Fig 5 - uploaded by Linqi Zhu
Content may be subject to copyright.
Correlation between well logging response and TOC In Fig. 5, the abscissa represents determination coefficient; the ordinate in RD represents the response value of deep resistivity curve; TH represents the response curve of thorium content value; KTH represents uranium gamma curve response value; CNL represents neutron log response value; PE represents the value of photoelectric absorption index log response curve; GR represents gamma-ray logging curve value; AC represents wave slowness response curve; U represents the response curve of uranium content value; DEN represents the density curve of response value. It can be seen from that TOC and the density value have the biggest correlation. This is because density of kerogen (a component of TOC) is much smaller than that of other lithology. Where content of TOC is higher, density of the rock must be smaller. Of course, other curves are more or less correlated with TOC. Although there are good single correlation between some well logging curves and TOC, the goal of accurately predicting TOC still cannot be achieved. Through correlation analysis, uranium content curve (U), density curve (DEN), natural gamma curve (GR), acoustic curve (AC), neutron curve (CNL), and photoelectric absorption eoss-section index (PE) were selected as input curves; the content of TOC was taken as output. They were all normalized. Samples were randomly grouped (the samples were divided into 4 groups, each containing 33 samples.). The network parameters were all determined by the K-CV method in order to obtain an ideal fitting-model (Jiang et al 2015). In order to facilitate the comparison with other nonlinear approximation models, we attempted to write Adaboost-BP neural network model, SVM model, Kernel Extreme Machine (KELM) model and in the MATLAB software, and obtained model parameters using the samples
Source publication
There is increasing interest in shale gas reservoirs due to their abundant reserves. As a key evaluation criterion, the total organic carbon content (TOC) of the reservoirs can reflect its hydrocarbon generation potential. The existing TOC calculation model is not very accurate and there is still the possibility for improvement. In this paper, an i...
Contexts in source publication
Context 1
... order to compare the models conveniently, the measurement of the TOC was carried out on 132 rock samples from the Longmaxi shale gas reservoir of the Jiaoshiba area in the Sichuan Basin (the first large shale gas field in China). With the sample points of the abnormal response value removed, the simple linear correlation of the log response and the TOC was analyzed using a crossplot single linear fit module in the commercial software Excel 2003; the cross-plot was drawn based on this result (as shown in figure 5). ...
Context 2
... figure 5, the abscissa represents the determination coefficient, the ordinate in the RD represents the response value of the deep resistivity curve, TH represents the response curve of the thorium content value, KTH represents the uranium gamma curve response value, CNL represents the neutron log response value, PE represents the value of the photoelectric absorption index log response curve, GR represents the gamma ray logging curve value, AC represents the wave slowness response curve, U represents the response curve of the uranium content value and DEN represents the density curve of the response value. It can be seen from figure 5 that the TOC and the density value have the biggest correlation. ...
Context 3
... figure 5, the abscissa represents the determination coefficient, the ordinate in the RD represents the response value of the deep resistivity curve, TH represents the response curve of the thorium content value, KTH represents the uranium gamma curve response value, CNL represents the neutron log response value, PE represents the value of the photoelectric absorption index log response curve, GR represents the gamma ray logging curve value, AC represents the wave slowness response curve, U represents the response curve of the uranium content value and DEN represents the density curve of the response value. It can be seen from figure 5 that the TOC and the density value have the biggest correlation. This is because the density of kerogen (a component of TOC) is much smaller than that of other lithology. ...
Context 4
... order to compare the models conveniently, the measure- ment of the TOC was carried out on 132 rock samples from the Longmaxi shale gas reservoir of the Jiaoshiba area in the Sichuan Basin (the first large shale gas field in China). With the sample points of the abnormal response value removed, the simple linear correlation of the log response and the TOC was analyzed using a crossplot single linear fit module in the commercial software Excel 2003; the cross-plot was drawn based on this result (as shown in figure 5). ...
Context 5
... figure 5, the abscissa represents the determination coefficient, the ordinate in the RD represents the response value of the deep resistivity curve, TH represents the response curve of the thorium content value, KTH represents the ura- nium gamma curve response value, CNL represents the neutron log response value, PE represents the value of the photoelectric absorption index log response curve, GR represents the gamma ray logging curve value, AC represents the wave slowness response curve, U represents the response curve of the uranium content value and DEN represents the density curve of the response value. It can be seen from figure 5 that the TOC and the density value have the biggest correlation. ...
Context 6
... figure 5, the abscissa represents the determination coefficient, the ordinate in the RD represents the response value of the deep resistivity curve, TH represents the response curve of the thorium content value, KTH represents the ura- nium gamma curve response value, CNL represents the neutron log response value, PE represents the value of the photoelectric absorption index log response curve, GR represents the gamma ray logging curve value, AC represents the wave slowness response curve, U represents the response curve of the uranium content value and DEN represents the density curve of the response value. It can be seen from figure 5 that the TOC and the density value have the biggest correlation. This is because the density of kerogen (a comp- onent of TOC) is much smaller than that of other lithology. ...
Similar publications
The shale gas potential of the early Cretaceous Talhar shale member of the lower Goru Formation in lower Indus basin of Sanghar district, Sindh province, has been evaluated using 2D seismic and well logs data. Graphical and empirical techniques were applied to calculate various geochemical parameters for the evaluation of shale plays. Total Organic...
Total organic carbon (TOC), a critical geochemical parameter of organic shale reservoirs, can be used to evaluate the hydrocarbon potential of source rocks. However, getting TOC through core analysis of geochemical experiments is costly and time-consuming. Therefore, in this paper, a TOC prediction model was built by combining the data from a case...
East Beni Suef Concession is located in the eastern part of Beni Suef Basin which is a relatively under-exploration basin and lies about 150 km south of Cairo city. Rock-Eval pyrolysis and well logs data from three wells (Gharibon-1x well, Gharibon NE-1X and Tareef-1X well) in the Gharibon oil field, East Beni Suef basin were used to evaluate the o...
Our analysis of a comprehensive well log database and complementary mineralogical and geochemical information indicates that the risk for Upper Jurassic shales on the Norwegian Continental Shelf (NCS) to permit severe leakage of hydrocarbons from the reservoir is generally low, even in the case of substantial uplift. The content of brittle minerals...
The evaluation of the exploratory potential for shale oil/shale gas of a source rock involves the characterization of three key factors: existence of high TOC, adequate maturity and susceptibility to fracking. Additional factors such as depositional and geological control over the distribution of organic matter in the source rock should also be inv...
Citations
... Numerous studies have made efforts to make connections between the geochemical parameters and well log responses (Fertl and Rieke 1980;Passey et al. 1990;Huang and Williamson 1996;Bordenave and Huc 1995;Alizadeh et al. 2012b;Sfidari et al. 2012;Mahmoud et al. 2017). Recent investigations are adopted as useful methods for the evaluation of the nonlinear relationship between TOC and well logs (Lewis et al. 2004;Kadkhodaie-ilkhchiy et al. 2009;Tan et al. 2015;Alizadeh et al. 2018;Bolandi et al. 2017;Zhu et al. 2018). According to the prior studies of the complicated and nonlinear relationships between TOC and well logs, an artificial neural network (ANN) is more applicable and advanced to evaluate TOC values. ...
Total Organic Carbon (TOC) is one of the most important geochemical parameters in source rock evaluation, utilized to characterize the hydrocarbon generation potential. Artificial intelligence (AI) methods have significant features, including decreased time and high cost-effectiveness, providing a sufficient database, reducing risk, and enabling better decision-making. AI methods based on these significant features have become efficient tools in various approaches to geological studies such as prediction and classification. This study uses an artificial neural network (ANN), and ensemble algorithms including random forest, bagging, and least-square boosting to evaluate the organic richness of the Kazhdumi Formation in the Persian Gulf. The Kazhdumi Formation with Cretaceous age is well-known as one of the most important source rocks in the Zagros and NW of the Persian Gulf. The obtained results of Rock-Eval pyrolysis and conventional well logs e.g., sonic, neutron, density, spectral gamma-ray, resistivity, and acoustic impedance are utilized for TOC estimation. The performance of AI algorithms was analyzed using root mean squared error (RMSE) and mean absolute error (MAE) equations and the obtained results from the error evaluation analysis show that the least square boosting (LSB) algorithm has the lowest error than the other proposed algorithms in this investigation. The cross-validation analysis between the determined TOC values of blind samples of Rock-Eval pyrolysis and the AI algorithms e.g., the ANN and random forest, bagging, and least-square boosting are evaluated 0.89, 0.90, 0.94, and 0.95, respectively. The geochemical study of the Kazhdumi Formation as a potential source of rock shows that this formation has a good potential for hydrocarbon generation. According to the geochemical evaluation based on the cross-plot analysis, the Kazhdumi Formation contains type II, and type II-III kerogen, which is affected by the sedimentation conditions of this formation. This investigation has confirmed the ability of ensemble methods for TOC value estimation, and it can further be applied to investigate other geological properties.
... TOC, which corresponds to the organic richness of the rock, is determined using core testing, wireline logging curves, or by combining the two methods . Zhu et al. (2018) used the Passey et al. (1990) method, coupled with results from X-ray diffraction (XRD) core analysis and mineral models, to estimate organic-rich source rock. G total was typically obtained from core data analysis through the element capture spectrum (ECS) coupled with the formation micro-imager tests (FMI), as the presence of gas in shales is in the form of free phase within pores and fractures and as gas sorbed onto organic matter Alipour et al. 2022). ...
Integrating petrophysical and geomechanical parameters is an efficient approach to evaluating shale gas reservoir potential. The high cost of corings and their limited number, coupled with time-intensive investigation, led researchers to use this alternative combination approach. In the Jiaoshiba area, from single-pilot well core data and log measurements, petrophysical and geomechanical parameters such as shale volume, total organic carbon, gas content, as well as pore pressure, stress components, and mineral brittleness were first estimated using established methods. In the second phase, based on logging curves, the reservoir electro-facies (EF) classification was performed using the unsupervised multi-resolution graph-based clustering method on a series of twenty wells, identifying five EF with different intrinsic characteristics. Unsupervised analyses were developed using the multilayer artificial neural network while incorporating the K-nearest neighbors and graphical classification algorithms. The results from the first and second phases indicate reservoir richness in organic matter, with the best reservoir exhibited by EF2 and EF3. In addition, effective stress components ( S V , S H , and S h ) evaluation shows a normal stress regime with hydraulic fracture systems perpendicular to the minimum horizontal stress at each measured depth of the reservoir ( S v > S H > S h ). This research workflow can efficiently evaluate shale reservoirs with a realistic approach for identifying favorable fracturing positions while reducing errors due to human interference.
... The burial history of shale reservoirs has a strong influence on porosity evolution (Zhang et al. 2013). The most significant processes in shale pore development are mechanical compaction and hydrocarbon generation (Loucks et al. 2012;Zhu et al. 2018b). During subsidence, most of the large pores in the fine-grained sediments are destroyed, dominated by mechanical compaction, particularly in the ductile minerals (Song et al. 2019a). ...
Pore volume, pore size and adsorption capacity are fundamental factors that control gas storage and production in shale reservoirs. We investigated the evolution of pore structure and adsorption capacity as functions of lithology and present-day burial depth. Nineteen shale samples from the Wufeng-Longmaxi Formation from shallow depths (present depths < 700 m) in the Sichuan Basin, China, were analyzed to elucidate the evolution of the pore system, applying N2/CO2 isotherm adsorption, high-resolution FE-SEM, FIB-SEM, and CH4 isotherm adsorption. Two comparison basin models with 700 m and 4000 m present-day burial depth were constructed to illustrate the impact of uplift on pore structure and methane adsorption. The results indicate that mineral structure, rock fabric, and TOC content are essential factors affecting the pore structure. Pores of various sizes are preserved through different mechanisms during the compaction. The development of micropores is mainly controlled by thermal maturity and TOC content. In contrast, mesopores and macropores are stress-sensitive and prone to be compacted during deposition. Methane adsorption capacity is highly related to the burial depth. During the process of reservoir uplift, the methane adsorption capacity of the reservoir gradually increases with decreasing burial depth, with a transition point occurring between 700 and 1300 m. Furthermore, as the reservoir continues to uplift to burial depths below 700 m, the methane adsorption capacity of the reservoir in normally pressured shale gas formations rapidly decreases. However, for overpressured shale gas reservoirs, there is no significant change in the methane adsorption capacity of the reservoir.
... Further, Tan [9] used three different regression algorithms and four different kernel functions to predict TOC content and Shi [8] used MLP-NN and Fuzzy logic to predict the TOC content data from wireline log to replace the Schmoke model. Extreme Learning Machine (ELM) were introduced in Shi [8] work for TOC estimation where ELM was quicker than ANN and Integrated hybrid neural network (IHNN) were introduced in Zhu [11] work in Jaioshiba zone to predict TOC. SVM with RBF outperformed ANN in Bolandi [12] work to predict TOC from wireline logs. ...
To evaluate the hydrocarbon generation potential of a rock strata, total organic carbon (TOC) is a significant factor. TOC estimation is considered as a challenge from well logs and direct measurement in laboratory from rock specimens is costly and time-consuming. In this paper, for the TOC prediction we proposed three artificial intelligence techniques including-Support vector machine (SVM), Deep Neural Network (DNN) and Genetic Algorithm-Deep Neural Network (GA-DNN) capable of fitting highly non-linear data. Specifically, 205 data points and seven well logs namely GR, DT, RHOB, SP, NPHI, LLD, and LLS were used from the Goldwyer Formation of the Canning Basin for training and testing the tree-based ensemble models to evaluate their efficiency and provide comparable results during the TOC estimation. From results it is validated that the accuracy of the three artificial intelligence techniques is at exemplary level for the TOC content estimation where the GA-DNN model for training and testing data sets (For testing R 2 94.39%, MAE 0.0447, MSE 0.0039) outperformed all the other AI techniques i.e., SVM (for testing R 2 90.63%, MAE 0.0583, MSE 0.0058) and DNN (for testing R 2 91.23%, MAE 0.0569, MSE 0.0053).
... In particular, convolutional neural networks (CNNs) (Le-Cun et al. 2015) and recurrent neural networks (RNNs) (Goltsev & Rachkovskij 2001;Tian & Noore 2004) have been successfully applied in many fields. CNNs have strong feature extraction abilities and are gradually being used to interpret logging information by identifying tight gas reservoirs (Zhu et al. 2020), estimating reservoir porosity (Feng 2020), and predicting total organic carbon content (Zhu et al. 2018) and lithology (Zhang et al. 2018b). The logging signal is a type of time-series signal; the logging information increases as the formation depth increases, and the dynamic changes between data are a rich source of information. ...
Well logging helps geologists find hidden oil, natural gas and other resources. However, well log data are systematically insufficient because they can only be obtained by drilling, which involves costly and time-consuming field trials. Additionally, missing or distorted well log data are common in old oilfields owing to shutdowns, poor borehole conditions, damaged instruments and so on. As a workaround, pseudo-data can be generated from actual field data. In this study, we propose a spatio-temporal neural network (STNN) algorithm, which is built by leveraging the combined strengths of a convolutional neural network (CNN) and a long short-term memory network (LSTM). The STNN exploits the ability of the CNN to effectively extract features related to pseudo-well log data and the ability of the LSTM to extract the key features from well log data along the depth direction. The STNN method allows full consideration of the well log data trend with depth, the correlation across different log series and the actual depth accumulation effect. The method proved successful in predicting acoustic sonic log data from gamma-ray, density, compensated neutron, formation resistivity and borehole diameter logs. Results show that the proposed method achieves higher prediction accuracy because it takes into account the spatio-temporal information of well logs.
... Furthermore, Tan et al. [8] used three different regression algorithms and four different kernel functions to predict TOC content and Shi et al. [7] used MLP-NN and fuzzy logic to predict the TOC from the well logs to replace the Schmoker's model. Extreme Learning Machine (ELM) was introduced by Shi et al. [7] for TOC estimation where ELM was found quicker than ANN and Zhu et al. [10] introduced Integrated Hybrid Neural Network (IHNN) in Jaioshiba zone. SVM with RBF outperformed ANN in Bolandi et al.'s work [11] to predict TOC from well logs. ...
To evaluate the hydrocarbon generation potential, Total Organic Carbon (TOC) of source/reservoir rocks is of vital importance. TOC estimation from well logs is challenging and in laboratory from rock specimens is costly as well as time-consuming. TOC prediction from Passey method is low whereas AI techniques such as Artificial Neu-ral Network (ANN), Support Vector Machine (SVM) get trapped in local optima, resulting in overfitting and are also considered ambiguous if the technique is not reasonable. In this paper, we proposed four efficient tree-based ensemble techniques that include Random Forest (RF), Extra Trees (ET), Gradient Boosting (GB), and eXtremely Gradient Boosting (XGB), capable of fitting highly non-linear data with minimum data pre-processing for TOC prediction. To evaluate the efficiency of these models, 205 data points and seven well logs from the Goldwyer Formation of the Canning Basin, Australia, were used for the training and testing purpose. Results validated that the accuracy of these tree-based ensemble techniques is at exemplary level for the TOC estimation, where the XGB model (for testing R 2 94.39%, MAE 0.0447, MSE 0.0039) outperformed the other techniques, i.e., RF (for testing R 2 90.59%, MAE 0.0549, MSE 0.0055), ET (for testing R 2 90.63%, MAE 0.0583, MSE 0.0058) and GB (for testing R 2 91.23%, MAE 0.0569, MSE 0.0053). These robust tree-based ensemble techniques have not only protected overfitting but also achieved better prediction results in dealing with the multidimensional data.
... ANN provided higher precision compared to ∆ log R for estimating TOC content. • An integrated hybrid neural network (IHNN) is developed in Zhu et al. [57] to work for TOC content estimation in Jiaoshiba area. From the result, it was observed that IHNN outperformed other AI models. ...
The most important element for the exploration and development of oil and
oil shale is total organic carbon (TOC). TOC estimation is considered a challenge for
geologists since laboratory methods are expensive and time-consuming. Therefore, due
to the complex and nonlinear relationship between well logs and TOC, researchers have
begun to use arti�cial intelligence (AI) techniques. Hence, the purpose of this research
is to explore new paradigms and methods for AI techniques. First, this article provides
a recent overview of selected AI technologies and their applications, including arti�cial
neural networks (ANNs), convolutional neural networks (CNNs), hybrid intelligent sys-
tems (HISs), and support vector machines (SVMs) as well as fuzzy logic (FL), particle
swarm optimization (PSO). Second, this article explores and discusses the bene�ts and
pitfalls of each type of AI technology. The study found that hybrid intelligence technology
was the most successful and independent AI model with the highest probability of infer-
ring properties of oil shale oil and gas �elds (such as TOC) from wireline logs. Finally,
some possible combinations are proposed that have not yet been investigated.
... Zhu et al. proposed an efficient and new integrated hybrid neural network method in order to estimate total carbon content (TOC) from conventional logging data. The model has a very high accuracy in order to estimate TOC [15]. Mohaghegh utilized neural networks (NNs) for prediction of parameters related to NMR log of one of the East Texas fields [16]. ...
This work highlights the application of Artificial Neural Networks optimized by Cuckoo optimization algorithm for predictions of NMR log parameters including porosity and permeability by using field log data. The NMR logging data have some highly vital privileges over conventional ones. The measured porosity is independent from bearer pore fluid and is effective porosity not total. Moreover, the permeability achieved by exact measurement and calculation considering clay content and pore fluid type. Therefore availability of the NMR data brings a great leverage in understanding the reservoir properties and also perfectly modelling the reservoir. Therefore, achieving NMR logging data by a model fed by a far inferior and less costly conventional logging data is a great privilege. The input parameters of model were neutron porosity (NPHI), sonic transit time (DT), bulk density (RHOB) and electrical resistivity (RT). The outputs of model were also permeability and porosity values. The structure developed model was build and trained by using train data. Graphical and statistical validation of results showed that the developed model is effective in prediction of field NMR log data. Outcomes show great possibility of using conventional logging data be used in order to reach the precious NMR logging data without any unnecessary costly tests for a reservoir. Moreover, the considerable accuracy of newly ANN-Cuckoo method also demonstrated. This study can be an illuminator in areas of reservoir engineering and modelling studies were presence of accurate data must be essential.
... Extreme Learning Machine (ELM) and ANN were used to predict TOC in Shi et al [6] work where ELM was quicker than ANN. Integrated hybrid neural network (IHNN) were used to predict TOC in Zhu et al. [17] work in Jaioshiba zone. In Bolandi et al. [18] work, ANN was outperformed by SVM with RBF to predict TOC content from well logs. ...
For shale oil and gas exploration total organic carbon (TOC) is the significant factors where TOC estimation considered as a challenges for geological engineers because direct laboratory coring analysis is costly and time consuming. Passey method and Artificial Intelligence (AI) technique have used on well logs extensively to determine TOC content. But, the prediction of Passey method is low and AI technique such as ANN, Support Vector Machine (SVM) trapped in local optima, overfitting and heavy computation work or error if the technique isn’t reasonable. In this paper, for the first time in TOC prediction we propose three feature selection-based algorithm which are Decision Tree (DT), Gradient Boosting Regressor (GBR) and Random Forest (RF) respectively. This feature selection-based algorithm select the best attributes among the input parameters for TOC content prediction. Then those best attributes works as an input for AI models for training and testing the AI models which illustrates that making a correlation between well logs and TOC content for the prediction. Specifically, 2069 core shale sample and well logging sample data of the Texas University Lands of Kansas Geologic Society were divided into 1448 training sample and 621 validating sample to evaluate the proposed AI models. This proposed AI model and feature selection-based algorithm jointly allows TOC content to be accurately and continuously predicted based on conventional well logs.
... Hydrocarbon Generation. TOC and S1+S2 were usually applied to evaluate hydrocarbon generation potential [40,41]. The bulk rock-normalized TOC and S1+S2 for bulk rock and residue after acidolysis indicate that acid treatment leads to hydrocarbon generable organic fraction loss, suggesting that conventional acid-treatment TOC measurements may underestimate the hydrocarbon generation potential. ...
It has been acknowledged that carbonate was identified as the source rocks of a series of oil-gas fields worldwide. For evaluating the carbonate source rocks, total organic carbon (TOC) contents act as an important indicator. However, the acid solution, which has been generated during conventional TOC measurements, contain organic matters. Hence, the released organic matters in acid solution during carbonate decomposition may lead to underestimate the hydorcarbon generation potential. In this study, rock-eval pyrolysis technique was applied to bulk rock and residue after acid treatment. Meanwhile, the organic matters in acid solution were measured by Gas Chromatography-Mass Spectrometer (GC-MS) to investigate the geochemical characteristics. In addition, the hydrocarbon generation and alteration of TOC contents of released organic matters by acid treatment were studied by hydrous pyrolysis experiments. The results show that the acid solution contains organic compounds, including n-alkanes, saturated fatty acids and fatty acid methyl esters. Meanwhile, total organic carbon (TOC) contents and hydrocarbon generation potential (S1+S2) significantly decrease for bulk rocks after acid treatment in low maturity samples. Moreover, organic CO2 (S3) decreased after treating of acid, revealing that acidolysis process can affect and release organic matters containing oxygen-bearing functional groups. The S1, S2, S3, and TOC loss are positive correlation with the proportion of rock loss during acidolysis, indicating that the organic matters in acid solution are associated with carbonate minerals. The organic fractions may exist as adsorption state on the surface of carbonate minerals and (or) exist as organic acid salts. Moreover, the thermal simulation experiments reveal that the organic matter in acid solution, which is not recovered by the conventional measurement approach, could contribute to hydrocarbon generation.
