tensile
ing, soft-sedimentary gravitational surface (topo- strength of the rock during unroofing. The practi-
graphic) instabilities, and contemporary primary cal implication of this is shown in Fig. 2, which sug-
and secondary stresses. gests that, for example, a sandstone may undergo
tensile fracturing at a level where a shale remains
Fractures related to burial and unroofing undeformed during erosion and uplift. Diagenetic
Fractures associated with loading and volume processes and fluid pressure contribute continu-
reduction may be initiated at very shallow ously to change the rheological properties of the
levels in the subsurface (a few tens of metres; sediments, and hence may promote fracturing
(Maltman 1988)), and continue to develop during burial and eventual erosion and upheaval.
throughout the burial history of the sediment However, mode II fractures generated through
(Fig. 1). Also agents other than sediment loading vertical loading are also frequently reported in
(e.g. ice) can contribute to their initiation. unconsolidated and poorly consolidated sedi-
ments, and are indeed reproduced in sand box
experiments (e.g. Mandl et al. 1977). The shear
strength of an unconsolidated to poorly con-
solidated sediment may vary from a few kilopas-
cals to some megapascals (Maltman 1994).
Differential stresses of this magnitude are obtained
near the surface, and development of shear frac-
tures may continue during burial, eventually pro-
moted by enhanced fluid pressures. In fact, all
fracture populations in hydrocarbon wells in the
Norwegian continental shelf logged by us are
dominated by mode II fractures, and layers with
a substantial number of mode I fractures are
found only rarely. In outcrop studies (e.g. Mon-
tserrat Fan-Delta northeast Spain (Alsaker et al.
1996) and Mesaverde Group, Utah (Alsaker
unpublished data 1995)), mode I fractures consti-
D tute a much larger part of the fracture populations.
E This is generally in accordance with observations
P in basins which have been buried, but which are
T presently exposed at the surface (Engelder 1993,
H p. 56), and can most easily be explained by stress
release during unroofing.
It should of course also be expected that differ-
ent tectonic histories and contrasting maximum
Mode I depths of burial of the investigated basins influ-
ence fracture frequencies and relation between
ModeH the fracture modes. Also, the sampling procedures
may contribute to the differences (most investi-
gated wells are vertical, implying that cores are
Mode II (+Mo:t¢ I) cut parallel to the vertical fractures, which accord-
ingly may be strongly under-representated in the
Fig. 1. Schematic relation and relative importance of fracture logs).
fractures generated during burial/uplift/unroofing and Fracture logging of siliciclastic sediments in
tectonic deformation. Symbols indicate dominant cores is most frequently performed in hydrocarbon
fracture mode. reservoirs which may constitute siltstone- and
 Downloaded from http://sp.lyellcollection.org/ at University of Birmingham on June 10, 2015

52 R. H. GABRIELSEN E T AL.

Fig. 2. Stress versus depth for sandstone and shale during burial and uplift. Lithification occurs at maximum depth
of burial. Note that the sandstone enters the area of tensile fracturing before the shale during upheaval. From
Engelder (1985).

sandstone-dominated intervals up to a few hun- the deforming sediment (Gabrielsen & Koestler
dred metres thick, presently buried at depths 1987), and that the two conjugate sets of mode
between c. 100 and 3000-4000 m. In vertical sedi- II fractures would be planar structures with a
ment intervals of, say, 400 m thickness buried at dip angle of 60 °. Experiments (Mandl et al.
depths of 1000-1400 m, changes in differential 1977; Bjornevoll & Gabrielsen, unpublished)
stress between the upper and lower part of the and field observations (Aydin & Johnson 1978;
sequence will typically be in excess of 10MPa, 1983; Gabrielsen & Koestler 1987) suggest that
and such stress differences may, given that all the fractures would be recognized as deformation
other parameters (rheological properties, fluid bands with the characteristics of the early stages
pressure, bed thickness etc.) are constant, be suffi- of deformation, i.e. grain compaction and por-
cient to explain contrasting fracture frequencies. osity reduction (Gabrielsen et al. 1993) (Fig. 4).
Fracture frequency studies in a 350m thick In fine- and medium-grained rocks, this type of
interval in the Njord Field of mid-Norway fracture has a high potential of preservation
(between 2750 and 3100m below present sea- (Maltman 1988). Unfortunately, however, such
bed), and in two intervals of 250 and 150m in textures are not diagnostic for compactional
the Oseberg Field in the northern North Sea, fractures, because tectonic deformation bands
where all sequences are dominated by siltstones possess similar textures and orientation. Perhaps
and sandstones interbedded with mudstones, the best criterion to distinguish compactional
show no simple relationship between fracture fractures is by their spatial distribution.
frequency and present depth of burial (Fig. 3). Water-escape fractures may be seen as a type
This is in concert with similar results from hom- of compactional fracture which was active
ogeneous and competent rocks (Haimson & Doe during the early stages of burial. In siltstones
1983), suggesting that the effect of changes in and sandstones these are typically high-angle,
differential stress in intervals of such magnitudes planar structures, usually a few millimetres
is almost completely overruled by other param- wide. They are often associated with symmetrical
eters (fluid pressure and lithology). Surprisingly, upward- or downward-pointing drag (Fig. 5).
however, no correlation between fracture Outside the zones affected by drag, there is
frequency and overpressured intervals was commonly no relative offset of strata (Burbridge
found in wells studied in the Njord Field in the et al. 1988). Water-escape fractures may further
mid-Norwegian shelf (Pedersen 1992). be subdivided into two classes (Gabrielsen &
In a subsiding, extensional basin where Aarland 1990): fractures (joints) through which
principal horizontal stresses (orn and ah) are sig- water has escaped (Lowe 1975; Cheel & Rust
nificantly different (oru > ah) and smaller than 1986; Guiraud & S6guret 1987), and fractures
the loading stress (i.e. crv = p g z = Crl), fractures (faults) associated with collapse due to volume
related to compaction and decompaction would loss associated with water escape (Burbridge
have predictable orientation and geometry. It is et al. 1988; de Lange et al. 1988). The deforma-
also to be expected that fractures related to com- tional style associated with water escape may
paction are more evenly distributed throughout vary considerably with lithology, from single
 Downloaded from http://sp.lyellcollection.org/ at University of Birmingham on June 10, 2015

FRACTURES IN SILICICLASTIC SEDIMENTS 53

fractures in sandstones to fragmentation in shales. for example in the form of synsedimentary

Texturally, water-escape fractures in fine- to faults or growth faults, slumps and slides
coarse-grained siliciclastic sediments are charac- (Hardin & Hardin 1961; Rider 1978; Crans
terized by lamination parallel to the fracture et al. 1980; Stow 1986). These structures are
walls, laminae being defined by clay particles related to surface slope, intrinsic instabilities
and other tabular mineral grains (Maltman such as inverse gravitational contrasts, or tilted
1988). Internal wall-parallel laminar elutriation contact surfaces between layers with contrasting
of mineral grains may occur (Lowe 1975; Bur- shear strength or fluid pressure (Cloos 1968;
bridge et al. 1988), and there are indications that Bruce 1973; Mandl 1988, pp. 28-29). It is
packing of grains takes place in this process. The frequently assumed that tectonic tilting or
textures may depend upon degree of water satura- seismicity may contribute to destabilization
tion in the sediment, grading from hydroplastic causing softsedimentary faulting. Local stresses
through liquefaction to fluidization, the latter easily exceed the shear strength of sand and
representing the highest degree of grain sorting. other unconsolidated or poorly consolidated
In conclusion, fractures related to water escape sediments in such circumstances, and may
in sandstones and siltstones are most commonly result in the development of discrete shear
strata-bound, and are usually associated with zones. Such shear zones may not be readily dis-
other structures related to liquefaction. Hence, tinguishable from shear zones generated during
they are generally easily identified on the scale of regional-scale tectonic deformation in the field,
a well core where they happen to be more abun- since morphological and textural characteristics
dant. Consequently, water-escape fractures con- of fractures may not be diagnostically different
tribute significantly to fracture frequency only on (Petit & Laville 1987). However, enrichment of
the local scale, and are accordingly not believed clay minerals (Sverdrup & Bjorlykke 1992) and
to represent a great problem in the study of scaling the association with fluidization processes
relationships of fracture populations in sandstones (Owen 1987) may help to identify fractures
and siltstones. In mudstones and shales, however, generated in soft-sedimentary, gravity-induced
the problems may be significant (Guiraud & deformation. In general, however, fractures
S6guret 1987; Petit & Laville 1987). associated with soft-sedimentary gravitative
features may sometimes only be identified as
Tectonic fractures such by their overall geometry and the geological
Tectonic fractures may affect the entire crust and environment in which they occur.
may be generated near surface as well as in rocks Fracture frequency and spatial distribution are
at a deep level of burial (Fig. 1). In many pub- not commonly reported in the literature for soft-
lished studies, tectonic fractures are implicitly sedimentary gravitational faults, but field studies
considered to dominate natural fracture popula- of slides in fluviodeltaic sediments (Sverdrup
tions. Under moderate stresses, morphologies & Bjorlykke 1992), in marine sediments (Farrell
and texture of the fracture fill of tectonic frac- & Eaton 1988) and in accretionary prisms
tures probably have characteristics that are (Behrmann et al. 1988) have indicated that frac-
similar to those of compaction fractures, and tures related to soft-sediment gravity-driven,
distinguishing between the two types may be dif- near-surface deformation may be abundant,
ficult. However, the textures related to tectonic and that they locally may develop frequencies
fractures will frequently reflect strains that comparable to that of regional (contemporary)
exceed those of consolidation fractures. Tectonic or local second-order tectonic deformation
fractures have been given thoroughly described (Ord et al. 1988; Martinsen & Bakken 1990). It
in the literature (see Hancock 1985; Groshong is therefore suggested that fracture systems
1988; Antonellini et al. 1994). associated with soft-sedimentary gravitational
Tectonic fractures may basically be considered instabilities may contribute significantly to the
to be responses to first-order plate stresses (con- bulk fracture strain in fracture populations in
temporaneous tectonic stresses), or secondary sedimentary rocks, and even dominate such
derivatives or effects of such stresses (Fig. 1). populations on the local scale. However, soft-
Thus, the second-order tectonic fractures include sedimentary gravitational instabilities may be
those associated with gravitational (surface) strictly local phenomena, where the strain (e.g.
instabilities and locally deflected stresses in, for extensional faulting in the upper extensional lis-
example, in the vicinity pre-existing faults, rock tric fan of a slump) can be compensated for in
borders, along geometrical irregularities in fault the outcropping (compressional) toe-zone.
planes etc. It is anticipated that fractures generated by
Faults related to soft-sedimentary gravitational near-surface gravity-induced deformation poten-
surface instabilities may reach large dimensions, tially represent a source of error in scaling
 Downloaded from http://sp.lyellcollection.org/ at University of Birmingham on June 10, 2015

54 R. H. G A B R I E L S E N E T AL.

Well N-I Well N-2 Well N~ W e l l N--4

~m.o 20 40 ~ o 20 40 ~m~ o 20 40 ~ o 20 40

Legend
l -S~istone ~ within
tl - lie F o m ~ c m Fmmti~s.

J -.Tilje/Aldrt Fm'm. o e ~ fommiem.

K -Am'Hitra From. n -l'redomimndy M~ale
m d ~ .

Fig. 3. Depth versus fracture frequency in wells from two Norwegian hydrocarbon fields (Njord (N) and Oseberg
(O) fields).
 Downloaded from http://sp.lyellcollection.org/ at University of Birmingham on June 10, 2015

FRACTURES IN SILICICLASTIC SEDIMENTS 55

Wdl0-1 Well O-2 Well 0-3

z~J Fmmn~ z~ Fr~a~.fmm~
anm. 0 20 40 60 80 ~ 0 20 40 60 80
.... ; ; ; ;

4140
N
Legend
A. Sbe~d C~.
C - Noes Fm'mefi~
D - Etive W m m a t i m
E - Raaaocb Femmi~

Fig. 3. Continued

relationship studies if one fails to identify such relations to the major faults, to be related to
features. Since gravity-induced sliding is not the contemporary (regional) and the secondary
uncommon in the Norwegian shelf at different (local) stressfield. Where ah and crH were signifi-
stratigraphic levels (Gabrielsen & Robinson cantly different during deformation, such struc-
1984; Alhilali & Damuth 1987), it may be tures are characterized by predictable fracture
expected that fractures associated with gravity- orientations and relations. In most extensional
induced near-surface sliding locally may contri- systems investigated by us, fault populations
bute significantly to the total fracture frequencies are characterized by relatively uniform fracture
obtained from core logging in this area. frequencies within each fault block, with the
We consider the major faults (those which are fracture frequency increasing both in the hang-
above the limit of resolution in reflection seismic ingwall and the footwall in the proximity of
data) and those fractures which have first-order major faults. In our studies, the width of the
 Downloaded from http://sp.lyellcollection.org/ at University of Birmingham on June 10, 2015

56 R. H. GABRIELSEN E T A L .

Fig. 4. Schematic zonal symmetric texture in 'complete' deformation band in sandstone. UD: sandstone unaffected
by deformation. Zone 1: grain compaction. Zone 2: grain-size reduction and grain reorientation. Zone 3:
mineralized central zone. C J: central joint. Grain compaction in full width of the deformation band usually occurs
before grain-size reduction. After Gabrielsen & Aarland (1990).

zones with enhanced fracture frequencies ('the

damage zone') in the hangingwalls and footwalls
of extensional faults with vertical throws in the
order of some tens of metres, is frequently
recorded to be less than 20-30 m (Fig. 6). This
is in general accordance with several previous
studies (Jamison & Stearns 1982; Chester &
Logan 1986), although the width of the damage
zone is reported to vary (Beach et al. 1995). It
/- is also evident that strain, and hence fractures,
is commonly concentrated in hangingwalls
adjacent to irregularities such as flats or splays,
so that abnormally wide damage zones can be
expected at such sites (see Fig. 8) (Koestler &
Ehrmann 1991; Aarland & Skjerven 1998).
i | ,; The present data may seem to confirm the gen-
eral model for growth of fractures (deformation
bands) in siliciclastic sediments from single frac-
tures, widening into zones of deformation bands
as described by Aydin & Johnson (1978). This
may be in general agreement with the model of
Walsh & Watterson (1988, 1989) which predicts
that shear fractures grow concentrically from a
point of nucleation into a surface with elliptic
circumference. It is noted, however, that in
plaster and sand analogue models faults are
me
frequently seen to nucleate at the surface or at
the contact between the deforming 'sediment'
and basement (Fossen & Gabrielsen 1995). The
Fig. 5. Fractures associated with water escape. After development of single deformation bands into
Burbridge et al. (1988). deformation zones also suggests that strain
 #fr. #~
a) 15
d) 16o 50
40
10 3O

40 32 24 16 8 0 8 16 I00 so 60 40 20 0 20 40 60 so loo 120 140 160 Iso

1

Distance ~ o m ~ u l (m)i Distance from fault (cm)i

III

#fr. #ft.
b) .,o
40
3O

".". . . . . . . . . . . . . . v • ~ . . ~ ; . . . . . . . . . T. - . . . ;:-, --.~-~;~

150 120 90 60 30 0 30 60 90 170 150 130 II0 90 70 50 30 I0 I0 30 50 70 90 II0 130 150 170 190 210 230
Distance f r o m fault ( c m ) / D i ~ n c e from fault(¢m)~
I
I I

#~

[ ::::- Fracturefrequency Inhanglng wallj

°> I::L I ~" "Fracturefrequency in foolwall J

5o 40 30 20 lo o to 2o 30 4o ~o 6o
Distance ~om ~ult (cm)l
I

Fig. 6. Fracture distribution in hangingwalls and footwalls of extensional faults. Examples from Price Canyon, SE Utah: Castlegate Formation (Mesaverde Group).
 Downloaded from http://sp.lyellcollection.org/ at University of Birmingham on June 10, 2015

58 R. H. GABRIELSEN E T A L .

hardening plays an important part in the devel- Some practical consequences of dealing with
opment of shear fractures in porous siliciclastic complex fracture data sets
sediments. It may also be questioned whether
there is an upper limit for such mechanisms to By use of established (Mohr-Coulomb and
prevail in the development of larger faults. Griffith) fracture criteria in combination with
stress-depth-strength relations and statistical
approaches in scaling relationships, structural
geologists possess powerful tools for the predic-
The influence of bed thickness, grain size and tion of fracture distribution in reservoirs. Still,
lithology on fracture frequency a question may be posed as to whether the frac-
ture sub-populations which are defined at differ-
It is well-established that bed thickness affects ent scales, and which make up the total fracture
joint spacing (mode I fractures) in sediments population, need to be genetically related to
(Harris et al. 1960; Price 1966; Hobbs 1967), obey the mathematical relations described in
and both linear (Price 1966; Narr & Suppe the literature. If so, this demands that fracture
1991) and non-linear (Ladeira & Price 1981) populations applied in predictive studies of
relationships have been proposed. Based on plas- scaling relationships need to be genetically
ter experiments, Mandal et al. (1994) arrived at a homogeneous. This puts strict constraints on
non-linear relationship, and suggested that this sampling procedures. Previous and present data
relation is also valid for mode II and mode III strongly suggest that fracture populations in
fractures. siliciclastic sediments are, by nature, inhom-
Lithology and grain size rule rock strength, and ogeneous in the context that they are commonly
accordingly strongly influence fracturing. Based composed of several sub-populations of different
on fracture mapping in fine- and medium-grained genetic origin. Unfortunately, morphologies and
sandstones of the Oseberg and Njord Fields in the textures of these structures are such that both on
Norwegian shelf, Gabrielsen e t al. (1993) reported the meso-scale and on the micro-scale, the differ-
enhanced fracture frequencies in fine- to medium- ent genetic populations may be identified only
grained sandstones as compared to siltstones and after careful investigation.
coarse-grained sandstones. In some cases, the Therefore, procedures for acquisition of
number of fractures was enhanced by a factor fracture data need to be carefully designed,
of five in the fine-grained sequences. Figure 7 with particular emphasis on the identification of
shows an example from two small, oblique faults fracture sub-populations during field mapping or
in sandstones of the late Cretaceous Mesaverde core analysis. For fractures in rocks exposed at
Group in Utah. In this example, fracture frequen- the surface, the potential for studying structures
cies have been recorded over a distance of in three dimensions may, to a large extent,
approximately 60 m in sandy sequences with dif- secure the recognition of fracture sub-popula-
ferent thicknesses and grain size. Fracture fre- tions, and frequently provides the possibility of
quencies in all the units increase in the vicinity finding criteria for determining the genesis of
of the faults, but the thinner layer with fine- individual fractures as well as for fracture sets.
grained sandstones displays an average fracture For fractures studied in drill-cores, less infor-
frequency enhanced by a factor of approximately mation is usually available on the genesis of frac-
2.6 compared to that of medium-grained sand- ture sub-populations, and a total analysis of the
stones of twice the thickness, and a factor of fracture population and its tectonic and sedimen-
four compared to coarse-grained sandstones five tary environment should be performed before
times its thickness. fracture data are mixed with data from elsewhere
It is also noted that tectonic and compac- in statistical analysis.The problems with fracture
tional fractures in mudstones and shales both modelling in, for instance, wells in the Norwegian
in the field and in well cores, represent a particu- continental shelf, are further increased because at
lar problem since intense deformation or volume the present stage of exploration, wells drilled at
reduction caused by compaction and water the crests of rotated fault blocks are over-
escape in such lithologies frequently generate represented in the available data.This implies
rubble zones, which completely prevent the that the data are recorded in a position where
evaluation of fracture frequency. This problem footwall collapse is frequent, so that the fracture
is exaggerated by the difficulty in separating frequency may be anomalously high.
tectonic fractures, particularly in cores, from Furthermore, deviation of the well path and
those associated with volume reduction and the dip angle of the fractures in the drilled unit
water escape in the most fine-grained sediments influence fracture statistics in wells. Such effects
(Petit & Laville 1987; Guiraud & S6guret 1987). can be corrected for relatively easily (Sikorsky
 Downloaded from http://sp.lyellcollection.org/ at University of Birmingham on June 10, 2015

FRACTURES IN SILICICLASTIC SEDIMENTS 59

Co~ to 5 m 0,6,0
F. ,=. D MediumSand

Fracture
Frequency
J /Vertic~ offset = 0,3 m
it t Cal~l~ displacement = 2~15 m

/ /
Vertic~ offset = 0,1 m I i
Calc~naed ~splacement = 0,7 m i ---- Zone A
/ i I
t t
/ i - - - Zone B
1 I
/ I ..... Zone C
i TM ,,. I /?.~. i I

Zone D
I

.i

metres
0 10 ~ ~ 40 50

Fig. 7. Fracture frequencies associated with two small, oblique faults from the Late Cretaceous Mesaverde Group,
Utah. Fracture frequencies have been recorded over a distance of approximately 60 metres in sandy sequences with
different thicknesses and grain size.
 Downloaded from http://sp.lyellcollection.org/ at University of Birmingham on June 10, 2015

60 R. H. GABRIELSEN E T AL.

(a)

®
®

O'

(b)

A Fracture frequency B Fracture frequency C Fracture frequency

- i

F
=.
iV
r~

Q) a)

,v
F F F F F
D E

Fig. 8. Schematic representation of shifting fracture distribution in rotated fault blocks. (a). (1) Gravity slide;
(2) and (3) sandstone units with different thicknesses; (4) zone of enhanced strain associated with ramp-flat-ramp
fault; (5) fault with synthetic accommodation faults. (b). Schematic fracture distribution diagrams from vertical
wells A and B, and from sections C-C', D - D t and E-E'.
 Downloaded from http://sp.lyellcollection.org/ at University of Birmingham on June 10, 2015

FRACTURES IN SILICICLASTIC SEDIMENTS 61

1991; Versteeg & Morris 1994), but such correc- continental shelf, and John Korstggtrd and M. Ridvan
tions do not seem to be frequently reported in Karpuz for constructive comments on an early version
published fracture studies where such data are of the manuscript. The detailed and constructive
applied. criticism from the referees, C. P. North and Juan
Figure 8 shows schematically some complex- Watterson, is much appreciated.
Thanks also to Jane Ellingsen, Geological Institute,
ities of fracture distribution which can be expected University of Bergen, who drafted Figs 4 and 8, and
in an area affected by gravity-driven soft-sedimen- Tore Odinsen who helped with mastering computer
tary faulting, compaction and tectonic extension. drafting.
Owing to differences in the fault geometries, frac- The work was supported by dr.scient grants from
ture logging in two wells crossing the master faults Norsk Hydro a.s. (R.-K. Aarland) and the Norwegian
(sections A and B) would give different fracture Council of Science (Einar Alsaker).
frequency-depth curves. In section A a minor
synthetic fault which is traversed in the hanging-
wall of the master fault gives two fracture
frequency maxima. Section B traverses a gravity
References
slide in its upper part and a high-strain area asso- AARLAND,R.-K. & SKJERVEN,J. 1998. Fault and fracture
ciated with a flat in the master fault. The flat has characteristics of a major fault zone in the northern
also caused the development of accommodation North Sea: analysis of 30 seismic and oriented cores
faults in the hangingwall and in the footwall. In in the Brage Field (Block 31/4). This volume
both these sections, it is reasonable to assume ALmLAL~,K. A. & DAMUTH,J. E. 1987. Slide block (?) of
that the tectonic fractures completely overrule Jurassic sandstone and submarine channels in the
the compactional fractures which are particularly basal Upper Cretaceous of the Viking Graben,
frequent in layer 3. Norwegian North Sea. Marine and Petroleum
If these structures were exposed on land, it Geology, 4, 35-48.
ALSAKER,A. et al. 1996. The significance of the fracture
may be assumed that fracture mapping in arbi-
pattern of the Palaeocene - early Miocene Mon-
trary sections C - C ' and D - D ' would give results tserrat Fan-Delta, Catalan Coastal Range (NE
which are significantly different from those Spain), Tectonophysics, 226, 465-491.
obtained from the well logs in sections A and ANTONELLINI,M. & AYDIN, A. 1994. Effect of faulting
B, because fractures associated with compaction on fluid flow in porous sandstones, petrophysical
of layer 3 would represent a significant fracture properties. AAPG Bulletin, 78, 355 377.
sub'population in section D - D ' (Fig. 8). , & POLLARD,D., D. 1994. Microstructure of
To obtain the best possible database for deformation bands in porous sandstones at Arches
statistical fracture studies, we therefore suggest National Park, Utah. Journal of Structural Geol-
ogy, 16, 941 959.
that the following steps should be followed in
AYDIN, A. 1978. Small faults formed as deformation
fracture mapping of cores.
bands in sandstone. Pure and Applied Geophysics,
116, 913-913.
(1) The position of the well relative to major - - & JOHNSON,A. M. 1978. Development of faults as
structures, the geometry of near-by faults, zones of deformation bands and slip surfaces in
and the general geological (sedimentary sandstone. Pure and Applied Geophysics, 116,
and tectonic) environment needs to be 931-942.
known. This implies that careful analysis & 1983. Analysis of faulting in porous
of reflection seismic data should be per- sandstones. Journal of Structural Geology, 5 19-31.
formed. BARNETT,J. A. M., MORTIMER,J., RIPPON,J. H., WALSH,
J. J. & WATTERSON,J. 1987. Displacement geome-
(2) Fractures should not be recorded in the
try in the volume containing a single normal
fracture logs without classification of frac- fault. AAPG Bulletin, 71,925 937.
ture mode and assumed genesis. BATES,R. L. & JACKSON,J. A. 1987. Glossary of Geology.
(3) Fracture frequencies should be compared American Geological Institute, Virginia.
to lithology and bed thickness. BEACH, A., BROCKBANK,P., BROWN, L., KNOTT, S.,
(4) Fracture frequency diagrams should be McCuLLUM, J. & WELBORN, A. 1995. Structural
corrected for eventual well deviation and characterisation of sandstone reservoirs-outcrop
fracture orientation relative to well axis. analogues for subsurface fault predictions.
(5) Core handling should follow standard (abstract). Structural Geology in Reservoir Charac-
terisation and Field Development. Imperial College,
procedures, and it is an advantage if
London, 1995, 6-7.
fracture logging is performed both prior to BEHRMANN,J. H., BROWN,K., MOORE,J. C., et al. 1988.
and after slabbing of the core. Evolution of structures and fabrics in the Barbados
Accretionary Prism. Insights from Keg 110 of the
The authors would like to thank Norsk Hydro a.s. for Ocean Drilling Program. Journal of Structural
permission to use data from wells in the Norwegian Geology, 10, 577-592.
 Downloaded from http://sp.lyellcollection.org/ at University of Birmingham on June 10, 2015

62 R . H . GABRIELSEN ET AL.

BRUCE, C. H. 1973. Pressured shale and related sedi- FOSSEN, H. & GABRIELSEN, R. H. 1995. Experimental
ment deformation, mechanism for development modeling of extensional fault systems by use of
of regional contemporaneous faults. AAPG plaster. Journal of Structural Geology, 18, 673-687.
Bulletin, 57, 878-886. GABRIELSEN, R. H. • AARLAND,R.-K. 1990. Character-
BURBRIDGE, G. H., FRENCH, H. M. & RUST, B. R. 1988. istics of pre-/syn-consolidation structures and
Water escape structures resembling ice-wedge casts tectonic joints and microfaults in fine- to medium-
in Late Quaternary subaqeuous outwash near grained sandstones. In: BARTON,N. 8z; STEPANSSON,
St.Lazare, Quebec, Canada. Boreas, 17, 33-40. O. (eds) Rock Joints. Balkema, Amsterdam, 4 5-50.
B~RGMANN, R., POLLARD,D. D. & MARTEL, S. J. 1994. & 1995. How can fracture analysis contribute
Slip distribution on faults, effects of stress gradi- to the understanding of previous and recent stress
ents, inelastic deformation, heterogeneous host- systems? In: FEJERSKOV, M. & MYHRVANG,A. M.
rock stiffness, and fault interaction. Journal of (eds) Rock Stresses in the North Sea. NTH/
Structural Geology, 16, 1675-1690. SINTEF, 38-48.
CHEEL, R. J. & RUST, B. R. 1986. A sequence of soft- -- 8~; KOESTLER, A. G. 1987. Description and struc-
sediment deformation (dewatering) structures in tural implications of fractures in the late Jurassic
the late Late Quaternary subaqeuous outwash near sandstones of the Troll Field, northern North
Ottawa, Canada. Sedimentary Geology, 47, 77-93. Sea. Norsk Geologisk Tidsskrift, 67, 371-381.
CHESTER, F. M. & LOGAN, J. M. 1986. Implications for & ROBINSON,C. 1984. Tectonic inhomogeneities of
mechanical properties of brittle faults from obser- the Kristiansund-Bodo Fault Complex, offshore
vations of the Punchbowl fault zone, California. mid Norway. In: SPENCER,A. M. et al. (eds) Petro-
Pure and Applied Geophysics, 124, 79-106. leum Geology of the North European Margin. Nor-
CHILDS, C., WALSH, J. J. & WATTERSON, J. 1990. A wegian Petroleum Society (Graham & Trotman),
method for elimination of the density of fault dis- 397-406.
placements below the limits of seismic resolution --, AARLAND, R.-K. & PEDERSEN, T. 1993. Distri-
in reservoir formations. In: BULLER,A. (ed.) North bution and genesis of natural fracture systems in
Sea Oil and Gas Reservoirs - II. Norwegian Petro- clastic sediments. In: BANKS, S. & BANKS, D. (eds)
leum Society (Graham & Trotman), 309-318. Hydrogeology of Hard rocks. Memoirs of the
CLoos, H., 1968. Experimental analysis of Gulf Coast XXIVth Congress International Association of
fracture patterns. AAPG Bulletin, 52, 420-441. Hydrogeologists, 52-73.
Cowia, P. A. & ScnoLz, C. H. 1992a. Physical explana- "GILLESPIE, P. A., WALSH, J. J. & WATTERSON, J. 1992.
tion for the displacement-length relationship of Limitations of dimension and displacement data
faults using a post-yield fracture mechanics from single faults and the consequences for data
model. Journal of Structural Geology, 14, 1133- analysis and interpretation. Journal of Structural
1148. Geology, 14, 1157-1172.
- - & 1992b. Displacement-length scaling rela- GROSHONG,R. H. Jr 1988. Low temperature deformation
tionship for faults; data synthesis and discussion. mechanisms and their interpretation. Geological
Journal of Structural Geology, 14, 1149-1156. Society of America, Bulletin, 100, 1329-1360.
CRANS, W., MANDL, G. & HARREMBOURRE,J. 1980. On GUIRAUD, M. & St~GURET, M. 1987. Soft-sediment
the theory of growth faulting. A geomechanical micro-faulting related to compaction within the
delta model based on gravity sliding. Journal of fluvio-deltaic infill of the Soria strike-slip basin,
Petroleum Geology, 2, 265-307. northern Spain. In: JONES, K. E. & PRESTON,
CRUIKSHANK, K. M., ZHAO, G. & JOHNSON, A. M. 1991. M. F. (eds) Deformation of Sediments and Sedi-
Duplex structures connecting fault segments in mentary Rocks. Geological Society, London,
Entrada sandstone. Journal of Structural Geology, Special Publication, 29, 123-136.
13, 1185-1186. HAIMSON, B. C. & DOE, T. W. 1983. State of stress, per-
DE LANGE, G. J., KOK, T. T. J., EBBING, J. & VAN DER meability, and fractures in the Precambrian granite
KLUGX, P. 1988. Microfault-like structures in of northern Illinois. Journal of Geophysical
unconsolidated Upper Quaternary sediments Research, 88, 7355-7372.
from the Madeora abyssal plain (eastern North HANCOCK, P. L. 1985. Brittle microtectonics. Principles
Atlantic). Marine Geology, 80, 155-159. and practice. Journal oj Structural Geology, 7,
DUNN, D. E., LEFOUNTAIN,L. J. 8~;JACKSON,R. E. 1973. 437-457.
Porosity dependence and mechanism of brittle HARDIN, F. R. & HARDIN, G. C. Jr 1961. Contem-
fracture in sandstones. Journal of Geophysical poraneous normal faults of Gulf Coast and their
Research, 78, 2403 2417. relation to flexures. AAPG Bulletin, 45, 238-248.
ENGELDER, T. 1985. Loading paths to joint propagation HARRIS, J. F., TAYLOR, G. L. d~; WALPER, J. L, 1960.
during a tectonic cycle, an example from the Appa- Relation of deformation fractures in sedimentary
lachian Plateau, U.S.A. Journal of Structural rocks to regional and local structures. AAPG
Geology, 7, 459-476. Bulletin, 44, 1853-1873.
- - 1 9 9 3 . Stress Regimes in the Lithosphere. Princeton HEALD, M. T. 1956. Cementation of Simpson and St.
University Press. Peter Sandstone in parts of Oklahoma. Journal of
FARRELL, S. G. & EATON, S. 1988. Foliation developed Geology, 64, 16-30.
during slump deformation of Miocene marine sedi- HOBBS, D. W. 1967. The formation of tension joints in
ments, Cyprus. Journal of Structural Geology, 10, sedimentary rocks, an explanation. Geological
567-576. Magazine, 104, 550-556.
 Downloaded from http://sp.lyellcollection.org/ at University of Birmingham on June 10, 2015

F R A C T U R E S IN SILICICLASTIC SEDIMENTS 63

JACKSON,P. & SANDERSON,D. J. 1992. Scaling of fault Rocks. Geological Society, London, Special Publi-
displacements from the Badajoz-Cordoba shear cation, 29, 11-24.
zone, SW Spain. Tectonophysics, 210, 179-190. PEDERSEN, T. A. 1992. Strukturelle stromningsbarrierer
JAMISON,W. R. & STEARNS,D. W. 1982. Tectonic defor- og deres betydningJor komminikasjon og stromning
mation of Wingate Sandstone, Colorado National av hydrokarboner pd Njordfeltet. Cand.scient.
Monument. AAPG Bulletin, 66, 2584-2608. thesis, University of Bergen.
KOESTLER,A. G. & EHRMANN,W. U. 1991. Description PETIT, J.-P. & LAVILLE,E. 1987. Morphology and micro-
of brittle extensional features in chalk on the crest structures of hydroplastic slickensides in sand-
of a salt ridge (NW Germany). In: ROBERTS,A. M., stone. In: JONES, K. E. & PRESTON, M. F. (eds)
YILEDING,G. & FREEMAN,B. (eds) The Geometry of Deformation of Sediments and Sedimentary Rocks.
Normal Faults. Geological Society, London, Geological Society, London, Special Publication,
Special Publication, 56, 113-123. 29, 107-121
KULANDER, B. R., DEAN, S. L. & WARD, B. J. J. 1990: PITTMAN, g. D. 1981. Effects of fault-related granula-
Fractured core analysis." Interpretation, logging, tion on porosity and permeability of quartz
and use of natural and induced fractures in core. sandstones. AAPG Bulletin, 65, 2381-2387.
AAPG, Methods in Exploration Series No. 8. PRICE, N. J. 1966. Fault and Joint Development in Brittle
LADEIRA, F. L. & PRICE, N. J. 1981. Relationship and Semi-brittle Rock. Pergamon, Oxford.
between fracture spacing and bed thickness. Jour- RAMSAY, J. G. 8~ HUBER, M. I. 1987. The Techniques of
nal of Structural Geology, 3, 179-183. Modern Structural Geology, Volume 2: Folds and
LowE, D. R. 1975. Water escape structures in coarse- Fractures, 309-700.
grained sediments. Sedimentology, 22, 157-204. RIDER, M. H. 1978. Growth faults in Carboniferous of
MALTMAN,A. J. 1988. The importance of shear zones in western Ireland. AAPG Bulletin, 62, 2191-2213.
naturally deformed wet sediments. Tectonophysics, SBAR, M. L. & SYKES, L. R. 1973. Contemporary com-
145, 163-175. pressive stress and seismicity in eastern North
- - 1 9 9 4 . Deformation structures preserved in rocks. America. An example of intra-plate tectonics.
In: MALTMAN,A. (ed.) The GeologicalDeformation Geophysical Society of America Bulletin, 80,
of Sediments. Chapman & Hall, London, 261-307. 1231-1264.
MANDAL,N., DEB, S. K. 8~ KHAN, D. 1994. Evidence for Scr~oLz, C.H & CowIE, P. A. 1990. Determining a total
a non-linear relationship between fracture spacing geologic strain from faulting. Nature, 346, 837 839.
and layer thickness. Journal of Structural Geology, SIKORSKY, R. I. 1991. A diagram for interpreting orien-
16, 1275-1281. tation data for planar features in core. Journal of
MANDL, G. 1988. Mechanics of Tectonic Faulting. Structural Geology, 13, 1085-1089.
Models" and Concepts. Elsevier, New York. STEARNS, D. W. & FRIEDMAN, M. 1972. Reservoirs in
, DE JONG, L. N. J. ~; MALTHA,A. 1977. Shear zones fractured rocks. AAPG Memoir, 16, 83-106.
in granular material, an experimental study of their STOW, D. A. V. 1986. Deep clastic seas. In: READING,H.
structure and mechanical genesis. Rock Mechanics, G. (ed.) Sedimentary Environments and Facies.
9, 95-144. Blackwell, Oxford, 399-444.
MARRETT, R. & ALLMENDINGER,R. W. 1991. Estimates SVERDRt:P, E. & BJORLYKKE, K. 1992. Small faults in
of strain due to brittle faulting. Sampling of fault sandstones from Spitsbergen and Haltenbanken.
populations. Journal of Structural Geology, 13, A study of diagenetic and deformational structures
735-738 and their relation to fluid flow. In: LARSEN,R. M.,
MARTINSEN, O. ~i; BAKKEN, B. 1990. Extensional and BREKKE, H., LARSEN, B. T. & TALLERAAS,E. (eds)
compressional slumps and slides in the Namurian Structural Modelling and its Application to Petro-
of County Clare, Ireland. Journal of the Geological leum Geology. Norwegian Petroleum Society
Society of London, 147, 153-164. Special Publication, 1, 507-517.
NARR, W. 8~; CURR1E, J. B. 1982. Origin of fracture TILLMAN, J. E. 1983. Exploration for reservoirs with
porosity - Example from the Altamont Field, fractured enhanced permeability. Oil & Gas
Utah. AAPG Bulletin, 66, 1231-1247. Journal, 81, 165-180.
- - & SUPPE, J. 1991. Joint spacing in sedimentary - - & BARNES,H. L. 1983. Deciphering fracturing and
rocks. Journal of Structural Geology, 13, 63-72 fluid migration histories in the Northern Appala-
NELSON, R. A. 1981. Significance of fracture sets chian Basin. AAPG Bulletin, 67, 692-705.
associated with stylolite zones. AAPG Bulletin, VERSTEEG, J. K. & MORRIS, W. A. 1994. Pitfalls and
65, 2417-2425. procedures for determining fabric orientations in
NYSTUEN, J. P. 1989. Rules and recommendations for non-oriented bore core. Journal of Structural
naming of geological units in Norway. Norsk Geology, 16, 283-286.
Geologisk Tidsskrift, 69, Suppl. 2. VlSSER, J. N. J., COLLISION,W. P. & TEREBLANCHE,J. C.
ORD, D. M., CLEMMEY,H. & LEEDER,M. R. 1988. Inter- 1984. The origin of soft-sediment structures in
action between faulting and sedimentation during Permo-Carboniferous glacial and proglacial beds,
Dinantian extension of the Solway basin, SW South Africa. Journal of Sedimentary Petrology,
Scotland. Journal of the Geological Society of 54, 1183-1196.
London, 145, 149-259. VOIGHT, B. & ST. PIERRE, B. H. P. 1974. Stress history
OWEN, G. 1987. Deformation processes in unconsoli- and rock stress. In: Advances in Rock Mechanics:
dated sands. In: JONES, M. E. & PRESTON,R. M. F. Proceedings of the 3rd Congress, International
(eds) Deformation of Sediments and Sedimentary Society of Rock Mechanics, 2, 580-582.
 Downloaded from http://sp.lyellcollection.org/ at University of Birmingham on June 10, 2015

64 R. H. GABRIELSEN ET AL.

WALSH, J. J. & WATTERSON,J. 1987. Distribution of WATTS, N. L. 1983. Microfractures in chalks from the
cumulative displacement and seismic slip on a Albuskjelt Field, Norwegian sector, North Sea.
single normal fault surface. Journal of Structural Possible origin and distribution. AAPG Bulletin,
Geology', 9, 1039-1046. 67, 201-234.
- -& - - 1 9 8 8 . Analysis of relationship between WESTAWAY,R. 1994. Quantitative analysis of popula-
displacements and dimensions of faults. Journal tions of small faults. Journal of Structural
of Structural Geology, 10, 239-247. Geology, 16, 1259-1273.
- & - - 1 9 8 9 . Displacement gradients on fault sur-
- WILKE, S., GvYON, E. & DE MARS~LY, G. 1985. Water
faces. Journal of Structural Geology, 11,307-316. penetration through fractured rocks. Test of a
, -- & YIELDING, G. 1991. The importance of tridimensional percolation description. Mathema-
small-scale faulting in regional extension. Nature, tical Geology, 17, 17-27.
351,391-393. WOJXAL, S. F. 1994. Fault scaling and the temporal
WATTERSON,J. 1986; Fault dimensions, displacements evolution of fault systems. Journal of Structural
and growth. Pure and Applied Geophysics, 124, Geology, 16, 603-612.
365-373.