Energy Procedia 63 ( 2014 ) 6335 – 6343

Available online at www.sciencedirect.com

ScienceDirect

1876-6102 © 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).Peer-review under responsibility of the Organizing Committee of GHGT-12doi: 10.1016/j.egypro.2014.11.666

GHGT-12

Reservoir characterization for site selection in the Gundih CCS project, Indonesia

Takeshi Tsujia*, Toshifumi Matsuokab, Wawan Gunawan A. Kadirc, Masami Hatod, Toru Takahashie, Mohammad Rachmat Sulec, Keigo Kitamuraa, Yasuhiro Yamadab,

Kyosuke Onishif, Djedi S. Widartog, Rio I. Sebayangg, Agung Prasetyog, Awali Priyonoc, Tutuka Ariadjic, Benyamin Sapiiec, Eko Widiantog, Ariesty Ratna Asikinc, and

Gundih CCS project team a International Institute for Carbon-Neutral Energy Research, Kyushu University, 744 Motooka Fukuoka, 819-0395, Japan

b Faculty of Engineering, Kyoto University, Kyotodaigaku-Katsura, Kyoto, 615-8540, Japan c Institut Teknologi Bandung, Bandung, 40132, Indonesia

d Waseda University, 3-4-1 Shinjyuku-ku Okubo, Tokyo, 169-8555, Japan e Fukada Geological Institute, 2-13-12 Bunkyo-ku Honkomagome, Tokyo, 113-0021, Japan

f Faculty of International Resource Sciences, Akita University, 1-1 Tegatagakuen-cho. Akita, Akita, 101-0852, Japan g Pertamina, Jakarta, 12950, Indonesia

Abstract

A pilot CCS project in Indonesia will be implemented in Gundih area, Central Java Province in Indonesia. Before the CO2 injection, the reservoirs for CO2 injection must be characterized carefully by conducting geophysical exploration as well as reservoir simulation, in order to make sure that the reservoir is suitable for CO2 storage. Here we report results of reservoir characterization and simulation for the determination of CO2 injection site in the Gundih area. Subsurface structures imaged on seismic reflection profiles indicate that the Ngrayong formation is one of the candidates for CO2 injection. We observed the outcrop of the Ngrayong formation and measured hydrological and geophysical properties (e.g., permeability, seismic velocity) of the rock samples obtained from outcrop and wells. The Ngrayong formation has layered structure and heterogeneous characteristics. Using (1) hydrological properties, (2) subsurface structures (i.e., geometry of the Ngrayong formation) and (3) physical properties predicted by integrating seismic and logging data via acoustic impedance inversion, we applied reservoir simulation and evaluated security of the CO2 injection sites. © 2013 The Authors. Published by Elsevier Ltd. Selection and peer-review under responsibility of GHGT.

* Corresponding author. Tel.: +81- 92-802-6875; fax: +81-92-802-6875.

E-mail address: tsuji@i2cner.kyushu-u.ac.jp

© 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).Peer-review under responsibility of the Organizing Committee of GHGT-12

6336 Takeshi Tsuji et al. / Energy Procedia 63 ( 2014 ) 6335 – 6343

Keywords: Gundih CCS project; Reservoir characterization; Reservoir simulation; Heterogeneous formation; Pore pressure

1. Introduction

Carbon Capture and Storage (CCS) has near-term impact on CO2 emissions. Since the CO2 in the atmosphere has been significantly increasing, we should decrease the CO2 emissions from human activities as soon as possible. CCS is attractive way to reduce the CO2 emissions using the almost established technology. The roadblocks for implementation of CO2 storage are (a) risk of CO2 leakage, (b) risk of injection-induced seismicity, and (c) high-cost. Because these roadblocks are strongly related to the highly uncertain local geological characteristics of potential storage sites, (1) reservoir characterization and simulation and (2) monitoring/modeling of injected CO2 are crucial procedures in the development of CCS. Especially, the CCS projects in tectonically active area near from the plate convergent margins (e.g., Japanese Island or Indonesia) have some difficulties, compared to those in the stable continental crust. To overcome these roadblocks, we have developed methods of reservoir characterization and monitoring /modeling of injected CO2.

For the CO2 storage for the tectonically active regions, the following issues should be considered. (a) Heterogeneous geological formation:

Because geological formations in plate convergent margins are heterogeneous compared to the large-scale reservoirs in the central part of the continental plates, we need to consider the heterogeneity (e.g., fractures) in constructing geologic models for reservoir simulation and in designing monitoring surveys. Furthermore, it is difficult to find stable structural closure (i.e., anticline structure) for CO2 injection, thus we need to use the mechanisms of residual trapping, dissolution trapping and mineralogical trapping [1]. Even if we use reservoirs without structural closure for CO2 storage, reservoirs around Japanese islands have a capacity of over 100 billion tons of CO2 [2]. This volume is corresponding to ~100 years of total CO2 emission from Japan.

(b) Limited information for CO2 storage: Geophysical data including well data are intensively acquired in CCS-EOR projects. However, there is

limited geophysical data available that can be brought to bear for CO2 injection into aquifer formations or new reservoir. Therefore, we need to characterize the reservoir (or construct geologic model) from limited geophysical /geological data.

(c) Awareness for earthquakes: Natural earthquakes are frequently occurred along the plate convergent margins. There is a possibility that

the stress state within the crust is close to the critical state (failure threshold stress), over which the earthquake is generated (i.e., critically stressed nature) [3]. The increased pore pressure due to CO2 injection reduces the frictional resistance to fault slip. Therefore, we must accurately monitor and control pore pressure variations due to CO2 injection. As discussed later, the estimation of stress state (including pore pressure) before the CO2 injection is crucial procedure.

(d) Long-term monitoring and modeling: Since monitoring in CCS projects should extend about hundred years, the requirements are much different

from the conventional approaches in oil production. Especially we should continuously monitor the injected CO2 if the lithology has heterogeneous characteristics (fractures). We need to develop the capability to monitor injected CO2 using effective methods (e.g., seismic monitoring using ambient noise [4]).

To establish the methods of reservoir characterization as well as monitoring/modeling of injected CO2 in tectonically active regions, the Nagaoka CCS project was conducted in Japan. The pilot CCS project demonstrated that the injected CO2 can be clearly monitored by seismic data [5, 6], logging data [7] and geochemical data [8]. Using these monitoring data, we confirmed that the injected CO2 is safely stored within injection reservoir (<~100m). Here, we focus on “Gundih CCS project” in Indonesia. This project will be a first pilot CCS project in Indonesia for research and development of technologies for assessing deep strata at CO2 injection and for monitoring of underground distribution of CO2. Indonesia has a plan to reduce CO2 emission by 26% by 2020 [9]. Since CO2 emission from gas production fields is a major problem in Indonesia, we plan to inject CO2 in the Gundih gas field, central Java Island (Fig. 1). The CO2 content within the produced gas is more than 20% in the Gundih gas field, so

Takeshi Tsuji et al. / Energy Procedia 63 ( 2014 ) 6335 – 6343 6337

that CO2 injection near the gas production wells could be effective way to avoid abundant CO2 emission from this area. In this study, we characterized reservoirs of the Gundih gas field mainly using seismic data, and applied reservoir simulation in order to evaluate the potential and security of CO2 injection sites. Presently we have two candidate sites for CO2 injection around the Gundih gas field; (1) Central Gundih Gas field and (2) Northern Gundih field (yellow circles in Fig 1b).

Fig. 1 (a) Map of the Java Island, Indonesia, provided by Google. (b) Locations of two candidates for CO2 injection (yellow circles): Central Gundih gas field and Northern Gundih field. The red and green stars indicate the locations of outcrops and shallow boreholes we drilled in this

project, respectively. From these locations (stars), we obtained discrete samples for laboratory measurements. (c) Lithology around the East Java Basin [13]. Tuban formation deeper than the Ngrayong formation is known as overpressure zone.

6338 Takeshi Tsuji et al. / Energy Procedia 63 ( 2014 ) 6335 – 6343

2. Geologic setting

The Gundih gas field is located at the vicinity of the east Java basin (Fig. 1), where contains of thousands meters of Tertiary sedimentary sequences. This sedimentary sequence has a good potential of hydrocarbon source rock and reservoir rock. The east Java basin is a back-arc basin [10]. Basement configuration of east Java basin is controlled by two main structural trends, that NE-SW trend are generally only found in northern shelf and E-W trend contained in Mandala Central high and south basin. Indeed, the geologic structures (e.g., fault) observed around the Gundih gas field are extended mainly for E-W (or NE-SW) direction (Fig. 2). The E-W or NE-SW trends would be influenced to the injected CO2 behavior within reservoir.

The lithology we will inject CO2 is the Ngrayong formation (Fig. 1b), because the formation is sandstone and because the depth of this formation is ~1 km in the Gundih area. Since the depth of the formation is deeper than ~800m in the candidates for CO2 injection reservoir [11], we can effectively inject CO2 as a supercritical state. The pore pressure at the Ngrayong formation is known as almost hydrostatic condition from previous wells. The physical and hydrological properties of the Ngrayong formation are much different between northern and southern regions; the northern Ngrayong formation is sandy and well sorted (Figs. 1 and 3). The permeability of the outcrop samples obtained in the northern Ngrayong formation is high. Whereas, in the southern formation, the lithology is mud dominant and has low permeability [12].

The Ngrayong formation is overlaid by the Wonocolo formation, composed by massive grey fossiliferous sandy marl (Fig. 1c). This formation is deposited during late Miocene in the outer neritic environment. Since this formation has low permeability, it may work as a seal layer. Indeed, the rock samples obtained from this lithology using the shallow borehole drilled in this project (Fig. 3) have low permeability. The Tuban formation deeper than the Ngrayong formation is known as overpressure zone.

Fig. 2 Geological structures extracted from 2D seismic profiles at (a) the central Gundih gas field and (b) the northern Gundih field. These

profiles are acquired for north-south direction. The red lines indicate the candidate for well location in panel (a) and the JEPON-1 site in panel (b). Light blue lines indicate the faults extending for east-west direction. Blue arrow in panel (a) shows the 3D seismic survey area.

Takeshi Tsuji et al. / Energy Procedia 63 ( 2014 ) 6335 – 6343 6339

3. Reservoir characterization and simulation

There are two main candidate sites for CO2 injection around the Gundih gas field; (1) Central Gundih Gas field and (2) Northern Gundih field (yellow circles in Fig 1b). Both candidates have advantages and disadvantages, so we summarize them by conducting reservoir characterization and simulation. We then provide information for the decision of CO2 injection site. Here we mainly use seismic reflection data, logging data and discrete rock samples for reservoir characterization. To construct geologic model from these geophysical data, we apply acoustic impedance (AI) inversion to the post-stack data [14] and Common Reflection Surface (CRS) stacking analysis to the pre-stack data [15]. To characterize the hydrological properties of the Ngrayong formation, furthermore, we obtain rock samples from outcrops as well as shallow boreholes (stars in Fig.1; Fig.3). We then apply reservoir simulation using the constructed geologic models.

Fig. 3 Pictures of outcrops of the Ngrayong formation (A-C) and shallow drilling operation (right). The location of each outcrop is noted in Fig. 1b. Hydrological properties (e.g., permeability) of this formation are much different between northern region (A and B) and southern region (C).

3.1 Central Gundih Gas field

The high-resolution 3D seismic reflection data and several 2D seismic data were acquired in this region and can be used for reservoir characterization. This site is located within the gas field and is close to CO2 capture facility (Fig. 1b), therefore CO2 can be supplied using pipelines. However, because of no borehole for CO2 injection, we need to drill new borehole if we inject CO2 in this region.

We could not find large-scale structural closure (i.e., anticline) for the Ngrayong formation within the 3D seismic survey area (Fig. 2a). Therefore, we need to store the CO2 using the residual trapping mechanism. Although the reverse faults are developed at the southern side of this region, the relatively stable formations are observed at the northern half of the 3D seismic area (Fig. 2a). The horizon of top of the Ngrayong formation extracted from 3D seismic data demonstrates that dislocation plane (i.e., fracture or fault) is not observed in the stable region [11]. Whereas, the strike-slip fault is developed at the north of the central Gundih gas field (Fig. 2a).

We applied reservoir simulation using realistic hydrological properties constructed by AI inversion and considering residual and dissolution trappings. Reservoir simulation is crucial step in CCS project to check the storage capacity of the reservoir and the risk of leak through faults. Because the field observation demonstrates that the lithology of the Ngrayong formation is fine grain in this region (Fig. 3c), the injectivity of this lithology is low

6340 Takeshi Tsuji et al. / Energy Procedia 63 ( 2014 ) 6335 – 6343

[12]. The reservoir simulation demonstrates that the injected CO2 is not arrived at the strike-slip fault over 1000 years and safely trapped within the CO2 injection reservoir [16], when CO2 is injected at the northern side of the reverse faults (red bar in Fig 2a). 3.2 Northern Gundih field

We can use the suspended well (i.e., JEPON-1 site) for CO2 injection in the Northern Gunidh field. The geological formations imaged on seismic profiles indicate that the JEPON-1 is located at the top of anticline structure whose axis is continued for E-W direction (Fig. 2b). The shallower part of the Ngrayong formation is shale dominant, but the logging data (gamma ray log) indicates several interbedded sand layers. Sidewall core samples of JEPON-1 show that the sandstones are well sorted, indicating the high permeability (good injectivity). These sand layers could be used for the CO2 injection. In the Gundih pilot CCS project, the injection amount of CO2 is limited (~10,000 tons/year), thus these thin sand layers could be enough for injection.

At the northern side of the anticline (JEPON-1 site), the strike-slip fault is existed [11]. The fault is dipping to south and extends beneath the well. In the depth of injection formation (~900m), the distance between the injection well and the closest fault is ~400 m. Therefore, we need to carefully consider the fault by performing reservoir simulation in order to estimate CO2 saturation as well as pore pressure. The pore pressure is believed to be nearly hydrostatic conditions in this site because the Pertamina drilled through the Ngrayong formation in many locations and did not observe overpressure in the Ngrayong formation. However, the overpressure zone is expected at deeper lithology (e.g., Tuban formation).

We characterized reservoir and constructed geologic model using AI inversion (Fig 4a). The AI estimated via inversion is well consistent with the value of logging-derived AI value. We further classified sand and mud layers by considering the relationship between AI and gamma ray log data. We then identified four sand dominant layers within the Ngrayong formation (red arrow in Fig 4b), and these sand layers continuously exist for horizontal direction. By considering depth and temperature of the injection reservoir, the depth of the shallowest sand layer is closed to the supercritical phase transition. Therefore, the deeper three sand layers would be candidates for CO2 injection.

To estimate future CO2 distribution as well as pore pressure, we have applied reservoir simulation around the Ngrayong formation. The geologic model includes four thin sandstone layers (with high permeability) in the Ngrayong formation as classified in AI inversion, and the other zones are assumed as mud rock layers (with low permeability). We calculate CO2 behavior in the case of high horizontal permeability (200 md) and low horizontal permeability (50 md) for sand layers, in order to recognize the CO2 behavior of the two extreme cases. The CO2 behavior would be within the range of these predictions. The horizontal permeability is assumed to be 10 times larger than the vertical permeability because the outcrop observations demonstrate sand-mud layered structure (Fig. 3c). We conducted the reservoir simulation at several reservoir conditions (temperature, porosity) as well as injection rates. Underground temperature is not strongly influenced to the flooding area of injected CO2. Here we show the results of the reservoir simulation of CO2 injection during 1 year with rate of ~10,000 tons/year (Fig 5).

The simulation results demonstrated that the injected CO2 is located around the borehole (within ~100m from borehole) in the low permeability case (Fig. 5a) and is stored far from the fault located at northern side (Fig 2b). In high permeability case (200 md), the injected CO2 migrates for shallower direction (western direction) along the axis of anticline (Fig. 5b), thus the injected CO2 would not move to northern direction (i.e. the closest fault). Flooding area and direction are strongly influenced by permeability of sandstone layers. Therefore, the acquisition of permeability of reservoir rock must be important task for CO2 injection project. Furthermore, since the formation geometry significantly controls the CO2 movement, it is important to know the detailed geological structures around the injection site. Pore pressure predicted by reservoir simulation is not much increased at the fault zone in the condition of the small amount of injection rate (~10,000 tons/year).

Takeshi Tsuji et al. / Energy Procedia 63 ( 2014 ) 6335 – 6343 6341

Fig. 4 Geologic model used for reservoir simulation. (a) Acoustic impedance (AI) profile derived from inversion. (b) Lithology classified into sand (yellow) and mud (green) from AI model by defining the threshold. Red arrows indicate four dominant sand layers within the Ngrayong

formation.

Fig. 5 Examples of reservoir simulations (map view of CO2 saturation) for two permeability cases (50md for horizontal absolute permeability in panel a; 200md in panel b). This figure show the CO2 saturation after 1 year from CO2 injection. Red indicates the higher CO2 saturated region. Black circles indicate the well location (JEPON-1). Yellow lines indicate the location of seismic profile displayed in Fig .4. These simulations

demonstrate that the injected CO2 is distributed around the injection well (<150m) in lower permeability case (panel a) and moves along the axis of the anticline (west direction) in higher permeability case (panel b). The strike-slip fault is located at the northern side of the JEPON-1 site.

6342 Takeshi Tsuji et al. / Energy Procedia 63 ( 2014 ) 6335 – 6343

4. Discussion

The most significant concerns in CO2 storage are CO2 leakage and injection-induced earthquakes. The shear stress at failure is obtained by the following equation;

(1) where C is cohesive strength, is coefficient of friction, is normal stress, and is pore pressure. If pore pressure increases due to CO2 injection, the shear stress at failure is decreased. When the shear stress along fault is higher than the critical value , the fault could be ruptured. Therefore, the pore pressure is one of the most important parameters to evaluate the stability of lithology (or fault) and should be accurately monitored during and after CO2 injection. From equation (1), furthermore, we recognize that the background stress state ( ) is also crucial information to evaluate stability of the faults. If the shear stress along the fault is close to (a little lower than) the shear stress at failure (critically stressed nature; [3]) before the CO2 injection, the small increase in pore pressure due to CO2 injection could generate fault rupture. Therefore, we should accurately estimate background stress state including pore pressure before CO2 injection.

Recent earthquake observations in fluid injection experiments indicate that the earthquakes are sometime occurred in the basement deeper than the injection reservoir [17]. The earthquakes in deep lithology could be occurred by CO2 injection-induced earthquakes. Even if the injection-induced earthquakes occurred around the injection interval are small, they can change the stress state around the reservoir (even deep crust). Because there is a possibility that CO2 injection indirectly generates large earthquake, the earthquake prediction in the CO2 storage is difficult issue. Although the northern Gundih field seems to be better environment (less pore pressure), the deeper lithology (i.e., Tuban formation) may be high pore pressure conditions. Therefore, we should carefully characterize the stress state, in order not to disturb the stress state around the reservoir. As discussed above, the stress state (including pore pressure) estimation is the priority work in the CO2 storage. Pore pressure can be directly measure at the borehole, and its spatial distribution can be estimated from seismic velocity [18]. Stress state can be estimated from the borehole breakout [19], earthquake source mechanisms and shear-wave splitting analysis [20]. From the GPS or InSAR observation [21], furthermore, the strain could be estimated. By using the geomechanical modeling [22], furthermore, we should evaluate the influence of CO2 injection to the stress state around the injection reservoir. By using these available methods, we should estimate stress state around the injection site in near future.

Acknowledgements

We acknowledge support of the SATREPS project by JICA-JST. Because of limited space, we cannot describe all contributors as authors. But, many other scientists contribute to this CCS project. Especially we thank Pertamina and MIGAS for this project. T. Tsuji and K. Kitamura gratefully acknowledge the support of the I2CNER, sponsored by the World Premier International Research Center Initiative (WPI), MEXT, Japan, and Sumitomo Foundation (Environment).

References

[1] Suekane T, Nobuso T, Hirai S, Kiyota M. Geological storage of carbon dioxide by residual gas and solubility trapping. International Journal of Greenhouse Gas Control 2006; 2: 58-64.

[2] Ogawa T, Shidahara T, Nakanishi S, Yamamoto T, Yoneyama K, Okumura T, Hashimoto T. Saline-aquifer CO2 sequestration in Japan-methodology of storage capacity assessment. International Journal of Greenhouse Gas Control 2011; 5: 318-326.

[3] Zoback MD, Gorelick SM. Earthquake triggering and large-scale geologic storage of carbon dioxide, Proceedings of the National Academy of Sciences (PNAS) 2012; 109:10164-10168.

[4] Minato S, Tsuji T, Ohmi S, Matsuoka T. Monitoring seismic velocity change caused by the 2011 Tohoku-oki earthquake using ambient noise records. Geophysical Research Letters 2012; 39, L09309, doi:10.1029/2012GL051405.

[5] Saito H, Nobuoka D, Azuma H, Xue Z, Tanase D. Time-lapse crosswell seismic tomography for monitoring injected CO2 in an onshore

Takeshi Tsuji et al. / Energy Procedia 63 ( 2014 ) 6335 – 6343 6343

aquifer, Nagaoka, Japan. Exploration Geophysics 2006; 37: 30–36. [6] Onishi K, Ueyama T, Matsuoka T, Nobuoka D, Saito H, Azuma H, Xue Z. Application of crosswell seismic tomography using difference

analysis with data normalization to monitor CO2 flooding in an aquifer. International Journal of Greenhouse Gas Control 2009; 3, 311-321. [7] Chiyonobu S, Nakajima T, Zhang Y, Tsuji T, Xue Z. Effect of reservoir heterogeneity of Haizume Formation, Nagaoka Pilot Site, based on

high-resolution sedimentological analysis. Energy Procedia 2013; 37: 3546-3553. [8] Mito S, Xue Z. Post-Injection Monitoring of Stored CO2 at the Nagaoka pilot site: 5 years time-laps well logging results. Energy Procedia

2011; 4: 3284-3289. [9] Kadir WGA, Sule R, Alawiyah S, Setianingsh, Santoso D, Widarto DS, Tamba R, Aasongko D, Widianto E, Matsuoka T. First Pilot Study of

CO2-Sequestration in Indonesia, Proceedings of PIT HAGI 2012. [10] Sribudiyani, Muchsin N, Ryacudu R, Kunto T, Astono P, Prasetya I, Sapiie B, Asikin S, Harsolumakso AH, Yulianto I. The collision of east

java microplate and its implication for hydrocarbon occurences in the east Java basin, Proceeding Indonesian Petroleum Association, 29th Annual convention and exhibition 2003; 1: 1-12.

[11] Tsuji T, Matsuoka T, Takahashi T, Kitamura K, Onishi K, Yamada Y, Sule MR, Kadir WGA, Widarto DS, Sebayang RI, Prasetyo A, Priyono A, Widianto E, Sapiie B. Pilot CCS project in Indonesia “Gundih CCS project”: Geological and geophysical surveys for site selection and monitoring, EGU General Assembly, 2013.

[12] Kitamura K, Yamada Y, Onishi K, Tsuji T, Chiyonobu S, Sapiie B, Bahar A, Danio H, Muhammad A, Erdi A, Sari VM, Matsuoka T, Kadir WGA. Gundih CCS project team, Potential evaluation of CO2 reservoir using the measured petrophysical parameter of rock samples in the Gundih CCS Project, Indonesia. Submitted to Energy Procedia.

[13] Ardhana W. A depositional model for the early middle Miocene Ngrayong formation and implications for exploration in the east Java basin, Proceedings Indonesian Petroleum Association, 22nd Annual Convention Proceedings, 1993.

[14] Riedel M, Bellefleur G, Mair S, Brent TA, Dallimore SR. Acoustic impedance inversion and seismic reflection continuity analysis for delineating gas hydrate resources near the Mallik research sites, Mackenzie Delta, Northwest Territories, Canada. Geophysics 2009; 74: B125-B137.

[15] Minato S, Tsuji T, Matsuoka T, Nishizaka N, Ikeda M. Global optimization by simulated annealing for common reflection surface stacking, and its application to low-fold marine data in southwest Japan. Exploration Geophysics 2012; 43: 59-69.

[16] Ariadji T, Putra AP, Ekowati PD, Matsuoka T, Tsuji T, Kadir WGA, Sule MR, Widarto DS. Quantification of Influencing Parameters of Storage for Gundih Area Considering Residual and Solubility Trapping Using Reservoir Simulation Model and Design Of Experiment. submitted to Journal of Petroleum Science and Engineering.

[17] Verdon JP. Significance for secure CO2 storage of earthquakes induced by fluid injection, Environmental Research Letters 2014; 9. [18] Tsuji T, Kamei R, Pratt G. Pore pressure distribution of a mega-splay fault system in the Nankai Trough subduction zone: Insight into up-dip

extent of the seismogenic zone, Earth and Planetary Science Letters 2014, 396: 165-178. [19] Chang C, McNeill LC, Moore JC, Lin W, Conin M, Yamada Y, In situ stress state in the Nankai accretionary wedge estimated from

borehole wall failures. Geochem. Geophys. Geosyst. 2010, 11, Q0AD04, doi:10.1029/2010GC003261. [20] Tsuji T, Hino R, Sanada Y, Yamamoto K, Park J-O, No T, Araki E, Bangs N, von Huene R, Moore G, Kinoshita M. In situ stress state from

walkaround VSP anisotropy in the Kumano basin southeast of the Kii Peninsula, Japan, Geochem. Geophys. Geosyst. 2011, 12, Q0AD19, doi:10.1029/2011GC003583.

[21] Khakim MYN, Tsuji T, Matsuoka T. Detection of Localized Surface Uplift by Differential SAR Interferometry at the Hangingstone Oil Sand Field, Alberta, Canada. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 2013, 6: 2344-2354.

[22] Khakim MYN, Tsuji T, Matsuoka T. Geomechanical Modeling for InSAR-derived surface deformation at steam-injection oil sand fields. Journal of Petroleum Science and Engineering 2012, 96-97: 152-161.