Integrating well log interpretations for lithofacies classification and permeability modeling through advanced machine learning algorithms

Abstract

In this paper, an integrated procedure was adopted to obtain accurate lithofacies classification to be incorporated with well log interpretations for a precise core permeability modeling. Probabilistic neural networks (PNNs) were employed to model lithofacies sequences as a function of well logging data in order to predict discrete lithofacies distribution at missing intervals. Then, the generalized boosted regression model (GBM) was used as to build a nonlinear relationship between core permeability, well logging data, and lithofacies. The well log interpretations that were considered for lithofacies classification and permeability modeling are neutron porosity, shale volume, and water saturation as a function of depth; however, the measured discrete lithofacies types are sand, shaly sand, and shale. Accurate lithofacies classification was achieved by the PNN as the total percent correct of the predicted discrete lithofacies was 95.81%. In GBM results, root-mean-square prediction error and adjusted R-square have incredible positive values, as there was an excellent matching between the measured and predicted core permeability. Additionally, the GBM model led to overcome the multicollinearity that was available between one pair of the predictors. The efficiency of boosted regression was demonstrated by the prediction matching of core permeability in comparison with the conventional multiple linear regression (MLR). GBM led to much more accurate permeability prediction than the MLR.

Interpreting the Subsurface Lithofacies at High Lithological Resolution by Integrating Information From Well‐Log Data and Rock‐Core Digital Images

“…Compared to direct rock-core analysis, the estimation of lithofacies based on well-log data can have significant uncertainty but is a cost-effective and viable method even for noncored or partially cored sections. Because of the cost benefits and information continuity, well-log data analysis methods have been widely applied in recent decades (Al-Mudhafar, 2017;Al-Mudhafar & Bondarenko, 2015;Lee & Datta-Gupta, 1999;Nashawi & Malallah, 2009;Tang et al, 2004). In these well-log data analysis methods, lithofacies that are separable from others in terms of petrophysical properties are identified and interpreted by using statistical characteristics of a set of well-log responses to produce electrofacies (Serra & Abbott, 1982).…”

“…To reduce the uncertainty of lithofacies interpretation solely based on well-log data, both direct and indirect data can be integrated (Akkurt et al, 2018;Al-Mudhafar, 2017;Coll et al, 1999;Faraji et al, 2017;Jeong et al, 2014;John et al, 2008;Kemp, 1982;Lindberg & Omre, 2014;Verma et al, 2014;Wedge et al, 2018;Worthington, 1994;Yan, 2002). Many of these studies were based on electrical borehole images (Donselaar & Schmidt, 2005;Folkestad et al, 2012;Linek et al, 2007;Luthi, 1994;Tilke et al, 2006;Yuan et al, 2017).…”

Integrated Study of Lithofacies Identification—A Case Study in X Field, Sabah, Malaysia

“…Various classification algorithms have been adopted for facies classification, such as generalized boosted regression model (GBM) [7], kernel support vector machines [8], partitioning algorithms [9], and neural networks [10,11]. These classification algorithms predict the discrete and continuous probability distribution of facies.…”

“…These classification algorithms predict the discrete and continuous probability distribution of facies. The interpretation of different log data such as neutron porosity, shale volume, and water saturation was considered for lithofacies classification in these studies [7,8].…”

“…Depending on their geologic application, attributes are classified and categorized into three groups: (1) structural mapping and geomorphological identification, (2) resolution of thin beds and improving interpretability, and (3) porefill and lithology discrimination [12]. Various classification algorithms have been adopted for facies classification, such as generalized boosted regression model (GBM) [7], kernel support vector machines [8], partitioning algorithms [9], and neural networks [10,11]. These classification algorithms predict the discrete and continuous probability distribution of facies.…”