Faults and their potential influence on fluid flow

Nicol, A.
1
, Seebeck, H.
1
, Hemmings-Sykes, S.
1,2
, Ilg, B.
3
, Childs, C.
4
,
Walsh, J.
4

1. GNS Science, Lower Hutt, New Zealand.
2.

Victoria University of Wellington, Wellington, New Zealand
3. New Zealand Petroleum & Minerals, Wellington, New Zealand
4. Fault Analysis Group, School of Geological Sciences, University College Dublin, Dublin 4,
Ireland.

Email: a.nicol@gns.cri.nz
Abstract
Faults have the potential to be both barriers to lateral flow and conduits for vertical
movement of fluids. We examine where and why faults locally enhance the flow of fluids
using mainly outcrops, tunnel excavations and seismic reflection lines. Outcrop studies of
normal faults within the Mount Messenger Formation reveal that the thickness of fault zones
(containing clay-rich fault rock and fractures) generally increases with displacement and
varies by several orders of magnitude on individual faults. Elevated densities of small-scale
faults, and the greatest fault zone widths, typically occur at irregularities on fault surfaces
(e.g., relays, bends and fault intersections). Data from ground water flow in tunnels and gas
chimneys imaged in seismic reflection lines show that for low permeability rocks (e.g., seals)
fluid flow is primarily achieved via fault zones, that flow rates are greatest on larger faults
and that migration of fluids is locally enhanced along fault irregularities (i.e. the sites where
the highest densities of small-scale faults occur). Geomechanical analysis of faults does not
appear to be a reliable predictor of gas chimney locations and migration pathways on
individual faults, perhaps because stress conditions near faults can depart significantly from
the regional stress field used in these analyses.

Keywords: fault zones, fluid flow, gas chimneys, geomechanical modelling

Introduction
In many sedimentary basins it has long been postulated that clay-rich fault rock and
associated small-faults/fractures strongly influence the sub-surface migration and
accumulation of hydrocarbons (e.g., Wade 1913; Illing, 1942; Neglia, 1979; Faulkner et al.,
2010). Individual faults can be both barriers to lateral flow and conduits for vertical movement
of fluids. The potential for faults to impede lateral flow of fluids within reservoir units has been
extensively studied (e.g., Yielding et al., 1997; Manzocchi et al., 1999, 2010). Faults can also
act as conduits to fluid migration enhancing their up-sequence flow and influencing the ability
of traps to retain oil or gas columns over geological timescales. Up-sequence fluid flow along
faults is thought to occur in the Taranaki Basin, where petroleum generated from Late
Cretaceous to Eocene source rocks have the ability to migrate through a thick (1-3 km)
Oligocene to Miocene mudstone-rich sequence, which has implications for the locations of oil
and gas fields. It is therefore important to determine where and under what circumstances
faults (and fractures) enhance the movement of petroleum fluids.

In this paper we utilise outcrop, tunnel and seismic reflection data (2D & 3D) from the
Taranaki Basin and the North Island of New Zealand to describe the structure of normal fault
zones, examine where and why fluid flow is focused within these fault zones and test the
predictive power of geomechanical techniques for identifying faults or parts of a fault that
may locally enhance up-sequence flow. Our analysis supports the view that channelized flow
of gas and water often occurs within fault zones. Geomechanical techniques (i.e. Slip
Tendency, Dilation Tendency and Fracture Stability) proved an unreliable means of
predicting the locations of channelized flow. Qualitative analysis of the data suggest that in
many cases fluid flow is observed close to geometric complexities on fault surfaces (e.g.,
fault relays, bends and intersections), where the densities of small-scale faults is likely to be
highest.

Fault Zone Structure
Faults typically comprise zones that accommodate heterogeneous shear strains (e.g., Caine
et al., 1996; Childs et al., 2009; Faulkner et al., 2010). The highest shear displacements are
focused within (or bounding) fault rock that typically contains fine-grained fault gouge and
breccia (products of crushing and intense fracturing). Fault rock in the Mount Messenger
Formation varies in content and dimensions depending on the host rock, fault-zone structure
and total displacement (Childs et al., 2007, 2009; Nicol et al., 2013)(Fig.1a). Fault rock is
generally (but not always) accompanied by fracturing and small-scale faulting and
collectively these structures form the fault zone (Fig. 1a). Fault zones typically comprise an
anastomosing system of intersecting fault segments which bound lenses of variably
fractured host rock. The thickness of fault zones shows a broad positive relationship with
displacement, however, there is significant heterogeneity of fault zone thickness for a given
displacement. The thicknesses of both fault zones and fault rock can change on individual
structures by an order of magnitude or more over distances of metres (e.g., Childs et al.,
2009; Nicol et al., 2013). For example, a recently studied normal fault with ~6-8 m of
displacement displayed a factor of five change in fault-zone thickness over 5 m horizontal
distance (10 cm to 50 cm) and a factor of 25 change over 40 m (10 cm to 2.5 m). These
changes indicate that the spread in fault-zone thickness for a given displacement on all
sampled faults may be approximately replicated by the variability on individual faults (Fig.
1b). The 1-2.5 orders of magnitude range of fault zone thickness for a given displacement on
all faults in Fig. 1b reflects the fact that fault zones comprise a combination of single planar
fault surfaces and multiple slip surfaces with variable amounts of associated secondary
fracturing. The widest fault zones are generally associated with fault complexities, including
segmentation and segment boundaries (e.g., relays), fault bending, fault intersections and
fault terminations (Childs et al., 2009; Nicol et al., 2013). In many cases these zones of fault
complexity are sites with increased numbers of fractures and are considered by many to be
zones of likely elevated fault-zone permeability (e.g., Gartrell et al., 2004; Ilg et al., 2012).
The role of these zones of fracturing on fluid flow and petroleum fluid migration are
discussed in the following sections.

Fluid Flow along Faults
Many quantitative and qualitative models have been proposed to explain where and why
fluid flow occurs along faults. Historically, one of the most significant impediments to testing
these models has been a lack of empirical data on fluid flow in areas where faulting is well
understood. Here we present the results of three studies in which data are available for both

a

b
faults and fluid flow (Hemmings-Sykes, 2012; Ilg et al., 2012; Seebeck et al., 2014). These
studies confirm that fluid flow (water or gas) occurs within fault zones and is highly
channelized and heterogeneous. They also permit testing of geomechanical methods for
predicting the locations of elevated fluid flow on fault surfaces.



Fig. 1. a) Schematic diagram showing a fault with definitions of fault rock (also referred to as fault
core) and fault zone illustrated (from Childs et al., 2009). b) Plot of fault zone width versus
displacement for normal faults in the Mount Messenger Formation on the north Taranaki coast.
Vertical bars show variations in fault zone thickness for individual faults.

Water flow into tunnels through fault zones has been analysed to determine the relationships
between faulting and fluid flow (locations and rates) (Seebeck et al., 2014). The dataset
includes information on fault geometries and their spatial relationships to water flowing into
tunnels located along the margins of the Taupo Rift, New Zealand (Fig. 2). Faults and water
flow data from engineering geological logs and reports have been used to examine the
factors influencing the rates and localisation of groundwater flows in relation to fault zone
architecture and connectivity of the fault-fracture network. As faults locally promote fluid
migration through seal and reservoir rocks our analysis may have implications for the
exploration and production of petroleum fluids (Manzocchi et al., 2010; Ilg et al., 2012). In
the tunnels localised ground water inflows occur almost exclusively (~90%) within, and
immediately adjacent to, fault zones. Fault zones in contrasting lithologies comprise fault
rock, small-scale faults, and fractures with thicknesses of 0.01 to ~110 m approximating
power-law distributions and bulk permeabilities of 10
-9
-10
-12
m
2
(Fig. 2a). Variability in fault-
zone structure results in highly heterogeneous flow rates and channelised flow. Within
basement rocks ~80% of the flow rate occurs from fault zones 10 m wide, with ~30% of the
total localised flow rate originating from a single fault zone (i.e. consistent with the golden
fracture concept) (Fig. 2a). No simple relationships are found between flow rates and either
fault strike or hydraulic head (Fig. 2b), with 50% of fault zones in any given orientation
flowing. A general positive relationship does however exist between fault zone thickness and
maximum flow rate. Higher flow rates from larger fault zones may arise because these
structures have greater dimensions and are more likely (than smaller faults) to be connected
to other faults in the system and the ground surface. These results suggest that faults
passing through the reservoir and seal in hydrocarbon systems are most likely to promote oil
and/or gas migration through the seal. However, it is clear that not all faults, or all parts of
faults, connecting the reservoir to strata above the seal will compromise petroleum traps
(see next section for further discussion).









Fig. 2: Fluid flow and fault data from tunnels in the Taupo Rift (Seebeck et al., 2014). a) Proportion of
localised flow carried by basement fault zones. Illustrates ~80% of total cumulative flow rate occur on
<3% of basement faults which typically have zones >10 m wide. Note, one fault accounts for ~30% of
total flow rate entering basement tunnels. b) Basement fault zones 1 m wide with and without water
flow. The flow and no flow fault populations have similar azimuth distributions suggesting that fault
strike may not be a strong determinant for which faults will accommodate water flow. Grey polygon
shows the approximate optimal fault azimuth for flow determined from geomechanical techniques
for a maximum horizontal shortening trending at ~060 (red line).
The role of normal faults in up-sequence flow of fluid has also been examined using 2D and
3D seismic-reflection data from the southern Taranaki Basin, New Zealand (Hemmings-
Sykes, 2012; Ilg et al., 2012). The spatial distributions of late-stage normal faults, gas
chimneys, thickness of the Oligocene mudstone-rich seal (Otaraoa Formation) and modelled
petroleum expulsion volumes have been compared in our studies. Gas chimneys are most
common above Cretaceous source rocks modelled to have expelled hydrocarbons. Most
(~70%) of the observed gas chimneys follow, and/or are rooted in, late-stage normal faults.
These faults are the primary seal bypass mechanism for hydrocarbons where they displace
the seal (or intersect faults that displace the seal) and the seal is thick (e.g., >~340 m). Gas
flow up along faults in low permeability mudstones (<1 mD) is channelized with steep
chimneys often occurring close to fault tips and relay ramps. In these cases gas flow may be
focused by the presence of high densities of open fractures (as is observed in outcrop),
locally elevating up-sequence bulk permeabilities in the seal.

Gas chimneys and normal faults imaged in a 3D seismic reflection volume provide a means
of testing the ability of geomechanical models to predict the locations of up-fault hydrocarbon
migration. The use of geomechanical methods for predicting leakage risk is only applicable
as a first-order estimate of which fault sets present the highest risk of up-dip migration
(Hemmings-Sykes, 2012). Slip Tendency and Dilation Tendency were able to differentiate
fault orientations in the Kupe Area most at risk of leakage and both indicated higher risk of
leakage for the fault set striking parallel to the maximum horizontal stress (approximately
northeast-southwest). In contrast, Fracture Stability was not able to differentiate fault sets
most at risk of leakage in the Kupe Area. The ability of geomechanical modelling methods to
Proportion total flow
F
a
u
l
t

z
o
n
e

w
i
d
t
h

(
m
)

F
r
e
q
u
e
n
c
y

b a
Azimuth
locate zones on fault surfaces with a high risk of up-sequence gas flow appears to be
limited. There was no statistical difference in leakage risk between chimney and non-
chimney locations on fault surfaces when applying Slip Tendency, Dilation Tendency and
Fracture Stability (Fig. 3).















Fig. 3: Oblique view of fault surfaces from the Kupe region with colours showing contours of slip
tendency on the fault surfaces (Hemmings-Sykes, 2012). White line polygons indicate the outline of
gas chimneys mapped on the fault surfaces. The chimneys are of limited extent and their locations do
not seem to be related to zones of higher Slip Tendency. Refer to Hemmings-Sykes (2012) for further
details.

The inability of geomechanical analyses to predict the locations of chimneys might arise
because: i) these techniques do not include small-scale information on the local stress
tensor and/or on the locations of open interconnected fractures and, ii) they do not take
account of whether individual faults breach the seal and are likely to intersect a petroleum
source. The poor relationship between fault orientation and fluid flow rates in the tunnel
dataset supports the view that in some cases geomechanical methods may not provide a
robust means of assigning risk of up-fault fluid flow for fault sets with widely varying strike
(Seebeck et al., 2014). In our experience geomechanical modelling should be used with
caution and does not seem to be a reliable predictor of fluid flow in the New Zealand
examples studied.

Conclusions
Analysis of outcrop and seismic-reflection studies of normal faults and fluid flow in New
Zealand support a number of conclusions. These are;
1) The thickness of fault zones (containing clay-rich fault rock and fractures) generally
increases with displacement and varies by several orders of magnitude on individual
faults. Elevated densities of small-scale faults (and greatest fault zone widths) typically
occur at irregularities on fault surfaces (e.g., relays, bends and fault intersections).
2) Ground water flow in tunnels and gas chimneys imaged on seismic reflection lines show
that in low permeability rocks (e.g., seals) fluid flow is primarily achieved via fault zones
and that flow rates are greatest on larger faults.
3) Geomechanical analysis does not appear to be a reliable predictor of where channelized
flow will occur on individual faults.
4) Up-fault migration is locally enhanced at fault irregularities (i.e. the sites where the
highest densities of small-scale faults occur).

References
Caine, J.S., Evans, J.P., Forster, C.B., 1996. Fault zone architecture and permeability
structure. Geology 24(11), 1025-1028.
Childs, C., Walsh, J. J., Manzocchi, T., Strand, J., Nicol, A., Tomasso, M., Schopfer, M.,
Aplin, A. 2007. Definition of a fault permeability predictor from outcrop studies of a
faulted turbidite sequence, Taranaki, New Zealand. In: Structurally Complex
Reservoirs. Geological Society of London, Special Publication 292, 235-258.
Childs, C., Walsh, J.J., Manzocchi, T., Bonson, C., Nicol, A., Schpfer, M.P.J., 2009. A
geometric model of fault zone and fault rock thickness variations. Journal of
Structural Geology 31, 117-127.
Faulkner, D.R., Jackson, C.A.L., Lunn, R.J., Schlische, R.W., Shipton, Z.K., Wibberley,
C.A.J., Withjack, M.O. 2010. A review of recent developments concerning the
structure, mechanics and fluid flow properties of fault zones. Journal of Structural
Geology 32, 1557-1575.
Hemmings-Sykes, S., 2012. The influence of faulting on hydrocarbon migration in the Kupe
area, south Taranaki Basin, New Zealand. Unpublished MSc, Victoria University of
Wellington, p. 224.
Gartrell, A., Zhang, Y., Lisk, M., Dewhurst, D. 2004. Fault intersections as crictical
hydrocarbon leakage zones: integrated field study and numerical modelling of an
example from the Timor Sea, Australia. Marine & Petroleum Geology 23, 1165-1179.
Ilg, B.R., Hemmings-Sykes, S., Nicol, A., Baur, J., Fohrmann, M., Funnell, R., Milner, M.,
2012. Normal faults and gas migration in an active plate boundary, southern Taranaki
Basin, offshore New Zealand. Amercian Association of Petroleum Geology Bulletin,
96(9), 1733-1756.
Illing, V.C., 1942. Geology applied to petroleum. Proceedings of the Geologists' Association
53(3-4), 156-187.
Manzocchi, T., Walsh, J.J., Nell, P., Yielding, G., 1999. Fault transmissibility multipliers for
flow simulation models. Petroleum Geoscience 5(1), 53-63.
Manzocchi, T., Childs, C., Walsh, J.J., 2010. Faults and fault properties in hydrocarbon flow
models. Geofluids 10(1-2), 94-113.
Nicol, A., Childs, C., Walsh, J., Schafer, K., 2013. A geometric model for the formation of
deformation bands. Journal of Structural Geology 55, 21-33.
Neglia, S., 1979. Migration of fluids in sedimentary basins. American Association of
Petroleum Geology Bulletin, 63(4), 573-597.
Seebeck, H., Nicol, A., Walsh, J.J., Childs, C., Beetham, R.D., Pettinga, J., 2014. Fluid flow
in fault zones from an active rift. Journal of Structural Geology 62, 52-64.
Wade, A., 1913. The natural history of petroleum. Proceedings of the Geologists
Association 24(1) 1-13, IN1-IN3.
Yielding, G., Freeman, B., Needham, D.T., 1997. Quantitative fault seal prediction. American
Association of Petroleum Geology Bulletin 81(6), 897-917.