Interval inversion as innovative well log interpretation tool for evaluating organic-rich shale formations

Abstract

The identification and detailed evaluation of unconventional hydrocarbon reservoirs require the application of advanced well logging data processing techniques. Since the number of the petrophysical properties such as porosity, water and hydrocarbon saturation, kerogen content, and fractional volumes of matrix components may exceed that of the observed wireline log types, currently used quick-look methods based on few well logs provide approximate solution to a part of the above petrophysical quantities. For instance, Passey's method basically utilizes two well logs for identifying the sweet spots and estimating the total organic content (TOC). Local inversion techniques estimating the petrophysical parameters in a given depth are also limited because of low overdetermination (data-to-unknowns) ratio and hence are very sensitive to data noises. Instead of using traditional empirical methods and to avoid ambiguous solutions by local inversion evaluations, data of all wireline logs measured in a longer depth interval are simultaneously inverted to give more accurate estimate to the petrophysical parameters in a highly overdetermined inverse problem. The suggested interval inversion approach allows the determination of the fractional volumes of rock constituents with their estimation errors in a joint inversion procedure. As part of the inversion process, the forward problem is improved by using kerogen and pyrite effects corrected probe response functions. Synthetic modeling tests are made to detect the noise sensitivity and outlier resistance of the proposed inversion algorithm, while a float-encoded genetic algorithm variant of the interval inversion method is applied to invert real data measured in an organic-rich shale reservoir in Alaska, USA. The feasibility is demonstrated by a well agreement between the inversion derived TOC values and those of measured on core samples. The interval inversion method not only works reliably in the presented shale gas formations, but it can be extended to evaluating other types of unconventional reservoirs.