Pulsed Neutron Advances in Through-Casing Evaluation of Multilayered Reservoirs: Case Studies in the Mature Fields of the San Jorge Gulf Basin, Argentina

Abstract

An alternative technique to evaluate layered formations through casing in the Golfo San Jorge basin in Argentina is presented. The proposed methodology is based on pulsed neutron spectroscopy logs to assess the lithology and evaluate the hydrocarbon type and potential using resistivity-independent methods. The basin's layered reservoirs present multiple challenges for resistivity-based methods, including low and variable formation water salinity, intricate pore systems affecting the rock electrical parameters, fine sediments with high irreducible water content suppressing the resistivity response, and high-resistivity tuffaceous sands not associated with hydrocarbons. In cased holes, simple correlations of resistivity and capture cross-section (sigma) or other techniques emulating openhole logs from basic pulsed neutron logs (PNL) can be used for interwell correlations, but are frequently inconclusive for assessing hydrocarbon potential. Our approach takes advantage of latest-generation PNL tools’ multiple measurements to reduce the uncertainty of water saturation assessment. The proposed methodology was tested with two new-generation PNL tools, which feature high-resolution detectors that provide elemental concentrations for better understanding mineralogy including direct measurement of the carbon concentration for a reliable estimation of total organic carbon (TOC), which is directly associated with the oil volume in the pore space. The first tool was originally intended for openhole logging and has a single large detector for high-quality spectroscopy analysis. The second device is a multiple-detector slim PNL tool, which, besides the TOC and other spectroscopy outputs, also provides sigma and neutron porosity and measures a new property, the fast neutron cross section (FNXS), which is useful to detect and quantify gas. The presented case studies include examples of application with both tools in new wells and workovers. More than 20 sands, typically 2- to 5-m thick, interbedded with thick shales, were counted over the 1000- to 1500-m zone of interest, and all the data could be acquired in an acceptable time frame in spite the long intervals. The log data were complemented with the mud logging information and/or correlations from the static geological model. In all cases, we achieved good correlation between the zones with TOC, the static model, and well test results. This experience illustrates the adaptation and application of new technologies to the development of mature fields where conventional openhole resistivity-based analysis is ambiguous. Future tasks include adjusting the technique for quantitative analysis and its use in deeper, more challenging, unconventional reservoirs in the basin.