Experimental Study of the Laws Between the Effective ... - Science Direct

Available online at www.sciencedirect.com

ScienceDirect Energy Procedia 63 (2014) 5654 – 5663

GHGT-12

Experimental study of the laws between the effective confining pressure and mudstone permeability ZENG Zhijiaoa,*, LI Xiaochuna, SHI Lua, BAI Binga, FANG Zhiminga, WANG Yingb a

State Key Laboratory of Geomechanics and Geotechnical Engineering, Institute of Rock and Soil Mechanics, Chinese Academy of Sciences, Wuhan 430071, China b CO2 Geological Storage and Enhanced Oil & Gas Recovery, Institute of Oil and Gas Engineering, Southwest Petroleum University, Chengdu, Sichuan, China

Abstract

The stress dependent permeability of a mudstone sample was measured using a transient pulse technique in a specially designed apparatus in which the confining pressure, pore pressure, and temperature were independently controlled. Experiments were conducted with the confining pressure first gradually increased from 10 to 30 MPa and then subsequently reduced back to 10 MPa. And the pore pressure was kept constant of 5 MPa during the whole process. During the loading process of effective confining pressure, the permeability of the mudstone had a nonlinear reduction with increasing stress and showed a high stress sensitivity (varying three orders of magnitude). The phenomenon can be explained that microcrack closure is the dominant mechanism controlling the evolution of mudstone permeability with the effective confining pressure during the loading process. The experimental results indicate that there is good fit of a model with exponential relationship of permeability and effective confining pressure to the data points. When unloading the effective confining pressure, the permeability of mudstone had very tiny increase and showed weak recovery capability. This is the result of the irreversibility of crack closure. © 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license

© 2013 The Authors. Published by Elsevier Ltd. (http://creativecommons.org/licenses/by-nc-nd/3.0/). Selection and peer-review under responsibility of GHGT.

Peer-review under responsibility of the Organizing Committee of GHGT-12

Keywords: permeability, mudstone, effective confining pressure, microcrack

* Corresponding author. Tel.: +86-13476826801; fax: 027-87198967. E-mail address: [email protected]

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-12 doi:10.1016/j.egypro.2014.11.598

5655

ZENG Zhijiao et al. / Energy Procedia 63 (2014) 5654 – 5663

1. Introduction CO2 capture and geological storage is considered as one of the most promising options and the only technology available to mitigate atmospheric emissions of CO2 from large-scale fossil fuel usage [1, 2]. The viability and longterm safety mainly depend on the caprock sealing capacity and integrity [3]. The permeability characteristics of the caprock will directly affect the safety and leakage risk assessment of CO2 storage. It is well known that the rock permeability changes with the stress condition. Early laboratory work has showed that permeability of rock will decrease with increasing effective confining pressure [4-26]. Based on the experimental results, several models of effective confining pressure - permeability relationship have been proposed, including exponential relationship, power law relationship and polynomial relationship [6, 9-11, 13, 15, 16, 20-22, 24-26]. Besides, it is well recognized that the permeability is dependent not only on the current loading condition, but also on the stress history within a sedimentary basin. Some researchers have made some further study about the influence of the stress history on the rock permeability. They increased the confining pressure and then subsequently reduced it back to the initial value. The rock permeabilities in loading and unloading processes were compared [5, 7, 8, 11, 12, 16, 22]. The test results indicates that there is obvious hysteresis effect for the restitution of permeability in the process of unloading of confining pressure. It should be noted that these researches were conducted mostly based on the experiments for sandstones which have a low content of matrix and high permeability and usually serve as reservoir formations for storage projects. But for caprocks such as mudstones, which have a high content of matrix and low permeability, what kind of relationship there is between permeability and effective confining pressure during the whole loading cycle? And what are the differences from that of sandstones? With this question in mind, we measured the permeability change of mudstone under loading-unloading of confining pressure. Based on the test results, we made further discussions about the caprock permeability change with the stress disturbance induced by fluid injection and its implications for the safety assessment of CO2 storage projects. 2. Experimental technique and materials tested 2.1. Experimental technique The permeability of mudstone was measured with a transient pulse technique originated by Brace [27]. The rock sample was placed in a pressure vessel and was connected to two fluid reservoirs, an upstream reservoir and a downstream reservoir. A heat shrinkable tube jacket was set on the circumference of the sample to prevent it from radial fluid flow at its lateral boundary and to insulate the pore pressure system from the hydrostatic pressure system. Fig. 1a shows the specimen assemblage for permeability measurements in the pressure vessel. After sample saturation and pore-pressure homogenization in the specimen, the pore pressure was raised instantaneously by a small amount (0.15 MPa is used in this test) at one end of the specimen. The pore pressure at this end then decreased, while the pore pressure at the other end increased with time as the fluid flowed through the specimen, i.e. the differential pressure decayed with time. The permeability of the specimen could be calculated according to the speed of the differential pressure decay. Fig. 1b illustrates the principle underlying the test.

P2 Peq

Peq

P1

P1 t

(a)

t (b)

Fig. 1. (a) Sample assemblage for permeability measurements under the effective confining pressure; (b) Principle of the transient pulse technique

5656


Figure 2 shows the experimental apparatus. The device can be divided into two symmetrical parts, which are called upstream and downstream respectively. There are two needle valves between upstream and downstream. When both of them are opened, the upstream and downstream are connected. When one of them is closed, the upstream and downstream are disconnected. We can raise the pore pressure upstream or downstream with a small amount by adjusting the other valve to make the fluid compressed. Confining pressure and pore pressure are applied by Teledyne Isco D Series Pump, which can be increased up to 60 MPa and 40 MPa, respectively. The differential pressure between upstream and downstream is measured by a DP15 variable reluctance differential pressure transducer, whose full-scale pressure range is 220 kPa. Both water and gas can be used as pore fluid medium. The whole device is placed in a thermostat water bath tank which keeps the system at a constant temperature to reduce the test error of pressure and differential pressure caused by temperature change. And the temperature can be set to any value between room temperature and 100 ° C.

Fig. 2. Experimental apparatus for the permeability test

In order to confirm the validity and accuracy of the apparatus in permeability test by transient pulse technique, comparative test was done to measure the permeability of a specimen under the same conditions with transient pulse technique and constant pressure steady state method. The two tests were both done at temperature of 30 ° C, using nitrogen for pore fluid medium at the confining pressure of 10 MPa. The transient pulse experiment was done after the pore pressure in the specimen equilibrated at 5 MPa. The upstream pore pressure was raised instantaneously by 0.15 MPa and the differential pressure decayed to zero after three minutes. If experimentally obtained values of § 'p · ln ¨ t ¸ were plotted vs. time, a straight line would be got (except for early times). Figure 3 shows the semi-log © 'p0 ¹ differential pressure vs. time curve. Based on this curve, the permeability of the specimen could be calculated according to the flowing equation: 14696m1 P Lf z k §1 1 · f1 Apm ¨ ¸ © V1 V2 ¹

(1)

where k is the permeability of the porous media; m1 is the slope of the linear part of the semi-log differential pressure vs. time curve; P is the fluid dynamic viscosity; L , A are the sample length and surface respectively; f z , f1 are the gas compressibility correction factor, mass flow correction factor respectively; pm is the mean pore pressure; V1 , V2 are the volume of upstream and downstream reservoir. The permeability measured by transient pulse technique was 20.77 ȝD.

ln(Ƹpt/Ƹp0)


5657

time(s)

Fig. 3. Semi-log differential pressure vs. time curve

For the constant pressure steady state method, the upstream and downstream pore pressures were 5.2 MPa and 5.0 MPa respectively. Thus the mean pore pressure was 5.1 MPa and the imposed pressure gradient was 0.2 MPa. The piston in the pumps connected to the upstream and downstream lines moved in a "push-pull" configuration, which allowed them to accommodate the fluid flow induced by the pressure gradient. The piston displacement in upstream pump was recorded (Figure 4). Once the evolution of displacement became linear, the slope corresponded to the successive flow rates. The permeability of the specimen could be calculated according to the flowing equation: 2 p1Q1 P L QP L k (2) A'p A P12 P22

volume(ml)

where Q is the average flow rate at the average pore pressure, 'p is the fluid pressure gradient, p1 , p2 are the upstream and downstream pore pressure respectively, and Q1 is the flow rate at the upstream pore pressure p1 . The permeability measured by constant pressure steady state method was 19.31 ȝD. 100 95 90 85

time(min)

Fig. 4. Piston displacements throughout the constant pressure steady state experiment

The measured values of this two methods agree with each other very well. This means that it is valid and reliable to measure permeability for this apparatus with the transient pulse technique. 2.2. Description of the material and test procedure The mudstone tested is a Cretaceous sedimentary rock, drilled from Dangyang city of Hubei province, which can basically represent the properties of caprock in Jianghan Basin. A cylinder specimen with a 50 mm diameter and 100 mm length was prepared with the axes perpendicular to the beddings. The mineral constituents were identified by Xray diffraction (XRD). The framework grains of mudstone mainly consist of quartz grains, calcite, feldspar, and they are cemented by mixed-layer minerals of illite and smectite, illite and chlorite. The pore structure was observed by mercury porosimeter. The results of the mercury porosimeter measurement show that the porosity is 9.94%, and pores with the width of about 10 nm are most abundant. Nitrogen was used for pore fluid medium in this test. The temperature of the system was kept at 30° C. We minimized the upstream and downstream storage volume to about 10 cc respectively so as to increase the sensitivity

5658


of pore pressure change during the tests, which can shorten the time for the permeability measurement. The test procedures were as follows: (1) The specimen was dried to constant weight; (2) The specimen was assembled as shown in Fig.1 and put in the pressure vessel; (3) A confining pressure of 5 MPa was exerted. (4) All the lines and the specimen were evacuated to ensure that there was no air or other fluid in the system. (5) The confining pressure and pore pressure were increased up to 10 MPa, 5 MPa respectively. At this pore pressure, Nitrogen in the pump would flow into the lines and then to the pore spaces of the specimen. After a period of time, the fluid flow stopped and the specimen was saturated with Nitrogen. From this time on, the test started. During the whole process of test, the pore pressure was kept constant and the confining pressure rose to 10 MPa, 12 MPa, 15 MPa, 18 MPa, 20 MPa, 25 MPa, 30 MPa and then dropped to 25 MPa, 20 MPa, 18 MPa, 15 MPa, 12 MPa, 10 MPa. At each confining pressure, the permeability of the specimen was measured. In order to get the stress and seepage in the mudstone specimen completely balanced, each load and relief process takes above 30 min and 60 min respectively. 3. Results The permeability change of mudstone specimen under loading-unloading of confining pressure was measured. Table 1 shows the permeability of specimen at each confining pressure. Figure 5 shows the permeability evolution of mudstone specimen in the process of loading-unloading. Table 1. Test results of permeability of mudstone specimen confining pressure(Mpa)

pore pressure(Mpa)

effective confining pressure(Mpa)

permeability(D)

10

5

5

184.2

12

5

7

66.6

15

5

10

9

18

5

13

1.67

20

5

15

0.529

25

5

20

0.23

30

5

25

0.116

25

5

20

0.122

20

5

15

0.134

18

5

13

0.145

15

5

10

0.194

12

5

7

0.312

10

5

5

0.605

Logk(ȝD)

Effective confining pressure (Mpa)

Fig. 5. Semi-log permeability vs. effective confining pressure plot

5659


Figure 5 shows a nonlinear reduction in permeability with increasing effective confining pressure. The permeability decreases rapidly at low effective pressure and stabilizes with increasing pressure. When the effective confining pressure was increased from 5 MPa to 25 MPa, the permeability of mudstone reduced to about 0.063%. This indicates that the permeability of mudstone is very sensitive to changes in the effective confining pressure. During the unloading path, there is basically no recovery for permeability of mudstone. When the effective confining pressure was dropped to the initial value of 5 MPa, the permeability increased to only 0.33% of the initial permeability. For this tested mudstone, an exponential relationship would be suitable for describing the stress dependent permeability under loading of confining pressure, which can be expressed as follows: k k0 e E pe (3) where k denotes the permeability under the effective confining pressure pe ; k0 represents the permeability under atmospheric pressure p0 , and E is a material constant. In this study, k0

2346.0 P D and E

0.509MPa 1 .

Fig. 6. Loading curve of the stress dependent permeability

4. Discussion 4.1. Interpretation for stress-dependency of permeability of mudstone There have been many experimental studies about the effective confining pressure dependency of permeability for sandstones previously [13, 15-17, 19, 21-23]. After comparing these experimental data with this work, we can find that the permeability of mudstone is more sensitive to the effective confining pressure than the sandstone during loading process, but the recovery capability of permeability of mudstone is smaller than that of sandstone during unloading process. This may be interpreted by the influence of the deformation mechanism on the stress dependency of rock permeability. David et al. [6] proposed three types of permeability evolution with the effective confining pressure induced by three different kinds of deformation mechanisms: Type I permeability evolution, typically for low porosity crystalline rock where the closure of microcracks plays an important role and there is a large stress sensitivity of permeability and significant reduction in permeability under a low effective confining pressure; Type II permeability evolution for porous clastic rock where the compaction is related to the relative movement and rearrangement of grains and there is relatively low stress sensitivity of permeability; and Type III permeability evolution, typically for unconsolidated materials, whose permeability evolution curve has a sigmoidal shape, with an inflection point at a given pressure called hereafter the "critical pressure" corresponds to the onset of grain crushing and pore collapse, and there is a rapid decrease in permeability when the effective pressure exceeds this critical value. The permeability evolution of most of sandstones can be categorized as type II, and the tested mudstone can be classified as having type I permeability evolution. Which means that relative movement and rearrangement of grains and microcrack closure are the dominant mechanisms controlling the evolution of sandstone and mudstone

5660


permeability with the effective confining pressure, respectively. This is supported by scanning electron microscope (SEM) images of the sample (Fig. 7). Microcracks with a width of about 8-10 ȝm can be clearly identified in Fig. 7.

Fig. 7. SEM image of mudstone sample

According to P. Gavrilenko and Y. Gueguen [4], the mudstone permeability can be expressed by the statistical distribution characteristics of microcracks as follows: 4S 0 3 (4) k N f D c 5 F x0 15 where N 0f is the crack number density at zero pressure (total number of cracks per unit volume), D is the aspect ratio (ratio of the semi-minor axis a to the semi-major axis c) of cracks represented by flattened revolution ellipsoids, F x0 is the fraction of cracks which contribute to flow. As a result of progressive closure of cracks during loading of confining pressure, cracks whose high aspect ratio is progressively reduced to smaller values, and the fraction of cracks contribute to flow is dropped because of the reduce of probability that two cracks intersect. This will reduce the rock permeability significantly. Smaller recovery capability of mudstone permeability during unloading of confining pressure may be resulted from the irreversibility of crack closures. In fact, the crack closures may alter the connected crack network by ‘cutting’ some branches, which makes many cracks contribute to flow become no longer efficient for fluid transport. 4.2. Implications for CO2 storage projects Our experiments showed that mudstone permeability decreased rapidly at low effective pressure and stabilized with increasing pressure. This reveals that permeability of shallow rock has a high stress sensitivity and has a significant reduction with depth. As the depth increases, the decrease of permeability slows down. When a certain depth is reached, the permeability becomes stable. Then the rock permeability keeps at its residual value, and will not have significant change with the increase of depth or stress. For this tested mudstone, the effective confining pressure


5661

at which the permeability stabilizes is 15 MPa, corresponding to the depth of about 1200 m below the ground. So, if this tested mudstone is the caprock of CO2 storage projects, the storage depth of more than 1200 m underground would be better. At this depth, the mudstone has very low penetration ability and its permeability is relatively stable with increasing pressure, which makes the storage projects has lower risk of leakage and more safe. Besides, in CO2 storage projects, the CO2 injection will change the stress in reservoir and caprock around injection wells. The injection and accumulation of CO2 in the pores increase the pore pressure, leading to the reduction of effective stress. Conversely, the process of sedimentation and consolidation for caprock before the CO2 injection causes the enhancement of effective stress in the rock. So, stress disturbance caused by CO2 injection corresponds to an unloading process of effective stress, and it will result in the enhancement of permeability in caprock. However, most caprocks, such as mudstone and shale, have type I permeability evolution as mentioned above, where microcrack closure is the dominant mechanism controlling the evolution of permeability with the effective confining pressure. Besides, these rocks usually have weak recovery capability of permeability during unloading of confining pressure because of the irreversibility of crack closures. For this tested mudstone, when the effective confining pressure was dropped to the initial value, its permeability was only about five times the one at maximum pressure. We can conclude that increase of permeability for caprock because of stress disturbance induced by fluid injection is small and it will not have significant impacts on the safety of storage project. 5. Conclusions We have built a new apparatus to measure the permeability of tight rock samples over a range of effective pressures. It has the capacity to measure both liquid and gas permeabilities using a transient pulse technique. By doing comparative test to measure the permeability of a specimen under the same conditions with transient pulse technique and constant pressure steady state method, the validity and accuracy of this apparatus in permeability test using transient pulse technique was confirmed. This apparatus was used to investigate the relation between permeability and effective confining pressure for mudstone from Dangyang city, Hubei province. The permeability of this mudstone has a high stress sensitivity during loading of effective confining pressure. There is significant reduction in permeability under a low effective pressure. As the effective pressure increases, the decrease of permeability slows down and the permeability stabilized at the effective confining pressure of 15 MPa. When the effective confining pressure was increased from 5 MPa to 25 MPa, the permeability of mudstone was reduced to about 0.063%. An exponential relationship can be used to describe the stress dependent permeability during loading process. It is postulated that microcrack closure in the tested mudstone contributes to the stress sensitivity of permeability under low effective confining pressure. This hypothesis is supported by the SEM image. The tested mudstone has weak recovery capability of permeability during unloading of confining pressure, which reflects the irreversibility of crack closure. Acknowledgements This study was financially supported by the National Natural Science Foundation Item (41172285) – Research of mechanism and laws for permeability evolution of mudstone in post-failure stage and its implication for CO2 leakage risk. References [1] IPCC special report on carbon dioxide capture and storage. In Prepared by Working Group III of the Intergovernmental Panel on Climate Change; Metz, B., Davidson, O., de Coninck, H. C., Loos, M., Meyer, L. A., Eds.; Cambridge University Press: Cambridge, UK/New York, NY, USA, 2005; p442. [2] Bennaceur, K.; Gielen, D.; Kerr, T.; Tam, C. CO2 Capture and Storage: A Key Carbon Abatement Option; IEA/OECD: Paris, 2008. [3] Juan Song, Dongxiao Zhang. Comprehensive Review of Caprock-Sealing Mechanisms for Geologic Carbon Sequestration. Environ. Sci. Technol. 2013; 47(1): 9-22.

5662


[4] P. Gavrilenko, Y. Gueguen. Pressure dependence of permeability: a model for cracked rocks. Geophys. J. Int. 1989; 98: 159-172. [5] Ohmyoung Kwon, Andreas K. Kronenberg, Anthony F. Gangi, Brann Johnson, and Bruce E. Herbert. Permeability of illite-bearing shale: 1. Anisotropy and effects of clay content and loading. J. Geophys. Res. 109, B10205, doi: 10.1029/2004JB003052. [6] CHRISTIAN DAVID, TENG-FONG WONG, WENLU ZHU and JIAXIANG ZHANG. Laboratory Measurement of Compaction-induced Permeability Change in Porous Rocks: Implications for the Generation and Maintenance of Pore Pressure Excess in the Crust. Pure Appl Geophys 1994; 143: 425-456. [7] Hong-Xing Wang, Guan Wang, Zhang-Xing (John) Chen, Ron C.K. Wong. Deformational characteristics of rock in low permeable reservoir and their effect on permeability. Journal of Petroleum Science and Engineering 2010; 75: 240-243. [8] A.J. Bolton. Some measurements of permeability and effective stress on a heterogeneous soil mixture: implications for recovery of inelastic strains. Engineering Geology 2000; 57: 95-104. [9] D. Jasinge, P.G. Ranjith, S.K. Choi. Effects of effective stress changes on permeability of latrobe valley brown coal. Fuel 2011; 90: 1292-1300. [10] M. Jobmann, Th. Wilsnack, H.-D. Voigt. Investigation of damage-induced permeability of Opalinus clay. International Journal of Rock Mechanics & Mining Sciences 2010; 47: 279-285. [11] Zhaoping Meng, Guoqing Li. Experimental research on the permeability of high-rank coal under a varying stress and its influencing factors. Engineering Geology 2013; 162: 108-117. [12] M. Darot, T. Reuschle. Acoustic wave velocity and permeability evolution during pressure cycles on a thermally cracked granite. International Journal of Rock Mechanics & Mining Sciences 2000; 37: 1019-1026. [13] Norbert Zisser, Georg Nover. Anisotropy of permeability and complex resistivity of tight sandstones subjected to hydrostatic pressure. Journal of Applied Geophysics 2009; 68: 356-370. [14] K. Okazaki, H. Noda, S. Uehara, T. Shimamoto. Permeability, porosity and pore geometry evolution during compaction of Neogene sedimentary rocks. Journal of Structural Geology 2014; 62: 1-12. [15] Yasser M. Metwally, Carl H. Sondergeld. Measuring low permeabilities of gas-sands and shales using a pressure transmission technique. International Journal of Rock Mechanics & Mining Sciences 2011; 48: 1135-1144. [16] Jia-Jyun Dong, Jui-Yu Hsu, Wen-Jie Wu, Toshi Shimamoto, Jih-Hao Hung, En-Chao Yeh, Yun-Hao Wu, Hiroki Sone. Stress-dependence of the permeability and porosity of sandstone and shale from TCDP Hole-A. International Journal of Rock Mechanics & Mining Sciences 2010; 47: 1141-1157. [17] HUANG Yuanzhi, WANG Enzhi. Experimental study of the laws between the effective confining pressure and rock permeability. J Tsinghua Univ (Sci & Tech) 2007; 47(3): 340-343. [18] XU Demin, HUANG Runqiu, ZHANG Qiang, DENG Ying’er. EXPERIMENTAL STUDY ON PERMEABILITY OF POROUS MEDIA UNDER HIGH CONFINING PRESSURE. Journal of Engineering Geology 2007; 15(06): 752-756. [19] QIAO Liping, WANG Zhechao, LI Shucai. EFFECTIVE STRESS LAW FOR PERMEABILITY OF TIGHT GAS RESERVOIR SANDSTONE. Chinese Journal of Rock Mechanics and Engineering 2011; 30(07): 14221427. [20] PENG Yong-wei, QI Qing-xin, DENG Zhi-gang, LI Hong-yan. Experimental research on sensibility of permeability of coal samples under confining pressure status based on scale effect. JOURNAL OF CHINA COAL SOCIETY 2008; 33(05): 509-513. [21] Xue Yongchao, Cheng Linsong. EXPERIMENTAL STUDY ON PERMEABILITY VARIATION WITH CONFINING PRESSURE IN MICRO-FRACTURE AND LOW-PERMEABILITY ROCK. PETROLEUM GEOLOGY & EXPERIMENT 2007; 29(01): 108-110. [22] He Yulong, Yang Lizhong. TESTING STUDY ON VARIATIONAL CHARACTERISTICS OF ROCK MASS PERMEABILITY UNDER LOADING-UNLOADING OF CONFINING PRESSURE. Chinese Journal of Rock Mechanics and Engineering 2004; 23(03): 415-419. [23] ZENG Ping, ZHAO Jin-zhou, LI Zhi-ping, GUO Jian-chun, YANG Man-ping. EXPERIMENTAL STUDY CONCERNING THE EFFECT OF TEMPERATURE, EFFECTIVE STRESS AND WATER SATURATION ON THE PERMEABILITY OF TIGHT SANDSTONE. NATURAL GAS GEOSCIENCE 2005; 16(01): 31-34.


5663

[24] Gangi, A. F. Variation of whole and fractured porous rock permeability with confining pressure. Int. J. Rock Mech. Min. Sci. 1978; 15: 249-257. [25] Luo, R.L., Cheng. 2008. Formation characteristics of rock in low permeability reservoir under confining pressure. [26] Evans JP, Forster CB, Goddard JV. Permeability of fault-related rocks, and implications for hydraulic structure of fault. J Struct Geol 1997; 19(11): 1393-1404. [27] Brace W F, Walsh J B, Frangos W T. Permeability of granite under high pressure. J. Geophy. Res. 1968; 73(B2): 2225-2235.