Abstract
Total organic carbon (TOC) as measured by laboratory techniques from core historically has been used to assess the quality of source rocks. Now, TOC measurements are widely used to help evaluate unconventional reservoirs/resource plays and to more optimally target and design lateral wells to achieve maximum productivity. This paper describes a method to estimate TOC from wireline logs and visualize its distribution across the entire Delaware Basin.
Introduction
Hydrocarbon exploitation in the Delaware Basin is currently focused on the Wolfcamp and Bone Spring Formations, however proven productive zones occur in the stratigraphic section from the overlying Delaware Mountain Group to the Ordovician Ellenburger. Known regional source rocks throughout the section include the Ordovician Simpson, Devonian Woodford, Mississippian Barnett, and potential local sources in the Pennsylvanian, Permian Wolfcamp and the Avalon shale member of the Bone Spring. To gain insight into the basin's petroleum systems, we describe a convenient approach to comparatively view source richness and distribution estimates from a basin-wide perspective.
Traditional sample based methodology uses TOC datasets from laboratory measurements that are displayed by plotting and contouring TOC values on a 2D map. This may be appropriate for thin homogenous shale formations, but for thicker heterogeneous source beds maps offer little flexibility in viewing and analyzing the data. More recently, petrophysical methods to derive TOC from wireline logs have been proposed and tested in several basins globally (eg Passey et al., 1990 & 2010, Issler et al., 2002). Those methods were not easily applicable here, the Passey method because of abundant calcite in the sediments and the need for good thermal maturity control, and the Issler method because of the general paucity of good quality sonic log coverage in the Delaware Basin.
Our project derived a petrophysical model to estimate TOC through the Bone Spring and Wolfcamp by calibrating ~1900 core measured TOC values from 57 wells to wireline curves. A calculated TOC curve was generated for each lithostratigraphic unit as appropriate using RHOB as the primary input (continuous DT curves were sparse in the wells with sample data). GR and borehole rugosity cutoffs were applied to constrain the calculation. The model was then applied to 872 wells across the basin and interpolated to provide a 3D volume of estimated TOC. The calculated curves and 3D model were QC'd visually and semi-statistically and found to be a reasonable match to the core data, given the methodology
