Geomechanical property evaluation of the Pennsylvanian Strawn-Canyon Group using machine-learning methods: Insights from the Salt Creek Field, Midland Basin, Kent County, West Texas

Authors:
Publisher website
Google Scholar
Abstract
Geomechanical properties of rock are critical in determining fracture shapes and designing hydraulic fracturing procedures. These properties are usually obtained from static laboratory measurements but are limited to borehole locations. However, in the absence of static properties, dynamically derived properties from surface seismic data can be used. By integrating 3D poststack seismic and wireline-logs, and using machine-learning procedures, we generated three dynamic properties: namely, Young’s modulus (E), Mu-Rho (MR), and brittleness (BRI) to characterize the Salt Creek carbonate reservoir located in the Midland Basin, Kent County, Texas. First, we generated a 1-D model of elastic properties at each well location. Next, we used model-based seismic inversion to generate acoustic impedance and then used probabilistic neural networks (PNN) and multiattribute analysis methods to generate S-wave impedance, and density volumes. These three parameters were then used to compute E, MR, and BRI and to generate crossplots to determine relationships between them. Finally, we generated time slices to examine the lateral distribution of each geomechanical properties. Our results show that (1) E and MR exhibit a very strong linear relationship with an R2 of 0.96, whereas the relationship between E and BRI has an R2 of 0.71. (2) On time slices through the computed volumes, porous zones can easily be identified, characterized by values ranging from 22–40 GPa on the E map, 21–45 GPa on the MR map, and from 25–47 GPa on the BRI map. (3) Because each of these properties can be correlated well with photoelectric-factor (PEF) logs, and because PEF log values can indicate calcite richness and porosity within a mixed carbonate-siliciclastic system, any of the three geomechanical properties, particularly E and MR, can be used to identify rigid rock layers within the reservoir and to deduce porous zones and possible lithologies.

This publication has 9 references indexed in Scilit: