A method is described for reservoir characterization in fine-grained and thinly bedded shales based on a probabilistic clustering procedure (PCP) of well logs followed by a forward modeling procedure that results in the calculation of profiles for porosity, water saturation (Sw), and permeability. The credibility of results relies on calibration using porosity, permeability, and mineralogy analyses of core samples. Complementary analysis using a nuclear spectroscopy log, if available, can add confidence to the results.

Kerogen needs to be included during the forward modeling because it is a matrix component of the bulk rock and the matrix grain density (GD) can be significantly affected by kerogen. Kerogen profiles can be estimated if kerogen core analyses are included in the PCP. In addition, a total organic carbon (TOC) profile, which is needed to determine the amount of adsorbed gas, can be estimated based on core analyses of TOC. The relationships between TOC and kerogen, including a discussion of Rock-Eval pyrolysis, are outlined.

The estimation of free gas gross pay in gas shales is fraught with difficulty because of the vagaries of estimating porosity and Sw. Although not realistic in terms of gross storage capacity, the use of the combination of total porosity (TPOR) and total water saturation (Swt) gives the same pay as the combination of effective porosity (EPOR) and effective water saturation (Swe). However, the combination of EPOR and Swt is ill-construed and results in underestimation of gross pay.

The estimation of adsorbed gas in gas shales relies on a methodology and equations adopted from the coalbed methane industry. The workflow is easily implemented, but the credibility of results hinges on the assumption that the adopted methodology and equations are valid for gas shales, and on having sufficient and proper laboratory-derived gas adsorption isotherm measurements to represent the TOC heterogeneity of the reservoir.

An example is given using analysis of a cored well from the Upper Jurassic Haynesville Shale of northwestern Louisiana and northeastern Texas. The analysis generated profiles for TPOR and EPOR, Swt and Swe, permeability, and net feet of pay.