Abstract - Progressive stages in petroleum exploration and exploitation require different considerations about faults, due to differences in flow processes and physical forces operative. Outcrop studies show that fault properties can vary markedly over short distances on a fault trace, and a consideration of this variability is important for making correct inferences about the effects of faults on flow in hydrocarbon systems. Hydrocarbon accumulation in fault-bounded traps depends on the properties of the hydrocarbon, the hydrodynamics of the basin and the properties of the fault. In the hydrostatic case, the column height trapped by a sealing fault is determined by the minimum capillary entry pressure of the fault; this pressure must be matched by the hydrocarbon if it is to establish a migration path through the pore spaces in the fault. An average fault property is of no value in determining the trapping potential of a fault, as leakage can occur through a focused flow path over geological time. In mature hydrocarbon provinces, empirical correlations between the limit capillary pressure and specific details of the faulted succession observed in proven accumulations, allows an estimation of the likely column height in an undrilled prospect.
Migration occurs at extremely low flow rates and pressure gradients, and is dominated by capillary and gravity forces. In contrast, production is achieved through exploiting natural or imposed fluid pressure gradients, and viscous forces are important. Flow rates are linked to pressure gradients through Darcy’s law, which contains viscosity and permeability terms. Hence fault permeability, which is not significant for migration studies, is important in production situations. Fault surface averages can be used to determine the appropriate permeability of a fault for incorporation in a production flow simulation model. An analytical consideration shows that the arithmetic average fault permeability and the harmonic average fault thickness yield the correct fault transmissibility. These averages are sensitive to extreme high and low values respectively, and flow is concentrated through these extremes. An example with realistic heterogeneity shows that 50% of the fault surface contributes 3% of the flow, and that 50% of the flow passes through less than 10% of the fault surface. The heavy influence of extreme fault properties highlights the importance of collecting representative samples at outcrop, and illustrates the large uncertainty in determining the flow through any particular fault in the sub-surface.
Abstract of talk given to:
Tectonic Studies Group Annual Meeting, University of Durham, December 1997