In the Rocky Mountain region of the US, nearly every well is hydraulic fracture stimulated to produce commercial volumes of oil and gas. The starting point for designing these treatments is an understanding of the in-situ stress profile. To calculate the in-situ stress profile, one must have an understanding of the mechanical rock properties and the pore pressure variations throughout the wellbore. Pore pressure can be measured in the permeable zones and in-situ stress can be calculated by the modeling of closure stress from pre-fracture pressure testing. But these tests are rarely performed in the nonreservoir rocks above and below the fracture stimulation treatment. The challenge for the stimulation design engineer is to determine the mechanical rock properties in and around the treatment interval. Calculating in-situ stress with the uniaxial strain equation requires the knowledge of Poisson's ratio, Young's modulus, pore pressure, and overburden pressure. Classically, the static Poisson's ratio (PR) and Young's modulus (YMS) are calculated by using the results from a dipole sonic log. However, in most fields, less than 1% of the wells requiring stimulation have dipole sonic data.
In the absence of dipole sonic information, conventional wireline log data can be used to quite effectively calculate mechanical rock properties by using basic petrophysical relationships and artificial neural networks. The composite method to determine mechanical rock properties uses five methods to estimate the compressional slowness (DTC), seven methods to estimate shear slowness (DTS), eight methods to estimate PR, and and nine methods to determine YMS. The composite model PR and YMS are error minimized by a weighted averaging technique that honors the most reliable correlations. The composite modeled PR and YMS values are then used as inputs for a continuous calculation of the minimum horizontal stress (Shmin).
The validation of the new composite model PR and YMS was confirmed in three ways. First, the composite model results are compared to the values determined in the lab from actual core samples. Second, the composite model results were compared to calculations using measured DTC and DTS in wells where dipole sonic tool measurements were recorded across a field in Wyoming. Finally, the composite model results were validated by using stimulation treatment pressure history matches on the Pinedale anticline in southwestern Wyoming. The goal of the composite model is to provide a robust rock property solution with or without sonic log data to eliminate the mechanical rock properties as a variable in fracture stimulation treatment design and history matching.
In the world of hydraulic fracture stimulation modeling, the starting place is always the determination of the mechanical rock properties used to derive the minimum horizontal stress profile. Where does one get such data? In an ideal world, the easy answer is to calculate them by using the compression and shear slowness from a dipole sonic log. In the real world, however, dipole sonic logs are not a common tool included in the typical wireline logging suite. What does one do in the frequent case in which no sonic data is available?
Most stimulation design models have some type of generic table of rock properties, based on a lithologic description or rock type. A stress profile built using this approach can easily lead to gross oversimplification of the stress profile throughout the wellbore. What about using multiple petrophysical relationships and a statistical error minimization approach to estimate the rock properties using conventional wireline log data?