Novel Method for Evaluating Shale Gas and Shale Tight Oil Reservoirs Using Well Log Data

摘要

This paper introduces a novel method for improved evaluation of shale gas reservoirs, shale tight oil reservoirs, and immature organic shale source rocks. The method provides exact and original algebraic equations for determining total porosities, producible fluid volumes, total hydrocarbon volumes, kerogen volumes, and immobile hydrocarbon volumes. The determination of accurate porosities and fluid volumes in organic shale reservoirs is critical for formation evaluation considering the low porosities (3 to 8 p.u.) typically found in shale reservoirs. We analyze the challenges associated with interpreting logging tool responses in poor quality shale reservoirs and propose a suite of logging tool measurements that are most likely to overcome the problems. A small set of logging tool measurements (i.e., three or four) that have known responses to fluid and solid volumes and have negligible surface effects were selected. Thus our approach aims to circumvent the interpretation problems caused by poorly understood complex fluid-surface interactions that are exacerbated in shales by high surface-to-volume ratios and clays and other conductive minerals. The logging tool measurements that we use to predict the shale reservoir properties are bulk densities, nuclear magnetic resonance (NMR) porosities, and total organic carbon (TOC) weight fractions derived from total carbon concentrations measured by a gamma ray spectroscopy tool. The tool response equations for these measurements are written as volume weighted averages of reservoir properties. The producible gas in shale gas reservoirs and the high API gravity producible oils in shale tight oil reservoirs cause the apparent (i.e., measured) NMR porosities to read too low and the apparent density log porosities to read too high. Kerogen also causes the density log porosities to read too high. The tool response equations are solved simultaneously and exactly to determine shale total porosities, fluid volumes, and kerogen volumes. The solution for the shale total porosity is automatically corrected for light hydrocarbon effects on the density and NMR porosity measurements and for kerogen effects on the density porosities. The exact algebraic solutions or "plug-in formulas" for shale reservoir properties are the main results of the paper. The robustness of these solutions in the presence of measurement noise is studied using Monte Carlo simulations. We discuss the standard deviations of the predicted reservoir properties for different measurement noise levels and the effects of errors in the assumed reservoir fluid and solid properties (e.g., gas density, gas hydrogen index, kerogen density) on the accuracies of the predicted reservoir properties. The algebraic solutions are used to predict reservoir properties from log data acquired in a shale gas well and a shale tight oil well. The results in both wells are shown to compare favorably with available core data.