New Technique to Determine the Total Organic Carbon Based on Well Logs Using Artificial Neural Network (White Box)

Abstract

Total organic carbon (TOC) is the amount of carbon present in an organic compound and is often used as an essential factor for unconventional shale resources evaluation. The previous models for TOC determination were either based on density log data only and considered the presence of organic matter is proportional to the bulk density, or based on resistivity log, sonic or density logs as well as the formation level of maturity (LOM), where these models assumed a linear relation between resistivity and porosity logs. The average absolute deviation (ADD) of the previous model was not less than 1.20wt% of TOC with a coefficient of determination (R2) of less than 0.85. The objective of this research is to develop new empirical correlation to determine the TOC based on well logs using artificial neural network for Barnett shale formation. Core TOC data (442 data point) and well logs (resistivity, gamma ray, sonic transit time, and bulk density) were used to develop the ANN model. For the first time, the ANN model will change to a white box by extracting the weights and biases of the model to form the empirical equation. The results obtained showed that TOC is strong function of bulk density, and moderate function of gamma ray, compressional sonic time, and week function of deep resistivity. The developed ANN model is able to predict the TOC based on conventional log data with high accuracy (the ADD is 0.91wt% of TOC and R2 between estimated and actual TOC is 0.93). The developed empirical equation for TOC determination from the ANN model outperformed the previous available models, which had an ADD of 1.20 wt% or more and R2 of less than 0.85. The developed TOC model and equation can be applied using simple computer without the need for a specific software. The novelty of this new research is the simplicity and high accuracy of the developed model for estimating the total organic carbon based on conventional log data. The developed empirical equation will help the geologists and reservoir engineers to predict the TOC without the need for hard lab work or complicated softwares.

Application of Machine Learning in Evaluation of the Static Young’s Modulus for Sandstone Formations

“…Machine learning techniques are used in several scientific and engineering fields since the early 1990s to solve complicated non-linear problems. Petroleum engineers and petroleum geologists use different machine learning techniques to solve problems related to petroleum industry, such as the characterization of the heterogeneous hydrocarbon reservoirs [33,34], evaluation of the reserve of unconventional reservoirs [35][36][37][38], estimation of the rock mechanical parameters, such as the static Poisson's ratio in carbonate reservoirs [39] and the static Young's modulus for sandstone reservoirs [24,40], evaluation of the integrity of wellbore casing [41,42], optimization of drilling hydraulics [43], evaluation of pore pressure and fracture pressure [44,45], hydrocarbon recovery factor estimation [46,47], determination of the alteration in the drilling fluids rheology in real-time [48,49], optimization of rate of penetration [50,51], prediction of the formation tops [52], and others.…”

Prediction of the Rate of Penetration while Drilling Horizontal Carbonate Reservoirs Using the Self-Adaptive Artificial Neural Networks Technique

“…AI techniques are used extensively in applications related to different engineering and scientific research areas [15][16][17][18][19][20], including in the petroleum industry where they can solve complicated problems such as prediction of drill bit wear from drilling parameters [21], real-time predictions of alterations in drilling fluid rheology [22,23], lithology identification [24], prediction of total organic carbon for the evaluation of unconventional resources [25][26][27][28][29], estimation of the oil recovery factor [30,31], estimation of pore and fracture pressures [32,33], evaluation of the static Young's modulus [34][35][36], estimation of the reservoir porosity [37], evaluation of the bubble point pressure [38], and the prediction of formation tops [39].…”

Evaluation of the Total Organic Carbon (TOC) Using Different Artificial Intelligence Techniques

“…Applying any of the previously discussed correlations to evaluate TOC in formations different than the one developed leads to inaccurate predictions. Recently, Mahmoud et al [18,19] suggested an artificial neural network (ANN)-based correlation for TOC estimation in Barnett formation using conventional well logs. Later on, Elkatatny [20] applied the self-adaptive differential evolution algorithm to optimize Mahmoud et al's [18,19] ANN model and he was able to improve the model predictability.…”

“…Since the early 1990s, AI techniques had been extensively applied in many scientific and engineering fields, including in the petroleum industry. Nowadays, AI has been used by petroleum engineers and geologists to solve problems related to unconventional hydrocarbon resources evaluation [18][19][20], reservoir characterization [21,22], bubble point pressure evaluation [23], prediction of real-time change in the rheological parameters of the drilling fluids [24,25], optimization of rate of penetration [26], estimation of rock mechanical parameters [27,28], prediction of pore pressure and fracture pressure [29,30], evaluation of the wellbore casing integrity [31,32], hydrocarbon recovery factor estimation [33,34] optimization of the drilling hydraulics [35], and others. AI techniques have also been applied successfully in other fields like social media [36,37].…”

“…The AI model's validation was completed using unseen data collected from the Devonian shale formation. The total number of core derived TOC data collected from Devonian shale are 22 data points, out of these data, only 20,19,19, and 15 were found to fit within the range of the training data that is used to develop TSK-FIS, M-FIS, FNN, and SVM models, respectively. The range for the training data are summarized in Table 1.…”