--> Abstract: Numerical Simulation of Reservoir Structures, Part I: Rheology of Reservoir Rocks, by S. Busetti, K. Mish, and Z. Reches; #90090 (2009).

Datapages, Inc.Print this page

Numerical Simulation of Reservoir Structures, Part I: Rheology of Reservoir Rocks

Busetti, Seth 1; Mish, Kyran 2; Reches, Ze'ev 1
1 School of Geology and Geophysics, University of Oklahoma, Norman, OK.
2 School of Civil Engineering and Environmental Science, University of Oklahoma, Norman, OK.

We simulated rock mechanics tests to model rock rheology for use in numerical simulations of the development of reservoir structures. Experiments show that during folding, fracturing, and faulting of the upper crust, rocks progress from quasi-linear elastic to non-linear elastic behavior. In order to solve complex mechanical processes under realistic in-situ conditions, we combine experimental and field observations with finite element simulations. In a series of three abstracts presented in this meeting we describe our efforts using the code Abaqus. Part I describes the elastic-plastic rheology of damaged rocks and implementation in numerical simulations of experiments with Berea sandstone. In the companion abstracts we apply the rock rheology to hydraulic fracturing (Part II) and ramp-folding (Part III) problems.

Experiments show that rocks are weakened by stress-induced damage that coincides with the non-linear portion of the stress-strain curve. Typically, this curve displays four stages: 1) elastic (linear or non-linear); 2) strain hardening and the onset of microcracking; 3) crack coalescence; and finally 4) strain softening and fracture propagation. Besides plasticity, non-linear elastic and visco-elastic rheology have been used as proxies to accommodate large deformation during strain hardening.

We converted stress-strain and damage characteristics for several reservoir rocks into numerical material models. Failure conditions are described by a yield surface that is curved in tension (failure occurs at 90°). We used fracture maps and acoustic emissions data to set the finite element damage parameter, d, a stiffness multiplier defined as the ratio of damaged to intact rock. Our benchmark tests are for Berea sandstone: four-point beam bending (Weinberger et al., 1994) and dog-bone shaped samples under triaxial extension (Ramsey and Chester, 2004; Bobich, 2005). The simulations agree with experiments for confining pressures, Pc, of 10 to 150 MPa and reflect the four stages of damage mentioned above. For Pc < 60 MPa, the onset of damage occurs at 0.05-0.1% extensional strain and increases exponentially until failure at 0.1-0.3%. Prior to first fracture, stiffness degrades 5-10%. Multiple fractures occur in regions of 15-20% stiffness reduction. In the dog-bone setup, damage widens as Pc increases. Results of simulations suggest that the rheology can be utilized in more complex problems.

This work is supported by funds from ConocoPhillips

 

AAPG Search and Discovery Article #90090©2009 AAPG Annual Convention and Exhibition, Denver, Colorado, June 7-10, 2009