Effect of Pressure-Dependent Natural-Fracture Permeability on Shale-Gas Well Production

Transcription

1 Effect of Pressure-Dependent Natural-Fracture Permeability on Shale-Gas Well Production Erdal Ozkan Colorado School of Mines Based on SPE159801, by, Cho, Y., Apaydin, O. G., and Ozkan, E. 1

2 Motivations Production from tight shale reservoirs requires a network of fractures Fractures in the stimulated reservoir volume are poorly propped, if at all Then, they should lose conductivity when pressure drops during production Productivity loss over the life of the well is sometimes attributed to an increasing skin effect due to closing fractures around the well 2

3 Objective Study the effect of stress-dependent natural-fracture permeability on the productivity of shale reservoirs 3

4 Methodology Experimental study using Bakken core samples to relate fracture permeability to pressure Compilation and screening of the existing correlations in the literature for use in our research Analytical study to incorporate pressure-dependent fracture permeability into a dual-porosity reservoir performance model Analysis of model and field data to understand the effect of fracture permeability reduction on reservoir performance 4

5 Assumptions & Simplifications The objective is not to model the stress dependency of shale natural fractures in its full complexity. Our interest is limited to obtaining a first approximation with minimum data requirement. Due to the uniaxial stress and isotropic permeability assumptions, effective stress can be incorporated in terms of pressure. The pressure-dependency of the matrix permeability is negligible compared with the natural-fracture permeability 5

6 Experimental Study Middle-Bakken core samples (from 9,026 ft) in the Williston Basin of North Dakota Measurements by CMS-300 Automated Permeameter (Core Labs, 2012) All cores were cut cylindrically into 1 diameter and 1 length The cores were cut mostly perpendicular to the bedding plane 6

7 Experimental Study To simulate natural fractures, cores were cut vertically at the center. The fractured cores were then held together with Teflon tapes. Confining stress was increased gradually From 1,000 psi to 5,000 psi by 1,000-psi increments to study the effect of stress on fracture closure. 7

8 Experimental Study The effective fracture permeability was measured at three different confining stresses. Fracture permeability, porosity, and width were calculated using the following equations: 8

9 Experimental Study Bakken Core #1 Fracture Permeability, D Fracture Porosity Confining Stress, psia Confining Stress, psia Fracture Width, Micron Confining Stress, psia 9

10 Experimental Study Bakken Core #3 Fracture Permeability, D Fracture Porosity Confining Stress, psia Confining Stress, psia Fracture Width, Micron Confining Stress, psia 10

11 Experimental Study Bakken Carbonate Cores #22 & 32 (Alamdari, 2011) 11

12 Experimental Study Example of Experimental Results Bakken Core #1 UNFRACTURED CORE FRACTURED CORE Confining Stress, psia Effective Stress, psia Matrix Permeability, k m, md Confining Stress, psia Effective Stress, psia Effective Fracture Permeability, k eff, md Fracture Permeability, k f, D Fracture Width, w f, micron Fracture Porosity, Φ, fraction E-04 1, E-04 1,726 1, E-04 1,992 1, E-04 2,435 2, E-04 2,992 2, E-04 4,007 3, E-04 4,478 4, E-04 12

13 Selected Correlations in Literature 13

14 Selected Correlations in Literature Correlations presented in the literature are mainly for pressure-dependent rock-matrix (primary) permeability. The constants and coefficients in these equations are available mainly for applications to rock-matrix permeability. Selected correlations were used to match the new experimental data and to determine the new constants and coefficients 14

15 Example: Matching the experimental data with Correlation # 3 15

16 Example: Change in fracture permeability as a function of pressure Correlation #3 16

17 Example: Pressure history match of the Haynesville shale-gas well (Wang and Liu, 2011 ) 17

18 Example: Pressure history match of the Haynesville shale-gas well (Anderson et al., 2010) 18

19 Example: Pressure history match of the Haynesville shale-gas well (Anderson et al., 2010) 19

20 Results of pressure-history match examples Both examples were for overpressured reservoirs and the initial pressure was not known; Initial pressure was a regression parameter. Due to the exponential nature of the fracture permeability correlation, it predicts much larger change in permeability if the initial permeability is taken at a higher initial pressure. In general, the large number of matching parameters used in these examples (which is also the case in most applications) causes severe non-uniqueness problems and reduces confidence in the history matching. 20

21 Synthetic Example 21

22 Synthetic Example 22

23 Sensitivity of Productivity to Natural Fracture Permeability 23

24 Sensitivity of Productivity to Various Parameters 24

25 Conclusions The results of this work indicate the following: 1. Unpropped natural fractures lose a significant portion of their initial permeability under pressure depletion. 2. However, the permeability which is retained in the fractures may still be very large compared with the shalematrix permeability (infinite-conductivity fracture effect) and sufficient to transmit the limited volume of fluid available to flow 3. Hence, fracture closure with pressure drop should not be used to infer the productivity loss as pressure drops without considering the complex interactions between the natural fractures and shale matrix 25