Total Organic Carbon Prediction from well Logs Data using the Multilayer Perceptron with Levnebreg Marquardt Algorithm with an Application to Shale Gas Reservoirs

The recognition of distinct rock types through log responses, referred to as electrofacies, is a fundamental role in mapping stratigraphic units that do not have any specific geological description. Lateral variability within adjoining intervals is differentiated by studying lithological characteristics such as petrography and mineralogy acquired from visual core description. In non-cored wells electrofacies analysis, therefore, is the most reliable way in determining reservoir zonations. The electrofacies’ accuracy is critically important in defining potentially desirable rock types for shale gas reservoirs in non-cored intervals, which can be obtained through an analogy of well log responses in identical lithofacies within cored wells. Considering the complexity of making a final prediction due to the unavailability of different well logs covering the whole area, only the gamma-ray log is used in determining electrofacies patterns within the studied shale gas intervals. The electrofacies patterns within identified lithofacies have been studied for the Kockatea Shale, which presented analogous patterns for identical lithological facies. The similarity has allowed for the correlation of lithofacies in cored and non-cored wells, and the evaluation of lithofacies variability and development within various wells. The correlation of the defined electrofacies indicates facies changes across the basin in association with thickening of some lithofacies. The thickest part of the electrofacies is shown at the Dandaragan Trough and the Beagle Ridge. Some electrofacies, however, have disappeared in some parts of these areas, such as lithofacies E in the Beagle Ridge, which is partially replaced by electrofacies C.

Download Full-text

A logging-curve separation scale overlay method for total-organic-carbon estimation in organic-rich shale reservoirs

Interpretation ◽

10.1190/int-2016-0156.1 ◽

2017 ◽

Vol 5 (3) ◽

pp. T387-T398 ◽

Cited By ~ 3

Author(s):

Jingling Xu ◽

Yangyang Li ◽

Baoying Zhang ◽

Lei Xu ◽

Yuxing Qin

Keyword(s):

Organic Carbon ◽

Total Organic Carbon ◽

Shale Gas ◽

Evaluation Method ◽

Gamma Ray ◽

Quantitative Estimation ◽

Well Logs ◽

Reservoir Evaluation ◽

Overlay Method ◽

Shale Reservoirs

Quantitative estimation of total-organic-carbon (TOC) content using well logs is very important for shale-gas reservoir evaluation, especially when core data are limited. Even though many techniques have been developed to calculate TOC from well logs, they all have their limitations and accurate assessment of TOC in organic-rich tight shales is still challenging. We have used an improved log-overlay method for evaluating TOC content in organic-rich tight shale reservoirs by overlying the properly scaled gamma-ray, sonic transit time, and bulk density curves on top of the compensated neutron log (CNL) curve based on the same concept to the original [Formula: see text] technique. These logging curves are overlapped with the CNL in nonsource intervals, and they are separated in organic-rich shale reservoirs. The separation magnitude increases along with the increase in TOC content of the shale, and this relationship is transformed to calculate TOC. This method was tested and verified by doing a case study using well-log data from the Jiaoshiba tight shale-gas play in the Sichuan Basin. The results illustrated that the new TOC evaluation method is more practical and effective compared with existing TOC evaluation methods.

Download Full-text