Total organic carbon (TOC) is the most significant factor for shale oil and gas exploration and development which can be used to evaluate the hydrocarbon generation potential of source rock. However, estimating TOC is a challenge for the geological engineers because direct measurements of core analysis geochemical experiments are time-consuming and costly. Therefore, many AI technique has used for TOC content prediction in the shale reservoir where AI techniques have impacted positively. Having both strength and weakness, some of them can execute quickly and handle high dimensional data while others have limitation for handling the uncertainty, learning difficulties, and unable to deal with high or low dimensional datasets which reminds the “no free lunch” theorem where it has been proven that no technique or system be relevant to all issues in all circumstances. So, investigating the cutting-edge AI techniques is the contribution of this study as the resulting analysis gives top to bottom understanding of the different TOC content prediction strategies.
TopIntroduction
For gas and oil exploration, it is critical to evaluate source rock property accurately. The abundance of organic carbon can be represented by Total organic carbon as a basic and important index (Passey, Creaney, Kulla, Moretti, & Stroud, 1990),(King, 2010). Generally, from core laboratory analysis, this parameter is obtained which are time-consuming and expensive (Delvaux, Martin, Leplat, & Paulet, 1990; Hare et al., 2014; Johannes, Kruusement, Palu, Veski, & Bojesen-Koefoed, 2006). This constrains the quick advancement of unconventional oil and gas exploration.
Then again, permeability, thermally mature, total organic carbon (TOC) content, porosity, saturation, mechanical properties and rock mineralogy etc. define the productivity of shale quality. Further, reservoir properties which are critical can judge qualitatively of most productive shale reservoir which are commercially potential typically has at least 2% TOC and Ro (more than 1.4 in gas dry window). For a good oil and gas flow capability and storage, it needs under 40% saturation and 100 nano-darcy permeability and over 2% porosity. Further, low differential stress, a certain degree of natural fractures and over 40% quartz or carbonate in mineralogy is needed additionally for commercial shales (Sondergeld, Newsham, Comisky, Rice, & Rai, 2010). For basic and important index, TOC content is the one among the all factors representing the organic matter.
Well logging and direct geochemical analysis are utilized conventionally for TOC determination in the present petroleum industry. Whatever, core data for TOC are not available due to the time and cost required for testing and the difficulties related to gather occasion an intact and representative sample. Despite the fact that laboratory test of TOC is difficult, they are still the preferred and necessary techniques (Guangyou, Qiang, & Linye, 2003; Jarvie*, Jarvie*, Weldon*, & Maende*, 2015).
For further prediction, these lab results are regularly applied as references for the mathematical approaches. With the fast advancement of unconventional exploration of gas and oil, the accurate and continuous study on the TOC is vital. High longitudinal resolution portrays well logging. For the fact of giving continuous TOC profiles that cover the entire interval of interest when continuity of the information and log-based TOC prediction are all the more universally applicable. By comparing with surrounding rocks, some specific geophysical responses (e.g., resistivity, neutron, density, and gamma-ray) of the source rock can be recognized. During utilization of log-based TOC prediction the empirical mathematical equations are commonly utilized. Notwithstanding, equation quality are incredibly dependent by the estimation result by logging data. Meanwhile, the gamma-ray and uranium correlation technique are some of the time not reasonable for shale reservoir. Having radioactivity in phosphatic fish plates in shale reservoir elevated gamma-ray and uranium counts can’t reflect TOC (Bessereau, Carpentier, & Huc, 1991; Passey et al., 1990).
A complicated non-linear function relationship is seen between the TOC content and the logging data. Between the logging data and the TOC content a complicated relationship is seen which are non-linear. By using simple linear regression, approximating the real function relationship is hard and utilizing the well log, it is impossible to predict the TOC content. Nowadays, most researchers have been attracted by AI. The real research demonstrates that AI strategies have extremely solid approximation abilities to non-linear implicit functions. On the other hand, the prediction of TOC content has been worked by AI strategies revealed by the current research result. Between TOC content and the logging parameters a correlation models have been established by utilizing the NN so as to accomplish a good prediction of TOC content. Indeed, the utilization of robust AI techniques approaches have been presented and effectively utilized in numerous engineering fields, for example, permeability estimation, lithology classification et al. (Shadizadeh, Karimi, & Zoveidavianpoor, 2010; Yuan, Zhou, Song, Cheng, & Dou, 2014; Zoveidavianpoor, Samsuri, & Shadizadeh, 2013). Table 1 abridge current TOC prediction strategies utilizing well logs.