Abstract Laboratory experiments on hydraulic fracturing were carried out to investigate the damage evolution during cyclic injection using acoustic emission monitoring. Cylindrical granite specimens with 50 mm in diameter and 100 mm in height were tested under a uniaxial vertical load of 25 MPa. A central borehole of 8 mm in diameter along the specimen long axis was drilled, and water was pumped in at a rate of 50 mm 3 /s. A total of 25 specimens were tested, 5 under continuous injection, and 20 under cyclic injection with different levels of maximum pressure starting from 76.9% to 101.2% of the average breakdown pressure under continuous injection. The results indicate that the level of maximum pressure during cyclic injection, is inverse to the number of cycles that lead to failure. Moreover, there was an average reduction of 25.9 dB of the maximum AE amplitudes recorded during cyclic injection. AE activity was detected from the early stages of cyclic injection, and only right before failure during continuous injection. The crack mode classification shows similar ratios of tensile to shear cracks for continuous and cyclic injection, and the 3D location of these events indicates that the 3.5% of shear cracks during cyclic injection are well distributed among the AE cloud. Finally, the hydraulic energy was higher for cyclic injection, especially for those cases with lower levels of maximum pressure. The radiated seismic energy ranges from 2.3 × 10 −14 J to 6.3 × 10 −11 J even for cases that failed at low number of cycles, however with increasing cycles the values are persistently low. 1. Introduction The stimulation of Enhanced Geothermal Systems through hydraulic fracturing, as a way to increase its permeability, has been associated with undesired levels of induced seismicity (Majer et al., 2007). An alternative cyclic hydraulic fracturing scheme was suggested with the aim to reduce induced seismicity. This concept is based on a hydro-mechanical model, and numerical simulation results show a reduction in breakdown pressure and seismicity (Zang et al., 2013). Later, this concept was implement in laboratory scale experiments that also resulted in a reduction of breakdown pressure and AE activity. Zhuang et al. (2016) reported a reduction of nearly 20% of breakdown pressure during cyclic hydraulic fracturing in granite specimens. Another study conducted with Tennessee sandstone, reported a 16% breakdown pressure reduction (Patel et al., 2017). However, no experimental study has covered the effect of maximum pressure during cyclic injection, fixed as a fraction of the average breakdown pressure under continuous injection. In this study, laboratory experiments were carried out on cylindrical granite specimens to investigate damage evolution during cyclic hydraulic fracturing using acoustic emission (AE) monitoring for fracture detection and propagation. The effect of maximum allowed pressure during cyclic injection on the number of cycles to failure and the maximum AE amplitude was investigated. Also, the AE crack type as well as the hydraulic and seismic energies were analyzed and discussed. The results were compared with a group of tests carried out under continuous injection, and some conclusion were drafted.

Abstract When interpreting hydraulic fracturing in low permeable rock such as granite, the effect of fluid infiltration into rock matrix is commonly neglected. There are laboratory experiments of hydraulic fracturing on granite samples indicating the fluid infiltration may play important role in the breakdown pressure. This paper presents a numerical simulation of hydraulic fracturing in low permeable rock with the consideration of fluid infiltration into rock matrix. A 2D Distinct Element Method code UDEC was employed where the polygonal blocks were generated by a Voronoi tessellation feature to provide incipient fractures to consider fluid diffusion into rock matrix. This Voronoi tessellated model was verified to be hydraulically representative as a low permeable rock. The simulations of hydraulic fracturing with varying pressurization rates detected three regimes in terms of the effect of pressurization rate on the magnitude of breakdown pressure, which is generally consistent with that predicted by the existing analytical models. It was demonstrated that the varying breakdown pressure with the pressurization rate can be attributed to the varied flow diffusion conditions in the surrounding rock matrix and associated change of stress at wellbore wall. Extending the flow diffusion length such as cyclic injection mode tends to reduce the breakdown pressure. 1. Introduction There is growing interest in performing hydraulic fracturing in low permeable formation for the economical production of resources and energies. The hydraulic fracturing in low permeable shale plays a critical role in shale gas production (e.g., Clerk et al., 2012). Another emerging application of hydraulic fracturing is to create fluid circulation pathways for extracting deep geothermal energy which is stored in deep crystalline formation (e.g., Legarth et al., 2005). There have been numerous studies involving hydraulic fracturing in low permeable rocks (e.g., Chen et al., 2015) and the infiltration of injection fluid into the surrounding rock has rarely been considered. Some laboratory hydraulic fracturing tests in low permeable rock demonstrated that the infiltration of fluid into the rock and its influence might not be negligible. The hydraulic fracturing experiment on Lac du Bonnet granite showed the penetration of fluid into the rock and its influence on the wave velocities (Falls et al., 1992). Ishida et al., (2004) carried out hydraulic fracturing on granite samples by water and oil injection, and found that the influence of fluid viscosity on hydraulic fracture was mainly caused by the varying fluid infiltration conditions. Recent hydraulic fracturing experiments on Pocheon granite demonstrated the water infiltration could significantly impact hydraulic fracturing behavior and the breakdown pressure was increased with the injection rate (Zhuang et al., 2018).

Abstract Cylindrical specimens of Pocheon granite were used for laboratory hydraulic fracturing (HF) tests, and sleeve fracturing (SF) tests where the borehole was covered with membrane and fluid infiltration into sample was ignored. To investigate the effect of borehole diameter and specimen height on hydraulic fracturing behavior, all specimens have a fixed external diameter of 50 mm, while borehole diameter varied at 5, 8, 12 and 14 mm, and specimen height varied at 50, 70 and 100 mm. Most tests were conducted at a constant injection rate of 50 or 100 mm 3 /s. A few SF tests were compared at different injection rates varying at 5, 25, 50 and 100 mm 3 /s using both oil and water as the injection fluid. Experimental results show that breakdown pressure (BP) decreases with increasing borehole diameter, while it is insignificantly affected by specimen height. BP increases with increasing injection rate while the change is not significant at higher injection rates. The pressurization rate estimated from injection pressure curves ranges from nearly 0.003 to 1.2 MPa/s, and its impact on the hydraulic fracturing breakdown pressure was investigated. BP predicted by a fracturing mechanics model were similar to those measured values during experiments under a constant injection rate of 100 mm 3 /s, when assuming a 3 mm length of preexisting crack. 1. Introduction Hydraulic fracturing (HF) method has been widely used in reservoir engineering development and in-situ stress measurement. Hydraulic fracturing behavior varies depending on rock types. For development of Enhanced Geothermal System (EGS), where the target reservoir formations are usually hard crystalline rocks like granite and tight sandstone, understanding of hydraulic fracturing behavior of these rocks, particularly fracture initiation and propagation and the resulted breakdown pressure, is very necessary. For an experimental study on hydraulic fracturing of rocks in laboratory, one has to decide sample and borehole size. The sample geometry will influence the experimental results. Table 1 lists sample sizes and borehole diameters applied in past experimental studies on hydraulic fracturing behavior of various granites. For cylindrical samples, the ratio of d/D has a wide range of 1:4 to nearly 1:50 and the ratio of D/L ranges from 1:1.2 to 1:3.5. Either through-going hole or half-going hole was used. For cubic samples in true tri-axial testing, the ratio of borehole diameter and the minimum side length d/s ranges from 1:6.4 to 1:20. Haimson and Zhao (1991) reported that laboratory tests in boreholes that are at least 20 mm in dia. yield breakdown pressures that are essentially unaffected by borehole diameter size and are directly usable in interpretation of field data. Experimental results by Morita et al. (1996) showed a clear decrease of breakdown pressure of Berea sandstone when the borehole size increased from 38 mm to 100 mm. Detournay and Cheng (1992) proposed a mathematical model interpreting pressurization rate and size effects on the magnitude of hydraulic fracturing breakdown pressure. The rate effect is seen as a consequence of the interaction of two lengthscales: a diffusion length and a microstructural length. In this study, we performed hydraulic fracturing tests on cylindrical granite samples, having three different ratios of D/L (1:1, 1:1.4 and 1:2) and four different ratios of d/D (1:3.6, 1:4.2; 1:6.3 and 1:10). Influence of D/L and d/D on the breakdown pressure and induced fractures was investigated. Additionally, sleeve fracturing tests where the borehole is covered with a latex membrane to prevent injection fluid infiltration into the rock specimen were performed at four different injection rates of 5, 25, 50 and 100 mm 3 /s. Oil and tap water, with viscosities of 52 cp and around 1cp at room temperature are used as injection fluids.

ABSTRACT The drill monitoring is a technique for predicting the risks at face by analyzing the mechanical quantities measured at hydraulic rock drill in the excavation of tunnels or rock mass for underground structures. Since drill monitoring can be conducted on realtime- basis when excavating blast holes or rock bolt holes in tunnel excavation, it enables fast and quantitative prediction and evaluation of tunnel face. Though a number of studies have been conducted on the drilling data, the selection of drilling parameters and numerical quantification of mechanical quantities for the evaluation of the ground characteristics have not been established yet. In this study, drilling tests were conducted with rock specimens to identify drilling parameters and the correlation between the drilling data which was analyzed with the data obtained in the test drilling. The average values of the drilling parameters were calculated using the average values of the sections of the drilling data, and the drilling mechanism was verified by correlation analysis between the drilling parameters. Introduction The drill monitoring which is a survey technique using drilling parameters can improve the work efficiency and safety by predicting and evaluating the state of the ground ahead of tunnel face. This technique measures and analyzes various drilling parameters, including instantaneous advance speed, torque pressure, percussion pressure, feed pressure, rotation speed, and number of percussion, which are measured with hydraulic sensors and data recorder installed on hydraulic rock drill of jumbo drill, during drilling process (Pfister, 1985; Peck and Vynne, 1993; Schunnesson, 1996; Toda corporation, 2005; Mituisumitomo corporation, 2005). The drilling parameters obtained in drill monitoring contain quantitative data of the ground, including weathering, fracturing (Barr, 1984; Scoble and Peck, 1987), hardness (Schunnesson and Sturk, 1997) of rock mass and different rock type boundaries (Schunnesson and Holme, 1997). A number of studies have been conducted to improve the efficiency in tunnel excavation by analyzing the correlation between drilling parameters and rock, rock mass and ground. Kahraman (2000) proposed drillability index which can predict penetration rate from the rock characteristics. Thuro (1997, 2003) proposed the concept of drillability using drilling parameters to classify rocks by the easiness of drilling, and proposed excavatability concept to define the relationship between the destruction work and specific consumption of explosive using drilling rate. In Korea, N. Y. Kim et al. (2001) studied on drilling parameters, however, the studies are at preliminary application step and the drilling data obtaining and analyzing system has not been developed yet. According to the reference survey, the preceding studies on drilling parameters have following problems, though they can provide ground data promptly and quantitatively. In drill monitoring, the reaction of the ground condition to the drilling is not monitored directly but the 1428 mechanical quantities of the hydraulic rock drills are measured.

ABSTRACT It has been widely reported that fault zones are the most influencing factor of tunnel collapse. This study reviewed the applicability of the 3-D absolute displacement measurement of a tunnel by using digital vision monitoring. Also analyzed is the behavior of a tunnel displacement when passing through a fault zone with various orientations by using 3-D finite element analysis. As a result of analyzing 3-D deformation, it was found that the variation of displacement trend on horizontal/vertical section per measurement section would be larger as it approaches to the core of the fault zone. It was also examined that the variation of crown settlement and trend line would be the most influenced at a point where a tunnel meets fault zone. However, since the variation of crown settlement trend line starts as it is very close to a fault zone(0.5D ~ 1D), it may be difficult to anticipate a fault zone from crown settlement trend line. It was also proposed the move vector concept of a center point in a plane, which is created by connecting displacement measuring points. After applying the proposed monitoring center vector(MCV) to a tunnel section through which a fault zone passes, it was found that the influence would be well reflected depending on the fault zone orientation. In addition, the study suggested the ‘stereonet projection’ as a method to express the move behavior of MCV in a 3-D space and suggested how to anticipate front fault zone through variations of measuring section center point vector on the stereonet projection diagram. The 3-D absolute displacement measurement of tunnel and the displacement analysis using the digital vision measurement developed in the study were applied to a tunnel site. Comparison with the survey results of fault zone showed good 1. Introduction For safe and economic construction of a tunnel in a fault zone, appropriate method have to be employed to predict and prepare for sudden change in ground condition ahead of the tunnel face as early as possible. Studies on the prediction of fault zone ahead of tunnel face using the convergence of tunnel have been conducted since late 1980s. Various methods have been tried, including analysis on the trend line and influence line of the displacement in tunnel axis, analysis on the change trend of the vector orientation which is the ratio between the displacements in radial and advance directions, and so on. Most of the studies on the prediction of fault ahead of tunnel face using tunnel convergence have been focused on the investigation of the deformation behavior of the displacement by excavation at the measuring point. However, it was not possible to take full consideration of the 3-dimensional behavior of displacement by fault zone ahead since the analyses were only limited to the displacement in single or two directions,. In this paper, a 3-dimensional displacement analysis method which can efficiently represent behavior of rock in fault zone is proposed.