Comparative study for the interpretation of mineral concentrations, total porosity, and TOC in hydrocarbon-bearing shale from conventional well logs
MetadataShow full item record
The estimation of porosity, water saturation, kerogen concentration, and mineral composition is an integral part of unconventional shale reservoir formation evaluation. Porosity, water saturation, and kerogen content determine the amount of hydrocarbon-in-place while mineral composition affects hydro-fracture generation and propagation. Effective hydraulic fracturing is a basic requirement for economically viable flow of gas in very-low permeability shales. Brittle shales are favorable for initiation and propagation of hydraulic fracture because they require marginal or no plastic deformation. By contrast, ductile shales tend to oppose fracture propagation and can heal hydraulic fractures. Silica and carbonate-rich shales often exhibit brittle behavior while clay-rich shales tend to be ductile. Many operating companies have turned their attention to neutron capture gamma-ray spectroscopy (NCS) logs for assessing in-situ mineral composition. The NCS tool converts the energy spectrum of neutron-induced captured gamma-rays into relative elemental yields and subsequently transforms them to dry-weight elemental fractions. However, NCS logs are not usually included in a well-logging suite due to cost, tool availability, and borehole conditions. Conventional well logs are typically acquired as a minimum logging program because they provide geologists and petrophysicists with the basic elements for tops identification, stratigraphic correlation, and net-pay determination. Most petrophysical interpretation techniques commonly used to quantify mineral composition from conventional well logs are based on the assumption that lithology is dominated by one or two minerals. In organic shale formations, these techniques are ineffective because all well logs are affected by large variations of mineralogy and pore structure. Even though it is difficult to separate the contribution from each mineral and fluid component on well logs using conventional interpretation methods, well logs still bear essential petrophysical properties that can be estimated using an inversion method. This thesis introduces an inversion-based workflow to estimate mineral and fluid concentrations of shale gas formations using conventional well logs. The workflow starts with the construction and calibration of a mineral model based on core analysis of crushed samples and X-Ray Diffraction (XRD). We implement a mineral grouping approach that reduces the number of unknowns to be estimated by the inversion without loss of accuracy in the representation of the main minerals. The second step examines various methods that can provide good initial values for the inversion. For example, a reliable prediction of kerogen concentration can be obtained using the ΔlogR method (Passey et al., 1990) as well as an empirical correlation with gamma-ray or uranium logs. After the mineral model is constructed and a set of initial values are established, nonlinear joint inversion estimates mineral and fluid concentrations from conventional well logs. An iterative refinement of the mineral model can be necessary depending on formation complexity and data quality. The final step of the workflow is to perform rock classification to identify favorable production zones. These zones are selected based on their hydrocarbon potential inferred from inverted petrophysical properties. Two synthetic examples with known mineral compositions and petrophysical properties are described to illustrate the application of inversion. The impact of shoulder-bed effects on inverted properties is examined for the two inversion modes: depth-by-depth and layer-by-layer. This thesis also documents several case studies from Haynesville and Barnett shales where the proposed workflow was successfully implemented and is in good agreement with core measurements and NCS logs. The field examples confirm the accuracy and reliability of nonlinear inversion to estimate porosity, water saturation, kerogen concentration, and mineral composition.