Evolution of permeability and microstructure of experimentally-created shear zones in Neogene siliceous mudstones from Horonobe, Japan

Journal of Structural Geology 60 (2014) 46e54

Contents lists available at ScienceDirect

Journal of Structural Geology journal homepage: www.elsevier.com/locate/jsg

Evolution of permeability and microstructure of experimentally-created shear zones in Neogene siliceous mudstones from Horonobe, Japan Shin-ichi Uehara a, *, Miki Takahashi b a

Department of Environmental Science, Faculty of Science, Toho University, 2-2-1 Miyama, Funabashi, Chiba 274-8510, Japan Active Fault and Earthquake Research Center, Geological Survey of Japan, The National Institute of Advanced Industrial Science and Technology (AIST), Central 7, 1-1-1 Higashi, Tsukuba, Ibaraki 305-8567, Japan b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 7 April 2012 Received in revised form 1 December 2013 Accepted 9 December 2013 Available online 31 December 2013

We report experimental measurements of bulk permeability changes due to a shear zone that is induced in siliceous mudstones collected from the Koetoi and Wakkanai Formations, northern Hokkaido, which are known to show different relationships between fault/fracture distribution and groundwater flow. We evaluate distributions of volumetric deformation in the induced shear zones by using micro-focus X-ray computed tomography. Measured permeability evolution while achieving the peak axial stress for specimens differed for the samples of the two formations. Permeability did not change obviously during shear for the Koetoi Fm. specimens, but in the Wakkanai Fm. specimens, the bulk permeability increased by a factor of 2.5 after reaching the peak stress. The difference in permeability change in these experiments can explain the differences in relationships between in situ groundwater flow and fracture distribution for the two formations. Analyses of the X-ray images reveal that this difference should reflect the differences of the volumetric deformation in the induced shear zones. Pore collapse occurred in the shear zone in the Koetoi Fm. specimen, which leads to porosity reduction, whereas fracture damages developed in the Wakkanai Fm. specimen, increasing porosity. These differences in the microstructure may reflect differences in yielding criteria for these host rocks. Ó 2013 Elsevier Ltd. All rights reserved.

Keywords: Fault Permeability Mudstone Rock mechanics Laboratory experiment Micro-focus X-ray CT

1. Introduction Characterizing the fluid flow properties of mudstones or siltstones is important when considering topics such as fluid flow in sedimentary basins (Bjorlykke, 1993) and accretionary prism (Behrmann, 1991; Brown et al., 1994), hydrocarbon accumulations (Fisher and Knipe, 1998) or geological sequestration of industrially-produced greenhouse gases (Cappa and Rutqvist, 2011). Fault zones play an important role in controlling the fluid flow in mudstones (Ingram and Urai, 1999). Fluid-flow properties of fault zones in mudstone depend on several factors such as mechanical properties of the host rock. For instance, in general, dilatant and permeable shear fractures form in rocks during deformation under small confining pressure and/or in relatively strong or brittle rocks, while compactive shear fractures occur during deformation of rocks at large confining pressure

* Corresponding author. Tel.: þ81 (0)47 472 5596; fax: þ81 (0)47 472 7039. E-mail addresses: [email protected], [email protected] (S.-i. Uehara), [email protected] (M. Takahashi). 0191-8141/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jsg.2013.12.003

and/or during deformation of weak or ductile rocks (Ingram and Urai, 1999). To fully understand how such factors affect flow properties of fault zones, it is important to study in detail the fluid flow properties of fault zones in mudstone, including in situ measurements, characterization of fault-zone structure, and the geological setting. The Neogene sedimentary basin in the Horonobe area, northern Hokkaido (Fig. 1), is a suitable place for this type of study. The Japan Atomic Energy Agency (JAEA) established Horonobe Underground Research Center in 2001 and has conducted very detailed geological and hydraulic studies, including drilling 11 boreholes and completing 66 in situ measurements of hydraulic conductivity (Horonobe Underground Research Center, 2004; Niizato et al., 2007; Ota et al., 2007; Kurikami, 2007; Kurikami et al., 2008; Funaki et al., 2009; Ishii et al., 2010, 2011; Uehara et al., 2012). At this site, the Koetoi Fm. (late Miocene to Pliocene, diatomaceous mudstone) and the Wakkanai Fm. (middle to late Miocene, siliceous mudstone) differ in their relationships between fault/ fracture distribution and groundwater flow. Funaki et al. (2009) detected fracture distributions and locations of fast water flow in a drill hole at this site from core logging, acoustic televiewer and

S. Uehara, M. Takahashi / Journal of Structural Geology 60 (2014) 46e54

47

Fig. 1. (a), (b) Location maps (upper-left). (c) Geological map of the study area (middle-left), a open circle is bore hole site. (d) Geological cross section (lower-left) along the AeA0 line in the geological map (compiled from previous studies, e.g., Ota et al., 2007). Borehole HDB-10 projected into profile as a black line showing penetration depth. (e) Sedimentary formations in the borehole of HDB-10 (right).

fluid electric conductivity logging. They determined that in the Wakkanai Fm., the fracture distribution strongly correlated to locations of fast groundwater flow into the drill hole, whereas a weak correlation existed for the Koetoi Fm. This difference in flow properties can be explained by differences in deformation, i.e., brittle or ductile, and can be characterized using a brittleness index (BRI) from laboratory tests, which is the ratio of unconfined compressive strength to effective vertical stress (Ishii et al., 2011). To further understand the relationship between deformation behavior and the flow properties at the Horonobe area, and to reveal factors controlling the flow properties, we conducted laboratory experiments to measure permeability evolution during shear deformation of these Horonobe mudstones. In addition, we imaged the pore distributions of the induced shear zones by using microfocus X-ray computed tomography (CT) to investigate microstructural textures and their relationship to flow and mechanical properties.

borehole penetrated the Koetoi and Wakkanai Fms. The Koetoi Fm. consists mainly of diatomaceous mudstone. The Wakkanai Fm. consists mainly of siliceous mudstone and is more strongly cemented than the overlying the Koetoi Fm. Fukusawa (1985)

2. Specimen preparation and experimental method The samples were taken from drill cores of the borehole HDB10 drilled by JAEA (Fig. 1; e.g., Ota et al., 2007). The HDB-10

Fig. 2. Schematic illustration and photograph of specimen geometry. Note that the mudstone specimen is intact and there is no pre-cut surfaces prior to the experiments.

48

S. Uehara, M. Takahashi / Journal of Structural Geology 60 (2014) 46e54

-16

20 10

15

Axial Stress

10

-18

10

Permeability

5

-19

0

1

2

3

4

Axial Displacement (mm)

Kt263m

20 -17

10

15

Axial Stress

10

Permeability

-18

10

5 0

10 5

b

10

Permeability (m2)

-17

0

-19

0

1

2

3

4

Axial Displacement (mm)

10 5

-16

25

10

c

Wk385m

20

Axial Stress

15 10

-18

10

Permeability

5 0

-17

10

Permeability (m2)

Axial Diff. Stress (MPa)

Kt43m

-16

25

Axial Diff. Stress (MPa)

a

10

Permeability (m2)

Axial Diff. Stress (MPa)

25

-19

0

1

2

3

4

Axial Displacement (mm)

10 5

Fig. 3. Axial differential stress and permeability vs. axial displacement for (a) Kt43m, (b) Kt263m, and (c) Wk385m.

suggested that the Koetoi and Wakkanai Fms. both consist mainly of diatomaceous mudstone, but that their properties differ from each other because the phase of amorphous silica changes from opal A to opal CT at the boundary between these formations. Opal CT is the intermediate phase at silica-phase transformations from biogenic opal (opal A) to quartz. The opal A to opal CT phase transformation is a solution-reprecipitation process, and decrease in porosity and pore size of Neogene siliceoucs rocks corresponding to this phase transformation has been reported (Tada and Iijima, 1983). Porosities for the Koetoi and Wakkanai Fms. mudstones are 0.55e0.65 and 0.35e0.45, respectively (e.g., Ota et al., 2007). The boundary between the Koetoi and Wakkanai Fms. is located at a depth of 290 m in the borehole. We collected samples from cores at a depth of 43 m (Kt43m, the Koetoi Fm.), 263 m (Kt263m, the Koetoi Fm.), and 385 m (Wk385m, the Wakkanai Fm.) (Fig. 1). We prepared 20 mm diameter and length, cylindrical specimens from the samples by re-coring in the laboratory using a diamond core bit. Experiments were performed using a gas medium, highpressure, high-temperature triaxial testing machine at the National Institute of Advanced Industrial Science and Technology

(AIST), Japan. The details of this apparatus have been described by Masuda et al. (2002) and Takahashi et al. (2007, 2009). To directly investigate permeability evolution along experimentally-induced shear zone, we adopted a sample assembly similar to Takahashi (2003) (Fig. 2). We put an intact mudstone cylindrical specimen between short cylinders of Berea sandstone. The sandstone cylinders have sawcut surfaces oriented at 30 to the axis (Fig. 2), which induces a shear zone in the intact mudstone specimen when differential stress is applied to the assembly. Note that the direction of s1 is vertical in these experiments and therefore a normal fault is induced, whereas the Horonobe area is in reverse-fault stress regime. Berea sandstone has a permeability of 1013e1012 m2 under confining pressures of less than 10 MPa (e.g., Zhu and Wong, 1997), which is more than four orders of magnitude larger than the permeability of the mudstone specimens used in this study (1017e1018 m2). As differential stress increases, in general, pores are likely to collapse in porous sandstones, decreasing permeability. However, the differential stress conditions during our deformation experiments conducted are insufficient for pore collapse in the Berea sandstone cylinders (Wong et al., 1997). Further, if the

S. Uehara, M. Takahashi / Journal of Structural Geology 60 (2014) 46e54

49

can be determined. The peak-to-peak amplitude of the upstream pressure change was 0.8 MPa. 3. Results 3.1. Shear stress and permeability measurements

Fig. 4. Mohr circles and stress conditions (open circles) at the peak axial stress, and Mohr-Coulomb criteria for peak stresses of the Koetoi and Wakkanai Fm. mudstone based on Matsui and Yamamoto (2004) and Sanada et al. (2009). Stress conditions were estimated by assuming that shear stresses at the mudstone and pre-cut surfaces of sandstone are the same. Sanada et al. (2009) also reported that the Koetoi Fm. mudstone yields under hydrostatic stress condition (s1 ¼ s2 ¼ s3) when a confining pressure is approximately 10 MPa (closed circle), which might reflect pore collapse.

Axial differential stress and permeability changed during shear deformation of the mudstone specimens (Fig. 3). Wk385m showed a marked reduction in axial stress after an axial displacement of 0.6 mm (Fig. 3c). For Kt43m, on the other hand, axial stress measurements did not reduce after reaching the peak axial stress (Fig. 3a). Kt263m did also show a clear reduction in the axial stress, but the magnitude was smaller than for Wk385m (Fig. 3b). The stress conditions at the peak stress were similar to those expected from Mohr-Coulomb criteria obtained from previous studies with specimens from Horonobe region (Matsui and Yamamoto, 2004; Sanada et al., 2009)(Fig. 4). Permeability changes after reaching the peak axial stress in the Koetoi Fm. specimens were also clearly different from that of the Wakkanai Fm. specimen, although the permeability values were similar to each other before the axial loading. For Kt43m and Kt263m, permeability did not change during loading up to peak stress, or when sheared afterward (Fig. 3a and b). For Wk385m, by contrast, the permeability clearly increased after reaching the peak stress (Fig. 3c), by a factor of approximately 2.5, although the permeability gradually decreased after that increase. 3.2. Microstructure observations by micro-focus X-ray CT

permeability of the sandstone blocks decreases due to concentration of deformation such as pore collapse where blocks contact each other, the effects on the measured bulk permeability should be small because the area of the surfaces that interrupt flow through the sample assembly is limited. Therefore, when we estimate permeability of the mudstone specimen from the laboratory experiments, we ignore the role of the sandstone cylinders. Each sample assembly was placed between perforated metal spacers and jacketed with the upper and lower pistons using polyolefin heat-shrink tubing, and installed in the pressure vessel. The sample assembly was loaded within the triaxial apparatus and deformed under a constant axial displacement rate of 0.2 mm/s. Confining and pore pressures were held constant at 8.3 MPa and 4.9 MPa throughout each permeability measurement, respectively. These conditions correspond to a depth of approximately 500 m, assuming an average rock density of 1.7 g/cm3, which is consistent with the Koetoi Fm. mudstone, and that pore pressure is equal to hydrostatic pressure. Axial load was increased independently to induce a shear zone in the initially-intact mudstone specimens. We used pure water as the pore fluid. Mudstones can react strongly with fluids whose compositions are different from those originally in the rocks (e.g., Müller, 1967). We cannot quantitatively assess this effect for the present study, but this effect may be small because we saturated the mudstone samples with pure water just prior to the experiments. All measurements were performed at room temperature conditions. We use the pore-pressure oscillation technique to determine permeability (Kranz et al., 1990; Fischer and Paterson, 1992). In this method, a sinusoidally varying pressure is generated in the upstream reservoir by means of a servo-controlled piston. Pore pressure at the downstream reservoir also varies sinusoidally, but the amplitude is smaller than that at the upstream end, and the phase shifts. From the amplitude ratio and phase shift of the upstream and downstream pressure changes, the bulk permeability

To investigate the relationship between deformation-related textural changes and the measured properties, we observed the final microstructure of the induced shear zone by using micro-focus X-ray Computed Tomography (CT) (TOSCANER-32251mhd) at Tokyo Metropolitan Industrial Technology Research Institute (Fig. 5). Xray CT scanner determines a spatial distribution of X-ray attenuation values within an object. The images in Fig. 5 show spatial distributions in slices through specimens of non-dimensional values proportional to X-ray attenuation values, which are called “X-ray CT values” in this paper. The X-ray CT value corresponds to density, and when the mineral assembly is the same, the value is negatively correlated with averaged porosity of the pixel (e.g., Wellington and Vinegar, 1987). In the X-ray CT images, the induced shear zones show geometries similar to those in geomechanical simulations of shear zones induced in previous studies (e.g., Couples et al., 2007). Comparing Figs. 5a and c, the X-ray CT values for the shear zone of Kt43m are greater than the host rock, whereas the CT values for the shear zones of Wk385m are less than in the intact host. For Kt263m, the difference in the CT values between the shear zone and the intact host is not clear (Fig. 5b). Open cracks are observed, but these cracks may not be present under load (Bésuelle et al., 2006; Lenoir et al., 2007). These differences in values should relate to changes of permeability and axial differential stress during experimental deformation. To characterize the pore volume change in the induced shear zone, we introduce the parameter Df0 for each pixel, the ratio of pore volume change by deformation to volume of the solid (mineral) part of intact host, defined by

Df0 ¼

f  fint ; 1  fint

(1)

where f and fint are an averaged porosity of a pixel and an averaged porosity of intact host, respectively. The parameter Df0 reflects the

50

S. Uehara, M. Takahashi / Journal of Structural Geology 60 (2014) 46e54

Fig. 5. X-ray CT images of slices through specimens after shear deformation. Left slices are parallel to the axis of the specimen and perpendicular to shear band. Right slices are perpendicular to the axis with position in left images shown by arrows. (a) Kt43m, (b) Kt263m, and (c) Wk385m. The pixel size is 52.8, 72.1 and 53.1 mm, respectively. Note that greater CT value corresponds to greater density. The images were modified by an image processing software ImageJ (the National Institutes of Health) and Slice (Nakano et al., 2006).

S. Uehara, M. Takahashi / Journal of Structural Geology 60 (2014) 46e54

51

difference in volumetric strain from an average of intact host. If volume change of the solid part can be ignored, volumetric strain of a pixel and intact host equal the porosity change, f  f0 and fint  f0, respectively, where f0 is a porosity before deformation. Therefore, the difference in volumetric strain from an average of intact host is (f  f0)  (fint  f0) ¼ f  fint. The parameter Df0 is derived as follows, on the assumption that there is a prorated share of contributions from pore and solid part for CT value of a pixel, x (e.g., Keller, 1998):

x ¼ f$xa þ ð1  fÞxm ;

(2)

where xa and xm are averaged CT values of air and a mineral part, respectively. Following the same assumption, an averaged CT value of intact host is given by

xint ¼ fint $xa þ ð1  fint Þxm : From Eqs. (2) and (3), the parameter

f  fint x  xint ¼ ¼ Df0 : xa  xint 1  fint

(3)

Df0

is given by

(4)

For Kt43m (Fig. 6a), the width of the induced shear zone was approximately 1 mm, and minimum Df0 was approximately 0.15. This value of Df0 corresponds to a f that is approximately 0.5e0.6 when fint is assumed to be 0.55 to 0.65, which is derived from porosity based on previous studies (e.g., Ota et al., 2007). The shear zone was not clear at the center of the X-ray CT image. For Wk385m (Fig. 6c), the width of the induced shear zone was about 2 mm. Df0 was generally large near the sandstone at up to 0.2 to 0.35, which corresponds to f that is approximately 0.48e0.64 when fint is assumed to be 0.35 to 0.45 (e.g., Ota et al., 2007), and Df0 was small at the center of the specimen. The Df0 distributes asymmetrically with respect to the center of the induced shear zone for Wk385m (Fig. 6c). For Kt263m, Df’ was near 0 through the sample (Fig. 6b). 4. Discussion 4.1. Yielding criterion as a factor affecting shear-zone pore structure

Fig. 6. The profile of Df’ averaged in white rectangles in the left CT image, perpendicular to the induced shear zone. (a) Kt43m, (b) Kt263m, and (c) Wk385m. Dash lines at the right indicate the position where Df’ ¼ 0 for each profile where CT value equals the average value of the intact host part.

In this study, we conducted experimental measurements of bulk permeability changes in mudstones during shear deformation, and found differences in the permeability changes between Koetoi and Wakkanai Fm. mudstones. We believe on the basis of the prior work by Ingram and Urai (1999) and Sanada et al. (2009) that, in this setting, the yield criterion of a rock can be one of factors that affected pore structure within the induced shear zone, which should cause the differences in the permeability changes. Ingram and Urai (1999) suggested that shear fracture is compacted during deformation when the stress state approaches the yield criterion for pore collapse. Accepting that suggestion and considering that the Koetoi Fm. mudstone is reported to yield at a confining pressure of approximately 10 MPa for hydrostatic stress (s1 ¼ s2 ¼ s3), we infer that yielding was achieved by pore collapse (a closed circle in Fig. 4) (Sanada et al., 2009). In contrast, the Wakkanai Fm. mudstone is reported not to yield even at greater confining pressures up to 40 MPa (Sanada et al., 2009). Given this behavior, we also believe that the Wakkanai Fm. mudstone is not undergoing pore collapse as adjacent Koetoi Fm. mudstone is doing so. We support this pair of interpretations with the observations from our experimental results that a induced shear zone for Kt43m is denser than the intact rock part (Figs. 5a and 6a), while that for Wk385m is less dense than the intact rock part (Figs. 5c and 6c).

52

S. Uehara, M. Takahashi / Journal of Structural Geology 60 (2014) 46e54

4.2. Porosity changes and fluid flow changes

104

The differences in permeability change between the Koetoi and Wakkanai Fm. mudstone specimens observed in the laboratory experiments can qualitatively explain how the deformation of these formations by shear zones can influence in situ groundwater flow, as suggested by Funaki et al. (2009). When the specimen Kt43m was deformed under the stress condition corresponding to a depth of several hundred meters, the induced shear zone was less porous than the intact rock part (Figs. 5a and 6a), and the permeability along the shear zone did not increase effectively (Fig. 3a). While in the case of the specimen Wk385m, the induced shear zone was more porous than the intact rock part (Figs. 5c and 6c) and the permeability along the shear zone increased effectively (Fig. 3c). These results suggested that shear zones developed in the Koetoi and Wakkanai Fm. under the depth of several hundred meters have different pore structures and different features on permeability as follows: shear zones developed in the Wakkanai Fm. tend to be porous and permeable, and therefore the shear zones become effective flow paths, whereas shear zones developed in the Koetoi Fm. under the depth tend to be compacted and not permeable, and therefore the shear zones do not become effective flow paths.

103

Range of data from Childs et al. (2009):

Evaluation of a fault width is important because it influences fluid flow in the crust and the deformation behavior of the rock such as during earthquakes (Cain et al., 1996; Scholz, 2002). Fault width shows a positive correlation with fault displacement (Childs et al., 2009; Mitchell and Faulkner, 2009; Faulkner et al., 2010, 2011; Savage and Brodsky, 2011). Childs et al. (2009) have compiled data characterizing a fault width, such as a thickness of fault rock, fault zone or damage zone, as a function of shear displacement. We compared the width of the shear zone for Kt43m, Kt263m and Wk385m with the compilations for thicknesses of fault rock and deformation band, which characterize a thickness of a single fault and therefore may be comparable to the width of the shear zone observed in this study. The relationship between shear displacement and the width of the shear zone in the specimens is in the range of the compilation (Fig. 7). 4.4. Estimation of local permeability for an induced shear zone Local permeability of the induced shear zone is greater than the initial bulk permeability for Wk385m (Fig. 3c). We estimated the local permeability, kS, by assuming that (1) the increase in the bulk flow rate at a sudden decrease in stress, DQB, is equal to the flow rate through the induced shear zone, QS, (2) the induced shear zone is homogeneous, (3) the flow direction in the induced shear zone is parallel to the zone, and (4) the cross section of the shear zone normal to the flow direction is constant and equal to AS ¼ DS WS (Fig. 8 shows the definition of DS and WS). The assumption (3) is reasonable if kS is several times larger than the initial bulk permeability. The assumption (4) is not valid at the shear zone close to the sandstone (Fig. 5c), but the area is limited and therefore the effect on this analysis is probably small. The assumption (2) is not strictly valid, and effects of inhomogeneity of kS is discussed later. The flow rate just before a sudden decrease in stress, QB, is given by.

kB AB DPP m Lm

Thickness (m)

102 101

Data from this study: Kt43m Kt263m, Wk385m

100 10-1 10-2 10-3 10-4 10-5 10-3 10-2 10-1 100

101

102

103

104

105

106

Displacement (m) Fig. 7. A width of a fault as a function of displacement. Data from this study are plotted with a range of data compiled by Childs et al. (2009).

difference between the two sides of the specimen, and Lm is the length of the specimen (Fig. 8). The increase in the bulk flow DQB, or QS, is given by

4.3. The relation between a shear zone width and shear displacement

QB ¼

Fault rock Deformation band

DQB ¼

kS AS DPP m LS

(6)

where LS is the length along the flow rate direction of the shear zone, which is Lm/cos30 (Fig. 8b). The ratio of bulk permeability change, DkB, to kB is equal to DQB/QB, and therefore

DkB kB

¼

DQB QB

¼

kS AS cos30o kB AB

(7)

Finally, kS is given by

A 1 kS ¼ DkB $ B $ AS cos30o

(8)

With DkB ¼ 9.4  1019 m2 (from Fig. 3c), AS ¼ 40 mm2 (DS ¼ 2 mm and WS ¼ 20 mm), and AB ¼ 102p mm2, we find from Equation (8) that kS is approximately 8.5  1018 m2. This value is an order of magnitude greater than kB, which is approximately 6.0  1019 m2. Note that in reality, as mentioned above, the induced shear zone is not homogeneous, and therefore kS should not be unique in the shear zone. For example, Df’ in the shear zone is less at the center of the specimen compared with near the

(5)

where kB is the bulk permeability, AB is the cross section of the specimen, m is the viscosity of water, DPP is the pore pressure

Fig. 8. Schematic illustrations of (a) the sample assembly before deformation and (b) after deformation, showing definitions of Lm, LS, DS and WS.

S. Uehara, M. Takahashi / Journal of Structural Geology 60 (2014) 46e54

sandstone, and shear zone permeability at the center may be less than that near the sandstone. Therefore the shear zone permeability at the center and near the sandstone may be less and greater than 8.5  1018 m2, respectively. The bulk permeability change during axial displacement was not observed in the cases of Kt43m and Kt263m. Therefore the local permeability for the induced shear zone is similar to or less than the initial bulk permeability of the mudstone. Consequently, the approach used for Wk385m cannot be applied to the deformed samples from the Koetoi Fm. 4.5. Scaling issues on permeability for the Wakkanai Fm Flow properties of fracture systems are generally scale dependent. We considered microfractures in a specimen at a scale of a few centimeters, whereas borehole measurements will likely involve fluid flow through fractures of a larger scale. We have compared the results of this study with the in situ borehole measurements in the Wakkanai Fm. while considering scale-dependent behavior. Schulze-Makuch et al. (1999) described a scale dependency of hydraulic conductivity K of a geological medium by

K ¼ cðVÞm

(9)

where c is a parameter characteristic of the medium, V is the volume of tested material, and m is the exponent of the relationship. The exponent m is close to 1 for fracture flow media. When V for the experiments in this study and borehole measurements are an order of 105 and 10 m3, respectively, Equation (9) predicts that a permeability value of borehole-scale media is approximately six orders of magnitude larger than that of experiment-scale media for fracture flow media. Therefore, if experiment-scale shear zone has a permeability of an order of magnitude of 1017 m2, based on the estimation in Section 4.4, the predicted permeability of a boreholescale shear zone is an order of magnitude of 1011 m2. When we assume that the intact host is homogeneous porous media, of which permeability does not depend on a scale, the predicted permeability of a borehole-scale intact host rock is an order of magnitude of 1019e1018 m2, based on the results of permeability measurements before a shear zone developed (Fig. 3c). Therefore, the range of permeability of a borehole-scale media can be 1019 (when a intact host part is dominant) to 1011 m2 (when shear zones work effectively as flow paths). This value of permeability corresponds to hydraulic conductivity of 1012e104 m/sec. The hydraulic conductivity from in situ borehole measurements was approximately 1011e106 m/sec at the depth of 500 m (Kurikami et al., 2008), which is in this range. 5. Conclusions We investigated the characteristics of permeability change along a shear zone developed in Neogene siliceous mudstone by using laboratory experiments. The Wakkanai Fm. mudstone is more strongly cemented and has lower porosity than the Koetoi Fm. mudstone, which is because the phase of amorphous silica changes from opal A to opal CT at the boundary between these formations. Permeability did not change noticeably during shear for the Koetoi Fm. specimens. By contrast, for the Wakkanai Fm. specimens, the permeability clearly increased after reaching peak stress. This permeability change during shear deformation can explain the differences in groundwater flow and fracture distribution suggested by Funaki et al. (2009) that, in the Wakkanai Fm., the fracture distribution strongly correlated to locations of fast groundwater flow into the drill hole, whereas a weak correlation existed for the Koetoi Fm.

53

Estimates of sample density as a proxy for porosity by using micro-focus X-ray CT reveals that this behavior difference between the two types of mudstone should reflect a difference of the volumetric deformation in the induced shear zone. The induced shear zone for Kt43m (the Koetoi Fm.) was compacted, and by contrast, the shear zone for Wk385m (the Wakkanai Fm.) was dilatant. For Kt263m, the difference in the CT values between the shear zone and the intact host is not clear. This differences in volumetric deformation for the induced shear zones may reflect a difference in yield criterion of the mudstones because the stress conditions during shear deformation for Kt43m and Wk385m are consistent with the conditions suggested by Ingram and Urai (1999) in which an induced shear fracture is compacted and dilatant, respectively. Our results suggest that flow properties of fault in mudstone may be able to be estimated from stress condition (or depth) and a yield criterion of the host rock. So, to improve the characterization of fluid flow properties for deformed mudstone, it might be useful to understand how fluid flow properties in a fault depend on mechanical properties of host rock and stress conditions at the faulting process.

Acknowledgments This research is supported by the Ministry of Economy, Trade and Industry of Japan. We express our gratitude to Dr. Niizato, JAEA, for his efforts to prepare drilling core samples from HDB 10. William M. Dunne, Gary D. Couples and Tom M. Mitchell have significantly improved this manuscript, and are very much appreciated.

References Behrmann, J.H., 1991. Conditions for hydrofracture and the fluid permeability of accretionary wedges. Earth Planet. Sci. Lett. 107, 550e558. Bésuelle, P., Viggiani, G., Lenoir, N., Desrues, J., Bornert, M., 2006. X-ray micro CT for studying strain localization in clay rocks under triaxial compression. In: Desrues, J., Viggiani, G., Besuelle, P. (Eds.), Advances in X-ray Tomography for Geomaterials. ISTE Ltd, pp. 35e52. Bjorlykke, K., 1993. Fluid flow in sedimentary basin. Sediment. Geol. 86, 137e158. Brown, K.M., Bekins, B., Clennell, B., Dewhurst, D., Westbrook, G., 1994. Heterogeneous hydrofracture development and accretionary fault dynamics. Geology 22 (3), 259e262. Cain, J.S., Evans, J.P., Forster, C.B., 1996. Fault zone architecture and permeability structure. Geology 24 (11), 1025e1028. Cappa, F., Rutqvist, J., 2011. Modeling of coupled deformation and permeability evolution during fault reactivation induced by deep underground injection of CO2. Int. J. Greenh. Gas. Control 5, 336e346. Childs, C., Manzocchi, T., Walsh, J.J., Bonson, C.G., Nicol, A., Schöpfer, M.P.J., 2009. A geometric model of fault zone and fault rock thickness variations. J. Struct. Geol. 31, 117e127. Couples, G.D., Lewis, H., Workman, G.H., Higgs, N.G., 2007. Insights into the faulting process from numerical simulations of rock-layer bending. In: Lewis, H., Couples, G.D. (Eds.), The Relationship between Damage and Localization. The Geological Society of London, pp. 161e186. 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. J. Struct. Geol. 32, 1557e1575. Faulkner, D.R., Mitchell, T.M., Jensen, E., Cembrano, J., 2011. Scaling of fault damage zones with displacement and the implications for fault growth processes. J. Geophys. Res. 116, B05403. http://dx.doi.org/10.1029/2010JB007788. Fischer, G.J., Paterson, M.S., 1992. Measurement of permeability and storage capacity in rocks during deformation at high temperature and pressure. In: Evans, B., Wong, T.-f. (Eds.), Fault Mechanics and Transport Properties of Rocks. Academic Press, pp. 213e252. Fisher, Q.J., Knipe, R.J. (Eds.), 1998. Faulting, Fault Sealing and Fluid Flow in Hydrocarbon Reservoirs. Geological Society, London. Special Publication, 147. Fukusawa, H., 1985. Late Neogene formations in the Tempoku-Haboro region, Hokkaido, Japan estratigraphic reinvestigation of the “Wakkanai” and “Koetoi” formations-. J. Geol. Soc. Jpn. 91 (12), 833e849 (in Japanese with English abstract). Funaki, H., Ishii, E., Tokiwa, T., 2009. Evaluation of the role of fracture as the major water-conducting feature in Neogene sedimentary rocks. J. Jpn. Soc. Eng. Geol. 50, 238e247 (in Japanese with English abstract).

54

S. Uehara, M. Takahashi / Journal of Structural Geology 60 (2014) 46e54

Horonobe Underground Research Center, Japan Nuclear Cycle Development Institute, 2004. International Workshop on Horonobe Underground Research Laboratory Project, JNC TN5400 2004e004. Ingram, G.M., Urai, J.L., 1999. Top-seal leakage through faults and fractures: the role of mudrock properties. In: Aplin, A.C., Fleet, A.C., Macquaker, J.H.S. (Eds.), Muds and Mudstones: Physical and FluidFlow Properties, Geological Society of London, Special Publication, vol. 158, pp. 125e135. Ishii, E., Funaki, H., Tokiwa, T., Ota, K., 2010. Relationship between fault growth mechanism and permeability variations with depth of siliceous mudstones in northern Hokkaido, Japan. J. Struct. Geol. 32, 1792e1805. http://dx.doi.org/ 10.1016/j.jsg.2009.10.012. Ishii, E., Sanada, H., Funaki, H., Sugita, Y., Kurikami, H., 2011. The relationships among brittleness, deformation behavior, and transport properties in mudstones: an example from the Horonobe Underground Research Laboratory, Japan. J. Geophys. Res. 116, B09206. http://dx.doi.org/10.1029/2011JB008279. Keller, A., 1998. High resolution, non-destructive measurement and characterization of fracture apertures. Int. J. Rock Mech. Min. Sci. 35, 1037e1050. Kranz, R.L., Saltzman, J.S., Blacic, J.D., 1990. Hydraulic diffusivity measurements on laboratory rock samples using an oscillating pore pressure method. Int. J. Rock Mech. Min. Sci. Geomech. Abstr. 27, 345e352. Kurikami, H., 2007. Groundwater Flow Analysis in the Horonobe Underground Research Laboratory Project -Recalculation Based on the Investigation until Fiscal Year 2005-, JAEA-research 2007-036 (in Japanese with English abstract). Kurikami, H., Takeuchi, R., Yabuuchi, S., 2008. Scale effect and heterogeneity of hydraulic conductivity of sedimentary rocks at Horonobe URL site. Phys. Chem. Earth 33, S37eS44. Lenoir, N., Bornert, M., Desrues, J., Bésuelle, P., Viggiani, G., 2007. Volumetric digital image correlation applied to X-ray microtomography images from triaxial compression tests on argillaceous rock. Strain 43, 193e205. Masuda, K., Fujimoto, K., Arai, T., 2002. A new gas-medium, highpressure and hightemperature deformation apparatus at AIST, Japan. Earth Planets Space 54, 1091e1094. Matsui, H., Yamamoto, T., 2004. Mechanical properties of Neogene sedimentary rocks in Horonobe. In: Proceedings of Annual Meeting of Japan Society of Civil Engineers, CS1e012, 23-24 (in Japanese). Mitchell, T.M., Faulkner, D.R., 2009. The nature and origin of off-fault damage surrounding strike-slip fault zones with a wide range of displacements: a field study from the Atacama fault system, northern Chile. J. Struct. Geol. 31, 802e 816. Müller, G., 1967. Diagenesis in argillaceous sediments. In: Larsen, G., Chilingar, G.V. (Eds.), Diagenesis in Sediments. Elsevier Publishing Company, Amsterdam. Nakano, T., Tsuchiyama, A., Uesugi, K., Uesugi, M., Shinohara, K., 2006. “Slice”Softwares for Basic 3-D Analysis. Japan Synchrotron Radiation Research Institute (JASRI). Slice Home Page (web). http://www-bl20.spring8.or.jp/slice/.

Niizato, T., Yasue, K.-I., Kurikami, H., 2007. Impacts of natural events and processes on groundwater flow conditions: a case study in the Horonobe Area, Hokkaido, northern Japan. In: Workshop Proceedings of Stability and Buffering Capacity of the Geosphere for Long-term Isolation of Radioactive Waste: Application to Crystalline Rock. OECD/NEA, Manchester, UK, pp. 269e275. Ota, K., Abe, H., Yamaguchi, T., Kunimaru, T., Ishii, E., Kurikami, H., Tomura, G., Shibano, K., Hama, K., Matsui, H., Niizato, T., Takahashi, K., Niunoya, S., Ohara, H., Asamori, K., Morioka, H., Funaki, H., Shigeta, N., Fukushima, T., 2007. Horonobe Underground Research Laboratory Project Synthesis of Phase I Investigations 2001e2005 Volume “Geoscientific Research”. JAEA-Research 2007-044 (in Japanese with English abstract). Sanada, H., Niunoya, S., Matsui, H., Fujii, Y., 2009. Influences of sedimentary history on the mechanical properties and microscopic structure change of Horonobe siliceous rocks. J. MMIJ 125, 521e529 (in Japanese with English abstract). Savage, H.M., Brodsky, E.E., 2011. Collateral damage: evolution with displacement of fracture distribution and secondary fault strands in fault damage zones. J. Geophys. Res. 116, B03405. http://dx.doi.org/10.1029/2010JB007665. Scholz, C.H., 2002. The Mechanics of Earthquakes and Faulting, second ed. Cambridge Univ. Press, Cambridge, U. K. Schulze-Makuch, D., Carlson, D.A., Cherkauer, D.S., Malik, P., 1999. Scale dependency of hydraulic conductivity in heterogeneous media. Ground Water 36 (6), 904e 919. Tada, R., Iijima, A., 1983. Petrology and diagenetic changes of Neogene siliceous rocks in northern Japan. J. Sediment. Petrol. 53 (3), 911e930. Takahashi, M., 2003. Permeability change during experimental fault smearing. J. Geophys. Res. 108 (B5), 2235. http://dx.doi.org/10.1029/2002JB001984. Takahashi, M., Mizoguchi, K., Kitamura, K., Masuda, K., 2007. Effects of clay content on the frictional strength and fluid transport property of faults. J. Geophys. Res. 112, B08206. http://dx.doi.org/10.1029/2006JB004678. Takahashi, M., Mizoguchi, K., Masuda, K., 2009. Potential of phyllosilicate dehydration and dehydroxylation reactions to trigger earthquakes. J. Geophys. Res. 114, B02207. http://dx.doi.org/10.1029/2008JB005630. Uehara, S., Shimamoto, T., Okazaki, K., Funaki, H., Kurikami, H., Niizato, T., Ohnishi, Y., 2012. Can surface samples be used to infer underground permeability structure? A test case for a Neogene sedimentary basin in Horonobe, Japan. Int. J. Rock Mech. Min. Sci. 56, 1e14. Wellington, S.L., Vinegar, H.J., 1987. X-ray computerized tomography. J. Petrol. Technol. 39 (8), 885e898. Wong, T.-f., David, C., Zhu, W., 1997. The transition from brittle faulting to cataclastic flow: mechanical deformation. J. Geophys. Res. 102, 3009e3025. Zhu, W., Wong, T.-f, 1997. The transition from brittle faulting to cataclastic flow: permeability evolution. J. Geophys. Res. 102, 3027e3041.