The entire process is grounded in careful petrophysical work. Using elastic parameters from prestack seismic inversion, probability distribution functions (PDFs) are ‘learned’ and then, using Bayesian probability theory, combined with prior knowledge about the reservoir petrofacies proportions to generate probability volumes for each petrofacies and a most probable petrofacies (MPP) volume. The approach is to first establish petrophysical rule-based petrofacies definitions, and then demonstrate where – and with what probability – these petrofacies classes are met within a seismic volume (Doyen, 2006 and Coulon et al., 2006). Lithology Prediction Workflowįigure 1 illustrates the lithology prediction workflow in general terms. This work was originally presented at the 2011 CSPG CSEG CWLS Convention. In this article, we describe a workflow to do just that: apply rule-based petrofacies to seismic to predict lithology, potentially improving well placement and completions in the Montney shale gas play. Extrapolating this petrophysical knowledge to seismic data for inter-well lithology prediction can be an important tool for developing the resource, but only if the predicted lithology remains anchored to real-world petrophysical measurements. While petrofacies can be determined carefully for each well in the field using log data integrated with core and sampled data, there is still the issue of predicting lithology between wells. Field development is enhanced by the ability to spatially predict different rock-types which in turn have different properties such as fracability and permeability. In any oil or natural gas development – conventional or unconventional – it is important to understand lithology, as it is a fundamental building block for all reservoir characterization work.
0 kommentar(er)
