Laboratory measurements indicate that both in situ shear and normal stressessignificantly affect fracture permeability. Therefore, it is their combinedinfluence on fracture permeability that should be taken into consideration inreservoir modeling as opposed to the standard practice of considering only theinfluence of the normal stress component. Spatial orientation of five of thesix common fold-related fractures in a producing field changes with changes ineither bed stake or bed dip. The spatial orientation of the sixth fold-relatedfracture, the type I extension fracture, changes with changes in bed stake butis not influenced by bed dip. On fold flanks, then, where the in situ stressstate is such that it is only the type I extension fracture on which the shearstress component remains zero independent of bedding dip. However on the otherfive fold-related fractures, the magnitudes of both the in situ shear andnormal stress components vary with bed dip and these variations in magnitude ofshear and normal stresses on fold-related fractures both contribute toanisotropic permeability in the reservoir even if fracture distribution isisotropic.
This study indicates that fracture permeability is sensitive not only to theabsolute values of the in situ shear and normal stress on the fracture planebut, also, depends upon their ratios. If pore pressure is reduced duringproduction, the normal stress on all fractures increases but shear stress onthese fractures remains constant. Therefore, the shear stress-normal stressratio changes with changes in pore pressure during production.
The data suggest that instead of using normal stress/fracture permeabilitycurves to determine fracture permeability in a reservoir it is more appropriateto use a surface plotted in normal stress/shear stress fracture permeabilityspace. Because shear stress/normal stress ratios on fractures vary withvariations in pore pressure, it is important to track fracture permeability onthe normal stress/shear stress/permeability surface throughout productionhistory.