Petroleum geochemistry is nowadays used as an essential knowledge to increase the probability of discovering hydrocarbon resources. It is accepted that geochemical methods significantly increase the chances of success in oil exploration. Among these methods, the pyrolysis family of methods are among the most accurate methods for determining the amount of organic carbon in the rocks. A standard method is used. Since petrophysical data is available in most drilled wells, extracting geochemical data from petrophysical data will be a valuable time and cost-saving. The fields studied in this study are Ahwaz, Ramshir and Rag-e Sefid. These fields are located from the center to the southeast of Khouzestan province, respectively, and are among the most potential fields in southwest Iran for more detailed research to estimate organic carbon content. The purpose and motivation of this study is to compare and compare the methods of ?logR, NMR log, mineralogical data and neural network for calculating total organic carbon in source rock, which is a higher priority method that is sufficiently similar to the results of pyrolysis experiments. Rock eval is closer and more accurate and relatively inexpensive. In this research software work has been done through IP software which has been used to calculate the petrophysical data of each field and finally its results are compared with real TOC values. Besides, in the NMR log method, the corresponding comparison is made through reports and logs drawn from the Ahwaz field with actual TOC values in this field. The neural network method is also implemented by coding in MATLAB software, the input data from petrophysical and real data and finally the output values are evaluated. Input data in the mineralogical data method including density logs, neutron porosity logs and gamma logs, in the neural network method, include logs that have a good correlation with the output data and do not cause network deviation. The input data in the ?logR method include acoustic and resistivity logs and in the NMR log method, the most important input data are the NMR log comfort times. The most suitable methods (considering R) in Ahwaz, Ramshir and Rag-e Sefid fields, are mineralogical data, ?logR method, respectively. Finally, the best method discussed in this research is the mineralogical data method with an R2-value of 0.94, 0.78 and 0.61 in Ahwaz, Ramshir, and Rag-e Sefid, respectively, and 0.78 in general. After this approach, neural network, ?logR and NMR methods are ranked second, third and fourth respectively.
