Evaluating Geologic Uncertainty and Its Impact on Log Interpretation for Precise Shale Gas Reservoir Assessment

The subject of this particular research is fractal patterns of mineralogy and lithology together with fluid type distribution on selected key well log parameters including gamma ray, resistivity and porosity logs. Following a quantitative method of research, secondary data on three Shale gas fields: The Marcellus, Barnett, and Haynesville Shales were collected and reviewed to measure the level of interpretational error occasioned by geological ambiguity. The dataset comprised 150 well profiles, each incorporating detailed core samples and $\log$ data at intervals of 0.5 to 1 meter. Quantileregression models and the multivariate analysis were used to estimate the measure of disparity between the log-based and core-based measurements, main sources of variation. The data reveal that interpretation of well logs differs between heterogeneous and homogenous zones from $\mathbf{1 5 - 2 5 \%}$, in terms of porosity, water saturation and lithology classification. More precisely, distribution of quartz and carbonate minerals was established to produce errors of up to $20 \%$ in resistivity and porosity logs. This paper brings into focus the need and scope of combining geological knowledge with enhanced logging approach in order to overcome the risks that may hinder the determination of the actual potential of shale gas reservoirs. These results are important for improving the approaches for exploration and optimizing the resource estimate of the unconventional gas reservoirs.