Abstract
Poroelasticity plays a vital role in many applications of rock mechanics in petroleum engineering. To model dual porosity stress sensitive reservoirs such as tight gas, it is necessary to quantify the stress-dependant hydraulic conductivities of their porous components, the matrix and natural fractures. The role of effective stress concept near the wellbore is more pronounced as the reservoir pressure declines rather rapidly approaching the bottom hole flowing pressure. The porous media changes when reservoir pressure is altered and therefore the elastic coefficient (α) becomes a function of reservoir pressure for both components; matrix and fractures.
Rock samples from tight formation were tested to determine the stress-dependant matrix permeability. Various combinations of net effective stresses were applied and the corresponding permeabilities were measured at each stress level. Applying the definition of effective stress with an iterative assumption of Biot's coefficient enabled the determination of (α) as a function of pressure.
Additionally, the Biot coefficient was determined for a natural fracture. A tensile fracture was simulated by splitting a whole core following a Brazilian test procedure. The stress dependent permeability was evaluated under varied effective stresses simulating a reservoir depletion scenario. The concept of effective stress within the natural fracture was totally different than that of rock matrix; therefore, the effective stress function for both matrix and natural fractures were evaluated and representative functions for both media were obtained. These stress-dependent permeability functions can then be used in any rock mechanics application in petroleum engineering. The developed experimental procedure is fairly simple and is discussed in details, and selective relavant applications are presented.
The new technique for determining the Biot coefficient is based on the application it is being used for; the effect of effective stress on matrix and natural fracture permeabilities. This paper shows that two poroelastic coefficients, and not one, should be used in any dual porosity system to obtain reasonable prediction of reservoir performance. Varying the coefficient as a function of pore pressure allowed for determining a pressure dependant function for a given porous medium.
