Abstract Easy oil has gone and now the focus of exploration and development in China has shifted to tight reservoirs deemed techanically challenging. One of the key challenges in tight reservoirs is how to place and land horizontal wells in sweet spots (with high reservoir quality and completion quality) and how to stage-fracture these wells efficiently to produce these tight reservoirs economically.

The paper presents a new 3D reservoir geomechanics workflow that has been applied to a tight gas reservoir in western China. The reservoir is very deep (up to 4500m) and the production rates from the wells are very low. Some hydraulic fracturing had been conducted for vertical exploration wells but the post-fracturing production rates were still not satisfactory. The best chance to produce this tight reservoir is to place horizontal wells in the areas with the best reservoir quality and completion quality and carry out optimized multistage hydraulic fracturing. To this end, a 3D full field geomechanics model was constructed through integration of seismic data, geological structure, core data and log data. This 3D geomechanics model enables a 3D identification of the high completion quality (high fracturability) zones in the reservoir and subseqently placement of a new horizontal well. A 1D mechanical model was then extracted along the planned trajectory from the 3D geomechanics model. Based on the 1D geomechanics model, optimization of the stage-fracturing design was conducted to obtain the optimal number of stages, optimum fracture half length and optimum staging. Introduction Multistage fracturing of horizontal wells has been successful in gas shale (Chong et al. 2010; King 2010) in North America and brings promise to development of many tight reservoirs around the world. However, experience shows that in many areas this technique has failed to achieve initial production rate goal (Ketter et al., 2006; Britt and Smith, 2009). One of the root causes is that heterogeneity of the reservoir was largely ignored, which led to poor placement of the horizontal wells and/or bypass of large portion of reservoir because of poor initialization and propagation of hydraulic fractures in multiple stages. These results highlighted the importance of 3D geomechanics modeling, to quantify the heterogeneity of the reservoirs and to enable the optimum well placement and hydraulic fracturing programs.

To this end, a new 3D reservoir geomechanics workflow was developed and applied to a tight field, named XG, near Karamay, Xinjiang, in western China. The target reservoir formation is the Jamuhe Permian tight sandstone with porosity of 8% and relative permeability of about 0.01mD. The preliminary exploration indicated a large gas reserve in the reservoir. For such tight reservoir, the economic success for gas production depends entirely on the effective generation by hydraulic fracturing of adequate surface area to flow (Economides and Nolte 2000). Since the reservoir is deep and very tight, the drilling and completion costs will be high, and a production rate up to 100,000 m3/d is required for economics. However, previous stimulation of vertical wells in this field did not yield satisfactory results, achieving a peak production rate of only up to 27,000 m3/d. To optimize the planning of horizontal well multistage fracturing, understanding the heterogeneity of the reservoir is the key. Through the 3D reservoir geomechanics workflow, a bettter understanding of heterogeneity of reservoir was obtained and optimization of hydraulic fracturing was able to be conducted. 3D Reservoir Geomechanics Workflow The workflow for 3D reservoir geomechanics in tight reservoirs is shown in Fig. 1. As illustrated, the workflow truly reflects the multidisciplinary nature of 3D reservoir geomechanics. The following sections give detailed explanations on each step of

IPTC 17115

3D Reservoir Geomechanics Workflow and Its Application to a Tight Gas Reservoir in Western China Kaibin Qiu, Schlumberger; Ning Cheng, Xiangui Ke, Yang Liu and Lirong Wang, PetroChina; Yingru Chen, Yong Wang and Pi Xiong, Schlumberger

Copyright 2013, International Petroleum Technology Conference This paper was prepared for presentation at the International Petroleum Technology Conference held in Beijing, China, 26–28 March 2013. This paper was selected for presentation by an IPTC Programme Committee following review of information contained in an abstract submitted by the author(s). Contents of the paper, as presented, have not been reviewed by the International Petroleum Technology Conference and are subject to correction by the author(s). The material, as presented, does not necessarily reflect any position of the International Petroleum Technology Conference, its officers, or members. Papers presented at IPTC are subject to publication review by Sponsor Society Committees of IPTC. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of the International Petroleum Technology Conference is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of where and by whom the paper was presented. Write Librarian, IPTC, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A., fax +1-972-952-9435

2 IPTC 17115

the workflow and results obtained for the XG field. Geological Modeling Five major geological surfaces were picked from seismic data interpretation (see Fig. 2). Horizon P3w is the top of the reservoir and Horizon P1j2 is the bottom of the reservoir.

There is no major fault visible from seismic data in the study area, so faults are not considered in the geological modelling, nor in the 3D stress initialization in the later step of the workflow. Spatially Continuous Velocity Analysis Building a 3D mechanical earth model (MEM) requires 3D seismic properties to delineate the lateral and vertical heterogeneities of the reservoir, overburden, and underburden. In many cases, seismic inversion data (including vP and vP/vS) are not available or are only available from the reservoir. Spatially continuous velocity analysis (SCVA) (Mao et al. 2000) becomes an effective way to obtain high-resolution 3D seismic velocity. SCVA is able to automatically generate seismic horizons and check the spatial continuity along horizons when picking velocity. It also uses automated velocity model building (AVMB) to filter bad picks, interpolates missing picks, and preserves high spatial resolution. This automatic picking method provides high-resolution velocity following key geological features in a very short time. It has been applied to many deepwater pore pressure prediction projects (Dutta 2002a, Banik et al. 2004, Dutta and Khazanehdari 2006) around the world in the last few years.

Fig. 3 shows the vertical cross section of SCVA velocity obtained for the study field. The SCVA velocity follows the general trend of increase with increase of depth until reaching the caprock of the reservoir. The caprock appears to be a low-velocity zone, and there is a clear increase of velocity when entering in the reservoir. To check the validity of the SCVA velocity, a comparison was made of the SCVA velocity (extracted at the well locations) and compressional slowness logs from the wells. A quite good consistency was observed (Fig. 4).

Petrophysical Evaluation Petrophysical evaluations were conducted for three wells in the study area. The reservoir properties and net pay for each well are listed in Table 1. The average porosity for gas layers and tight gas layers is approximately 8%, and the average permeability is 0.6 mD. Typically, the petrophysical evaluation represents gas permeability from a dry sample, and it is absolute permeability. Further investigation of well testing data and a history match of hydraulic fracturing data revealed the relative permeability (gas and water two phase) is about 0.01 mD. The difference between absolute permeability and relative permeability in this case highlights the importance of analyzing well testing data to more accurately estimate the “real” permeability in the reservoir.

Table 1: Petrophysical properties and net pays for XG field

Well Top (m)

Bottom (m)

Thickness (m)

Porosity (%)

Permeability (mD)

Water Saturation

(%) Comments

X1

4553.2 4569.0 15.8 8.9 0.55 45 Gas layer 4570.1 4575.0 4.9 8.3 0.49 53 Gas layer 4577.0 4579.1 2.1 8.9 0.76 44 Tight gas layer 4579.1 4582.0 2.9 3.5 0.02 100 Dry layer 4582.0 4590.0 8.0 7.4 0.43 42 Tight gas layer 4590.0 4591.0 1.0 3.0 0.02 100 Dry layer 4591.0 4597.0 6.0 7.1 0.29 49 Tight gas layer 4597.0 4598.4 1.4 3.0 0.01 80 Dry layer 4598.4 4602.1 3.7 6.9 0.25 59 Tight gas layer 4602.1 4616.0 13.9 4.0 0.10 62 Dry layer 4616.0 4618.0 2.0 7.8 0.46 49 Tight gas layer 4618.0 4658.0 40.0 4.5 0.15 60 Dry layer 4663.5 4728.0 64.5 4.6 0.10 70 Dry layer

G2

4746.0 4749.7 3.7 10.8 2.28 45 Gas layer 4749.7 4752.3 2.6 4.0 0.02 100 Dry layer 4752.3 4789.0 36.7 8.2 0.61 46 Gas layer 4791.5 4797.0 5.5 4.0 0.10 70 Dry layer 4798.0 4801.2 3.2 4.9 0.15 60 Dry layer 4808.0 4810.0 2.0 4.0 0.06 82 Dry layer 4811.0 4814.0 3.0 4.5 0.06 70 Dry layer 4820.0 4823.0 3.0 4.9 0.20 55 Dry layer 4828.9 4878.0 49.1 7.6 0.47 42 Gas layer

IPTC 17115 3

Well Top (m)

Bottom (m)

Thickness (m)

Porosity (%)

Permeability (mD)

Water Saturation

(%) Comments

G3

4609.0 4613.6 4.6 3.5 0.02 80 Dry layer 4613.6 4625.0 11.4 6.4 0.13 45 Gas layer 4626.0 4629.0 3.0 4.0 0.04 75 Dry layer 4638.9 4665.0 26.1 8.7 0.63 31 Gas layer4666.0 4701.5 35.5 7.6 0.33 41 Gas layer 4701.5 4733.0 31.5 4.8 0.10 60 Dry layer 4734.0 4768.5 34.5 4.0 0.01 80 Dry layer 4768.5 4776.0 7.5 3.0 0.01 100 Dry layer 4776.0 4780.5 4.5 9.1 0.11 85 Water layer 4780.5 4792.0 11.5 2.0 0.01 100 Dry layer4792.0 4800.0 8.0 9.0 0.08 88 Water layer

Geomechanical Core Testing Geomechanical core testing is an important step for both 1D and 3D geomechanical evaluations, which rely on empirical correlations to derive mechanical properties from sonic data or petrophysical properties (Coates and Denoo 1980, Morales and Marcinew 1993, Plumb 1994, Chang et al. 2006). These correlations were derived from specific rock type, age, depth range, and field, and their applications to other rocks may not be reliable unless calibrated with specific field conditions (Khaksar et al. 2009).

The minimum geomechanical laboratory test program includes the following: Triaxial compression test. This test determines the static elastic moduli including Young’s modulus and Poisson’s

ratio, and rock strength parameters including unconfined comressive strength (UCS) and friction angle. A triaxial test should be carried out at different confined stresses; normally three to five confined stress levels from 1 MPa to the in-situ confined stress in the reservoir. The test rquires three to five core samples cut from adjacent locations. However, should there be any heterogeneity from these core samples, deriving a consistent failure envelope from test data from these samples will be difficult (Pagoulatos 2004, Khaksar et al. 2009). To save core samples and avoid inconsistency of intrinsic mechanical properties among different samples, a multistage triaxial test (Kovari et al. 1983) is recommended.

Ultrasonic test: This test determines the dynamic elastic moduli including Young’s modulus and Poisson’s ratio. An ultrasonic test should be conducted simultaneously with the triaxial test. Compressional slowness and shear slowness from ultrasonic tesets, if comparable with sonic log data, can be used to establish correlation between dynamic moduli and static moduli.

CT scanning before and after the compression test. The baseline test CT scan provides important quality control for the core samples. Defects in core samples, such as vugs or coring-induced fractures, may not be evident from visual examinations. Samples with defects detected by the baseline CT scan will be discarded, or the test results will be used with caution. The post-test CT scan is used to describe the failure plane morphology, which is also important for interpretation of test data. For parameters used for the Mohr-Coulomb failure criterion, such as UCS and friction angle, the test data will not be representative unless the core sample fails from conventional shear failure.

For the XG project, multistage triaxial tests were performed on six core samples, and ultrasonic tests were simultaneously conducted at each stage for each test. Each multistage triaxial test was able to determine static elastic moduli, including Young’s modulus and Poisson’s ratio, and rock strength parameters, including UCS and friction angle, from one core sample. The ultrasonic tests conducted simultaneously for each stage and each core sample were able to determine dynamic elastic moduli including Young’s modulus and Poisson’s ratio.

The mechanical properties obtained from the core tests are shown in Table 2. Table 2: Geomechanical properties from multistage triaxial tests

Well Sample Depth

(m)

Static Young's Modulus

(Gpa)

Static Poisson's

Ratio

Cohesion (Mpa)

Friction Angle

(°)

Dynamic Poisson's

Ratio

Dynamic Young's Modulus

(Gpa)

UCS (Mpa)

G2 3077B 3077 19.1 0.2 25.9 31.3 0.20 42.3 92.1 X1 3825A 3825 22.3 0.16 33.9 26.8 0.20 46.2 110.2 X1 4490B 4490 14.1 0.16 27.1 21.6 0.22 34.5 79.8 X1 4552A 4552 14.0 0.26 2.3 43.5 0.23 37.1 10.8 X1 4552C 4552 15.2 0.25 3.2 46.1 0.27 35.6 15.9 G3 4933A 4933 19.8 0.16 19.6 33.7 0.22 37.3 73.3

4 IPTC 17115

Building 1D MEM The MEM is a numerical representation of the state of in-situ stresses and rock mechanical properties for a specific stratigraphic section in a field or basin. It includes elastic properties, rock strength data, and geostresses (Plumb et al. 2000, Ali et al. 2003). The ‘1D’ in ‘1D MEM’ stands for that the MEM is built along the wellbore trajectory, which can be either vertical or deviated.

The procedure used to build a 1D MEM has been documented elsewhere (Qiu et al. 2005, Qiu et al. 2006) and is not covered here. This paper presents only the results of the final 1D MEM from the two study wells in the XG field.

Fig. 5 displays the output window of the wellbore stability prediction from the 1D MEM for Well G3. See the in-situ stress state on Track 1, the mechanical properties on Track 2, the boundaries for wellbore deformation and the stable mud window on Track 3, and the synthetic borehole failure image (Qiu et al. 2008) on Track 4. A comparison of the predicted synthetic borehole failure image with the caliper logs presented on Track 5 (Fig. 5) shows a good match. This suggests the MEM was a close representation of the actual geomechanics in the locality of Well G3, and the MEM was validated.

Fig. 6 shows the comparison of the 1D MEM with the caliper log for another study well, G2; again a good match is observed. Gridding Gridding is to dividing the formation rock from top to bottom into an equivalent system of finite elements and generate the finite-element mesh to be used for numerical simulation. Gridding should be based on the geological framework from horizons and faults to better honor the geological structure of the field. The following guidelines should be followed when gridding:

Balance has to be achieved on grid size and number of grid that allows the grid is fine enough to give useable results and yet large enough to reduce computational effort (Logan, 2010). Normally, if the number of elements exceeds 10 million, the computation time will be excessively long if there is no high-performance cluster server.

Finer grids are desirable for the complex structure or where there is dramatic change of stress and strain (high stress or strain gradient).

For the XG project, the study area is 10×10 km2, and the vertical coverage is 6000 m. The gridding was conducted by following the main geological horizons, and the horizontal grid size (XY) was set at 50 m. High vertical resolution was ensured in the reservoir and caprock (25 m), and the grids were coarsened in the overburden and underburden. The total number of grids reaches up to 3 million (Fig. 7).

3D Mechanical Properties Mechanical properties should be assigned to the 3D finite-element grid to reflect horizontal and vertical heterogeneity of the mechanical properties of the 3D model. There are many ways to populate 3D mechanical properties to a 3D finite-element grid, either through stochastic method such as kriging (Goovaets 1997), or through deterministic ways such as directly interpolating from 3D seismic data.

For the XG project, the high-resolution SCVA velocity cube was directly interpolated to the 3D finite-element grid. Shear slowness was then calculated by using the following equation, which was derived from sonic data from the field:

1.05 / (-0.00222 0.748 / )s ct t , ……………………………………………….………….. (1)

where ts is shear slowness in units of us/ft, and tc is the compressional slowness in units of us/ft. Density is calculated by using the Gardner equation (Gardner et al. 1974): 0.25= 0.23 (1 000 000 / t )b c , ………………………………………………….……………. (2)

where ρb is the density in units of g/cm3. By using the same correlations employed in building the 1D MEM, 3D mechanical properties were populated in the 3D finite-element grid. The following properties were included: static Young’s modulus, static Poisson’s ratio, UCS, friction angle, and tensile strength.

Pore Pressure Prediction A 3D methodology for pore pressure prediction based on seismic data has been extensively documented by Dutta (2002b), Sayers et al. (2002) and Dutta and Khazanehdari (2006). All seismic velocity-based pore pressure prediction methods rely on the premise that seismic velocity is sensitive to effective stress and overpressure. This premise generally holds true for cases of young sediments with fast deposition. For these sediments, undercompaction is the main overpressure mechanism for which seismic velocity is sensitive to effective stress and then overpressure. In the XG field, the formation is relatively old (Cretaceous at top, Triassic in middle, and Permian at bottom), and the sensitivity of seismic velocity to effective stress was not observed. It was known from previous drilling experiences that pore pressure above the geological surface T1b is hydrostatic, and below it is slightly overpressured (up to 12.1 KPa/m). Then a 3D pore pressure cube was constructed by assigning the pore pressure gradient to be 1.03 g/cm3 above the geological surface T1b and 1.23 g/cm3 below it. This pore pressure cube was consistent with previous drilling experience, well test data in the reservoir, and wellbore stability analysis results (see Fig. 5 and Fig. 6). Stress Initialization The 3D finite-element method (FEM) is used to initialize the stress tensor in the 3D model. The detailed theory of the FEM is

IPTC 17115 5

involved in a good number of mathematical equations which had been well documented elsewhere (Zienkiewicz and Taylor 2005, Deb 2006, Logan 2010) and is not covered in this paper. Instead, we give an overview in an easily understandable language for those who practice geomechanics but are not experts on the FEM. When a FEM software tool is used to calculate stresses, it is actually solving a set of equations that relate forces, mechanical properties, and deformation at each point in the grid. The equations have the form Ku = F, ……………………………………………………………………………………. (3) where K is a global stiffness matrix and includes information about the grid structure and mechanical properties, u is a vector containing the displacement at each point, and F is a vector containing the various forces driving the deformation. An intuitive way to interpret the equation is imagining a steel spring is pulled by a force F at its end, the displacement of the end is proportional to the force F, and K represents the stiffness of the spring. Through solving equation (3), the displacement u is determined.

Since the derivative of the displacement u is the strain , and strain can be related to stress by constitutive law,

C ,………………………………………………………………………………….. (4) where C is stiffness. Then both and can be solved for each node in the finite-element elements.

In the 3D model used for XG field, the forces, or loads, in this type of simulation are of three types. First, there are gravitational loads depending on the density at each point. Second, there are pore pressure loads. Third, there are boundary loads on the edges of the model that represent the regional state of stress transmitted from neighboring areas.

In the 3D finite-element software tool used in the project, the preproduction stress solution was calculated in two stages. For the first step, only the gravity and pore pressure loads were applied; the second step added depth-dependent horizontal loads on the side boundaries. It is important to emphasize that these side loads were applied on the boundaries only; inside the grid, stress magnitudes and orientations will vary according to the local loads and properties. The magnitudes of the side loads were initially estimated based on the knowledge of horizontal stresses gained from building the 1D MEM, leading to a gradient of maximum horizontal stress with depth equal to 17 kPa/m, and a maximum/minimum horizontal stress ratio of around 1.15. Then the side loads were adjusted until the resulted 3D stress field was consistent with the in-situ stress profiles from respective1D MEM at each study well location.

After stress initialization, a 3D MEM was obtained. This 3D MEM illustrates the 3D in-situ stress field and the heterogeneity of mechanical properties for the tight gas field. Placement of Horizontal Wells With the 3D MEM, the placement of a proposed horizontal well was carried out. Since the minimum horizontal stress direction is in the north-to-south direction, the horizontal well should be placed in the same direction to allow generation of transverse hydraulic fractures during the hydraulic fracturing treatment (Economides and Nolte 2000).

The first horizontal well was a sidetrack of the existing vertical well, G2, making use of the existing vertical section of the wellbore. Then a vertical north-south section of the minimum horizontal stress crossing Well G2 was made (Fig. 8). The figure clearly shows the variation of stress magnitude in the reservoir. With consideration of the petrophysical properties (from Well G2 and nearby wells), formation dip, and the minimum horizontal stress, the horizontal well was landed in the zone with good gas saturation and low minimum horizontal stress. The reason to land the well into in the low minimum horizontal stress zone is because hydraulic fractures are prone to initialize and propagate in the low-stress zones. Optimization of Multstage Hydraulic Fracture of Horizontal Well Once the trajectory of the horizontal well was defined, the 1D mechanical properties and stress profiles were extracted from the 3D MEM, and optimization of the hydraulic fracturing treatment for the horizontal well was carried out.

For a multistage hydraulic fracturing treatment in a horizontal well, three questions should be answered during the design stage: What is the optimal number of stages? What is optimal fracture half-length? What is the optimal staging along the horizontal trajectory?

The objective way to select the optimal number of stages is through sensitivity analysis on production rate vs. number of fractures. The normal practice is to place four clusters of perforations for each stage in a cased hole completion. So, most likely, there will be four fractures created in each stage. Sensitivity analysis on cumulative production rate vs. number of fractures was carried out (Fig. 9a), and the cumulative production rate vs. number of fracture plot (Fig. 9b) shows the optimal number of fractures is from 32 to 34. It was therefore concluded that eight to nine stages would be optimal.

To obtain the optimal half-length, sensitivity analysis was conducted on the cumulative production rate vs. fracture half-length (Fig. 10a), revealing the answer to be from 220 to 240 m (Fig. 10b).

After the number of stages was determined, the next step was to place stages along the horizontal section of the borehole. Previous experience shows that even-spacing of each stage is not optimal and will not be able to exploit the full potential of the reservoir. For tight reservoirs, both reservoir quality (CQ), represented by porosity, saturation, and permeability of the reservoir, and completion quality (CQ), mostly represented by stress and mechanical properties, are equally important, and both must be considered during staging (Suarez-Rivera et al. 2011). For the horizontal well, a proprieptory software tool was used to evaluate both RQ and CQ along the horizontal section (Fig. 11). The fourth track from the right shows the RQ evaluation and the third track from the right shows the CQ evaluation. The second track from the right shows the default

6 IPTC 17115

staging plan (even-spacing). Based on the evaluation results, the optimal staging was obtained and is shown on the last track from the right. Conclusion Making production from tight reservoirs economically viable is a major challenge for many operators. To address the challenge, a new 3D geomechanics workflow was formulated and applied to a tight reservoir in western China. Through the 3D reservoir geomechanics workflow, the heterogeneity of the reservoir was characterized and 3D stress field was mapped, the horizontal well was placed in the sweet spots, and optimization of the hydraulic fracturing design was carried out. It was demonstrated that 3D reservoir geomechanics is able to provide a new insight of the tight reservoirs and will be a promising technology to enable efficient production of many tight reservoirs. Nomenclature ts = shear slowness t/L, us/ft tc = compressional slowness t/L, us/ft b = bulk density m/L3, g/cm3 K = global stiffness matrix m/t2, Pa.m u = displacement vector L, m F = force mL/t2, N = stress m/Lt2, KPa = strain dimensionless C = stiffness m/Lt2, KPa Acknowledgement The author thanks PetroChina and Schlumberger for permission to publish this paper. The authors also thanks Schlumberger colleagues Guo Wei, Hong Tian for their invaluable works that were used in this manuscript; and Ping Wang and Hai Liu for their contribution to the project. References Ali, A.H.A., Brown, T., Delgado, R., et al. 2003. Watching Rocks Change—Mechanical Earth Modeling. Oilfield Review 15 (1): 22–39. Banik, N.C., Schultz, G., Wool, G., et al. 2004. Seismic Pore Pressure Imaging in Deepwater Offshore West Africa. Paper SPE 89994

presented at the SPE Annual Technical Conference and Exhibition, Houston, Texas, 26–29 September. Britt, L.K. and Smith, M.B. 2009. Horizontal Well Completion, Stimulation Optimization and Risk Mitigation. Paper SPE 122526 presented

at the SPE Eastern Regional Meeting, Charleston, West Virginia, USA, 23–25 September. Chang, C., Zoback, M.D., and Khaksar, A. 2006. Empirical Relations Between Rock Strength and Physical Properties in Sedimentary

Rocks. Journal of Petroleum Science and Engineering 51: 223–237. Chong, K.K., Grieser, B., Jaripatke, O., et al. 2010. A Completion Roadmap to Shale Play Development: A Review of Successful

Approaches Toward Shale Play Stimulation in the Last Two Decades. Paper SPE 130369 presented at the CPS/SPE International Oil & Gas Conference and Exhibition in China, Beijing, 8–10 June.

Coates, G.R. and Denoo, S.A. 1980. Log Derived Mechanical Properties and Rock Stress. Paper U presented at SPWLA 21st Annual SPWLA Logging Symposium, 8–11 July.

Deb, D. 2006. Finite Element Method: Theory and Application to Geomechanics. New Delhi: Prentice-Hall of India. Dutta, N.C. 2002a. Geopressure Detection Using Reflection Seismic Data and Rock Physics Principles: Methodology and Case Histories

From Deepwater Tertiary Clastics Basins. Paper SPE 77820 presented at the SPE Asia Pacific Oil and Gas Conference and Exhibition, Melbourne, Australia, 8–10 October.

Dutta, N.C. 2002b. Geopressure Prediction Using Seismic Data: Current Status and the Road Ahead. Geophysics 67 (6): 2012–2041. Dutta, N.C. and Khazanehdari, J. 2006. Estimation of Formation Fluid Pressure Using High-Resolution Velocity From Inversion of Seismic

Data and a Rock Physics Model Based on Compaction and Burial Diagenesis of Shales. The Leading Edge 25:1528–1539. Economides, M.J. and Nolte, K.G. 2000. Reservoir Stimulation, Third Edition, John Wiley & Sons. Gardner, G.H.F., Gardner, L.W., and Gregory, A.R. 1974. Formation Velocity and Density—The Diagnostic Basics for Stratigraphic Traps.

Geophysics 39 (6): 770–780. Goovaerts, P. 1997. Geostatistics for Natural Resources Evaluation. Oxford, England: Oxford University Press. Ketter, A.A., Daniels, J.L., Heinze, J.R., et al. 2006. A Field Study Optimizing Completion Strategies for Fracture Initiation in Barnett Shale

Horizontal Wells. Paper SPE 103232 presented at the SPE Annual Technical Conference and Exhibition, San Antonio, Texas, USA, 24–27 September.

Khaksar, A., Taylor, P.G., Fang, Z., et al. 2009. Rock Strength from Core and Logs: Where We Are Stand and Ways to Go. Paper SPE 121972 presented at the SPE EUROPEC/EAGE Annual Conference and Exhibition, Amsterdam, The Netherlands, 8–11 June.

King, G.E. 2010. Thirty Years of Gas Shale Fracturing: What Have We Learned? Paper SPE 133456 presented at the SPE Annual Technical Conference and Exhibition, Florence, Italy, 19–22 September.

Kovari, K., Tisa, A., Einstein, H.H., et al. 1983. Suggested Methods for Determining the Strength of Rock Material in Triaxial Compression: revised version. International Journal of Rock Mechanics & Mineral Science 20 (6): 283–290.

Logan, D.L. 2010. A First Course in the Finite Element Method. Stamford, Connecticut: Cengage Learning. Mao, W., Fletcher R., and Deregowski, S. 2000. Automated interval velocity inversion. 2000 SEG International Conference and Exhibition,

Calgary, Canada, Aug. 6-11.

IPTC 17115 7

Morales, R.H. and Marceinew, R.P. 1993. Fracturing of High-Permeability Formations: Mechanical Properties Correlations. Paper SPE 26561 presented at the SPE Annual Technical Conference and Exhibition, Houston, Texas, 3–6 October.

Pagoulatos, A. 2004. Evaluation of Multistage Triaxial Testing on Berea Sandstone, MS Thesis, University of Oklahoma, Norman, Oklahoma.

Plumb, R.A. 1994. Influence of Composition and Texture on the Failure Properties of Clastic Rocks. Paper SPE 28022 presented at the SPE/ISRM Rock Mechanics in Petroleum Engineering Conference, Delft, The Netherlands, 29–31 August.

Plumb, R., Edwards, S., Pidcock, G., et al. 2000. The Mechanical Earth Model Concept and Its Application to High-Risk Well Construction Projects. Paper SPE 59128 presented at the IADC/SPE Drilling Conference, New Orleans, Louisiana, 23–25 February.

Qiu, K., Marsden R., Solovyov Y., et al. 2005. Downscaling Geomechanics Data for Thin Beds Using Petrophysical Techniques. Paper SPE 93605 presented at the SPE Middle East Oil & Gas Show and Conference, Bahrain, 12–15 March.

Qiu, K., Marsden, J.R., Alexander, J., et al. 2006. Practical Approach To Achieving Accuracy in Sanding Prediction. Paper SPE 100944 presented at the SPE Asia Pacific Oil & Gas Conference and Exhibition, 11-13 September, Adelaide, Australia.

Qiu, K., Felgueroso, J.G., Lalinde, G., et al. 2008. Geomechanics Enables the Success of Horizontal Well Drilling in Libya: A Case Study. Paper IADC/SPE 111384 presented at the IADC/SPE Drilling Conference, Orlando, Florida, USA, 4–6 March.

Sayers, C.M., Johnson, G.M., and Denyer, G. 2002. Predrill Pore Pressure Prediction Using Seismic Data. Geophysics 67 (4): 1286–1292. Suarez-Rivera, R., Deenadayalu, C., Chertov, M. et al. 2011. Improving Horizontal Completions on Heterogeneous Tight Shales. Paper

CSUG/SPE 146998 presented at the Canadian Unconventional Resources Conference, Calgary, Alberta, Canada, 15–17 November. Zienkiewicz, O.C. and Taylor, R.L. 2005. The Finite Element Method for Solid and Structural Mechanics, Sixth Edition. Elsevier.

1D MEM

Geological Modeling

SCVA Velocity Analysis

1D-MEM Building

3D Mechanical Properties

Pore Pressure 

Prediction

StressInitialization 3D MEM

Geomechanical Core Test

PetrophysicalEvaluation

WellPlacement

FractureOptimizationGridding

Fig. 1 — The workflow of 3D reservoir geomechanics for tight reservoirs.

Ground floor

K1tg

T1b

J1b

P3w

P1j2

Figure 2 — Geological horizons.

8 IPTC 17115

(a) Inline (b) Crossline

Fig. 3 — The vertical cross-section of SCVA velocity result across Well G3.

Fig. 4 — The comparison of SCVA velocity to compressional slowness logs from study wells.

IPTC 17115 9

Fig. 5 — Validation of 1D MEM for Well G3.

Fig. 6 — Validation of 1D MEM for Well G2.

10 IPTC 17115

Fig. 7 — The grid of 3D model (the horizontal spacing of the grid is evenly 50mx50m, the vertical grid lines are intentionally omitted to avoid overshadowing the rest information in the figure).

Top of reservoir

Fig. 8 — Vertical section of minimum horizontal stress across Well G2.

Days

Cum

. Pro

duct

ion

(MM

m3)

Fractures

No. of Fractures: 32-34

Cu

m. P

rod

uctio

n (

MM

m3

)

(a) Predicted cumulative production vs. time (b) Cumulative production vs. number of fractures

Fig. 9 — Optimization of number of stage.

IPTC 17115 11

Days

Cum

. Pro

duc

tion

(M

M m

3)

Half Length

Half length: 220-240m

Cum

. P

roductio

n (

MM

m3)

(a) Predicted cumulative production vs. time (b) Cumulative production vs. half-length of fractures

Fig.10 — Optimization of fracture half-length.

RQ CQ Default Optimized

Fig. 11 — Optimization of staging.