A Real-time Lithological Identification Method based on SMOTE-Tomek and ICSA Optimization

In petroleum engineering,real-time lithology identification is very important for reservoir evaluation,drilling decisions and petroleum geological exploration.A lithology identification method while drilling based on machine learning and mud logging data is studied in this paper.This method can effectively utilize downhole parameters collected in real-time during drilling,to identify lithology in real-time and provide a reference for optimization of drilling parameters.Given the imbalance of lithology samples,the synthetic minority over-sampling technique(SMOTE)and Tomek link were used to balance the sample number of five lithologies.Meanwhile,this paper introduces Tent map,random opposition-based learning and dynamic perceived probability to the original crow search algorithm(CSA),and establishes an improved crow search algorithm(ICSA).In this paper,ICSA is used to optimize the hyperparameter combination of random forest(RF),extremely random trees(ET),extreme gradient boosting(XGB),and light gradient boosting machine(LGBM)models.In addition,this study combines the recognition advantages of the four models.The accuracy of lithology identification by the weighted average probability model reaches 0.877.The study of this paper realizes high-precision real-time lithology identification method,which can provide lithology reference for the drilling process.

A real-time intelligent lithology identification method based on a dynamic felling strategy weighted random forest algorithm

Real-time intelligent lithology identification while drilling is vital to realizing downhole closed-loop drilling. The complex and changeable geological environment in the drilling makes lithology identification face many challenges. This paper studies the problems of difficult feature information extraction,low precision of thin-layer identification and limited applicability of the model in intelligent lithologic identification. The author tries to improve the comprehensive performance of the lithology identification model from three aspects: data feature extraction, class balance, and model design. A new real-time intelligent lithology identification model of dynamic felling strategy weighted random forest algorithm(DFW-RF) is proposed. According to the feature selection results, gamma ray and 2 MHz phase resistivity are the logging while drilling(LWD) parameters that significantly influence lithology identification. The comprehensive performance of the DFW-RF lithology identification model has been verified in the application of 3 wells in different areas. By comparing the prediction results of five typical lithology identification algorithms, the DFW-RF model has a higher lithology identification accuracy rate and F1 score. This model improves the identification accuracy of thin-layer lithology and is effective and feasible in different geological environments. The DFW-RF model plays a truly efficient role in the realtime intelligent identification of lithologic information in closed-loop drilling and has greater applicability, which is worthy of being widely used in logging interpretation.

Stratigraphic identification using real-time drilling data

Identification of stratigraphic interfaces and lithology is a key aspect in geological and geotechnical investigations.In this study,a monitoring while-drilling system was developed,along with a corresponding data pre-processing method.The method can handle invalid drilling data generated during manual operations.The correlation between various drilling parameters was analyzed,and a database of stratigraphic interfaces and key lithology identification based on the monitoring parameters was established.The average drilling speed was found to be the most suitable parameter for stratigraphic and lithology identification,and when the average drilling speed varied over a wide range,it corresponded to a stratigraphic interface.The average drilling speeds in sandy mudstone and sandstone strata were in the ranges of 0.1e0.2 m/min and 0.2e0.29 m/min,respectively.The results obtained using the present method were consistent with geotechnical survey results.The proposed method can be used for realtime lithology identification and represents a novel approach for intelligent geotechnical surveying.

Quantitative identification and prediction of mixed lithology, Bohai Sea, China

The Paleogene Shahejie Formation in the KL16 oilfield, Bohai bay, is characterized by a thinly interbedded mixed sedimentary system, with complex sedimentary facies, lithologic types and distributions. It is hard for conventional logging methods to identify the lithology therein. In order to solve the difficulty in lithologic identification of mixed sedimentary system, analyses based on graph data base using elemental capture energy spectrum log have been proposed. Due to the different composition for the various minerals, we innovatively established the molar numbers of silicon, calcium, magnesium, and aluminum as characteristic parameters for sandstone, limestone, dolomite, and mudstone, and a graph clustering analysis method was applied to identify lithology. Considering the seismic waveforms corresponding to lithologic impedance of reservoir, three seismic phases were identified by neural network clustering analysis of seismic waveform, and the seismic attributes with high sensitivity to reservoir thickness were then selected to realize the fine description of the mixed carbonate-siliciclastic reservoir. Drilling results confirmed that the sedimentary facies were accurately identified, with reservoir prediction accuracy reaching up to 80%. Under the guidance of reservoir research, the oil-in-place discovered in the oilfield were estimated to be more than 5 million tonnes. This technology provides reference for the exploration and development of oilfields of mixed sedimentary system.

Log identification method in complex sand reservoirs

In order to obtain effective parameters for complex sand reservoirs,a log evaluation method for relevant reservoir parameters is established based on an analysis in the gas-bearing sandstone with high porosity and low permeability,low porosity and permeability and on various characteristics of log responses to reservoir lithologies and physical properties in the Neopleozoic sand reservoir of the Ordos basin.This log evaluation method covers the Cook method that is used to evaluate the porosity and oiliness in high porosity and low permeability reservoirs and another method in which the mineral content,derived from geochemical logs,is used to identify formation lithologies.Some areas have high calcium and low silt content,not uniformly distributed,the results of which show up in the complex formation lithologies and conventional log responses with great deviation.The reliability of the method is verified by comparison with conventional log data and core analyses.The calculation results coincide with the core analytical data and gas tests,which indicate that this log evaluation method is available,provides novel ideas for study of similar complex reservoir lithologies and has some reference value.

Multi-scale data joint inversion of minerals and porosity in altered igneous reservoirs—A case study in the South China Sea

There are abundant igneous gas reservoirs in the South China Sea with significant value of research,and lithology classification,mineral analysis and porosity inversion are important links in reservoir evaluation.However,affected by the diverse lithology,complicated mineral and widespread alteration,conventional logging lithology classification and mineral inversion become considerably difficult.At the same time,owing to the limitation of the wireline log response equation,the quantity and accuracy of minerals can hardly meet the exploration requirements of igneous formations.To overcome those issues,this study takes the South China Sea as an example,and combines multi-scale data such as micro rock slices,petrophysical experiments,wireline log and element cutting log to establish a set of joint inversion methods for minerals and porosity of altered igneous rocks.Specifically,we define the lithology and mineral characteristics through core slices and mineral data,and establish an igneous multi-mineral volumetric model.Then we determine element cutting log correction method based on core element data,and combine wireline log and corrected element cutting log to perform the lithology classification and joint inversion of minerals and porosity.However,it is always difficult to determine the elemental eigenvalues of different minerals in inversion.This paper uses multiple linear regression methods to solve this problem.Finally,an integrated inversion technique for altered igneous formations was developed.The results show that the corrected element cutting log are in good agreement with the core element data,and the mineral and porosity results obtained from the joint inversion based on the wireline log and corrected element cutting log are also in good agreement with the core data from X-ray diffraction.The results demonstrate that the inversion technique is applicable and this study provides a new direction for the mineral inversion research of altered igneous formations.

Deep learning of rock microscopic images for intelligent lithology identification: Neural network comparison and selection

An intelligent lithology identification method is proposed based on deep learning of the rock microscopic images.Based on the characteristics of rock images in the dataset,we used Xception,MobileNet_v2,Inception_ResNet_v2,Inception_v3,Densenet121,ResNet101_v2,and ResNet-101 to develop microscopic image classification models,and then the network structures of seven different convolutional neural networks(CNNs)were compared.It shows that the multi-layer representation of rock features can be represented through convolution structures,thus better feature robustness can be achieved.For the loss function,cross-entropy is used to back propagate the weight parameters layer by layer,and the accuracy of the network is improved by frequent iterative training.We expanded a self-built dataset by using transfer learning and data augmentation.Next,accuracy(acc)and frames per second(fps)were used as the evaluation indexes to assess the accuracy and speed of model identification.The results show that the Xception-based model has the optimum performance,with an accuracy of 97.66%in the training dataset and 98.65%in the testing dataset.Furthermore,the fps of the model is 50.76,and the model is feasible to deploy under different hardware conditions and meets the requirements of rapid lithology identification.This proposed method is proved to be robust and versatile in generalization performance,and it is suitable for both geologists and engineers to identify lithology quickly.

Discovery and identification of clay-sized tuffaceous rocks in Chang 7 Member of Upper Triassic Yanchang Formation in Ordos Basin,China

This study for the first time demonstrates that some of the so-called clay-sized mudstones observed by the naked eye,such as clay-sized black mudstones and clay-sized oil shales,which are rich in black organic matter(including oil and asphaltene),in the Chang 73 Submember of the Upper Triassic Yanchang Formation in the Ordos Basin of China are actually clay-sized tuffaceous rocks(including tuff,sedimentary tuff and tuffaceous sedimentary rock)with high hydrocarbon generation capacities.Thus,these rocks can be defined as clay-sized tuffaceous source rocks.Identification of this lithology has important theoretical and practical significance for the exploration and development of shale oil in the Chang 7 Member.Through the macroscopic observation of drill cores and outcrop profiles,microscopic observation of electron probe thin sections and whole-rock inorganic geochemical analysis(including major,trace and rare earth elements),this work demonstrates that the organic matter-rich clay-sized tuffaceous rocks,especially clay-sized tuffs,have the following characteristics.First,the clay-sized tuffaceous rocks with little black organic matter are mainly greyish white,yellowish brown and purplish grey,and mixed colors occur in areas with strong bentonite lithification.Second,the clay-sized tuffaceous rocks have experienced strong devitrification and recrystallization,forming abundant flaky aluminosilicate minerals with directional arrangement.In thin sections under a polarizing microscope,the interference colors generally show regular alternation between the lowest interference color of first-order yellow and the highest interference color of second-order blue-green.Third,the rock samples plot in the igneous rock field in the TiO2-SiO2 cross-plot and exhibit similar trace element and rare earth element patterns on spider diagrams,indicating that the samples are derived from the same source.The results prove that clay-sized tuffaceous rocks may be widespread in the Chang 73 Submember of the Upper Triassic Yanchang Formation in the Ordos Basin,China.

A Study on the Classification and Well-Logging Identification of Eclogite in the Main Hole of Chinese Continental Scientific Drilling Project

Eclogite, one of the important lithologies in the main hole of the Chinese Continental Scientific Drilling （CCSD） Project, exists above the depth of 3 245 m and has distinctive responses of gamma-ray, compensating density and neutron well-logging, and so on. In this study, according to the diversities of minerals and chemical components and well-logging responses, edogites are classified from three aspects of origin, content of oxygen, and sub-mineral. We studied the logging identification method for eclogite sub-classes based on multi-element statistics and reconstructed 11 kinds of eclogite. As a result, eclogites can be divided into 6 types using well logs. In the light of this recognition, the eclogite in the main hole is divided into 20 sections, and the distribution characters of all sub-classes of eclogite are analyzed, which will provide important data for geological research of CCSD.

Adaptive Identification of Logging Lithology Based on VPSO-ENN Hybrid Algorithm

Particle swarm optimization （PSO） was modified by variation method of particle velocity, and a variation PSO （VPSO） algorithm was proposed to overcome the shortcomings of PSO, such as premature convergence and local optimization. The VPSO algorithm is combined with Elman neural network （ENN） to form a VPSO-ENN hybrid algorithm. Compared with the hybrid algorithm of genetic algorithm （GA） and BP neural network （GA-BP）, VPSO-ENN has less adjustable parameters, faster convergence speed and higher identification precision in the numerical experiment. A system for identifying logging parameters was established based on VPSO-ENN. The results of an engineering case indicate that the intelligent identification system is effective in the lithology identification.

Quantitative evaluation methods for waterflooded layers of conglomerate reservoir based on well logging data

The rapid changing near source, multi-stream depositional environment of conglomerate reservoirs leads to severe heterogeneity, complex lithology and physical properties, and large changes of oil layer resistivity. Quantitative evaluation of water-flooded layers has become an important but difficult focus for secondary development of oilfields. In this paper, based on the analysis of current problems in quantitative evaluation of water-flooded layers, the Kexia Group conglomerate reservoir of the Sixth District in the Karamay Oilfield was studied. Eight types of conglomerate reservoir lithology were identified effectively by a data mining method combined with the data from sealed coring wells, and then a multi-parameter model for quantitative evaluation of the water-flooded layers of the main oil-bearing lithology was developed. Water production rate, oil saturation and oil productivity index were selected as the characteristic parameters for quantitative evaluation of water-flooded layers of conglomerate reservoirs. Finally, quantitative evaluation criteria and identification rules for water-flooded layers of main oil-bearing lithology formed by integration of the three characteristic parameters of water-flooded layer and undisturbed formation resistivity. This method has been used in evaluation of the water-flooded layers of a conglomerate reservoir in the Karamay Oilfield and achieved good results, improving the interpretation accuracy and compliance rate. It will provide technical support for avoiding perforation of high water-bearing layers and for adjustment of developmental programs.

Study of S-wave ray elastic impedance for identifying lithology and fluid

In this paper, we derive an approximation of the SS-wave reflection coefficient and the expression of S-wave ray elastic impedance （SREI） in terms of the ray parameter. The SREI can be expressed by the S-wave incidence angle or P-wave reflection angle, referred to as SREIS and SREIP, respectively. Our study using elastic models derived from real log measurements shows that SREIP has better capability for lithology and fluid discrimination than SREIS and conventional S-wave elastic impedance （SEI）. We evaluate the SREIP feasibility using 25 groups of samples from Castagna and Smith （1994）. Each sample group is constructed by using shale, brine-sand, and gas-sand. Theoretical evaluation also indicates that SRE1P at large incident angles is more sensitive to fluid than conventional fluid indicators. Real seismic data application also shows that SRE1P at large angles calculated using P-wave and S-wave impedance can efficiently characterize tight gas-sand.

Laser-induced breakdown spectroscopy applied to the characterization of rock by support vector machine combined with principal component analysis

Laser-induced breakdown spectroscopy（LIBS） is a versatile tool for both qualitative and quantitative analysis.In this paper,LIBS combined with principal component analysis（PCA） and support vector machine（SVM） is applied to rock analysis.Fourteen emission lines including Fe,Mg,Ca,Al,Si,and Ti are selected as analysis lines.A good accuracy（91.38% for the real rock） is achieved by using SVM to analyze the spectroscopic peak area data which are processed by PCA.It can not only reduce the noise and dimensionality which contributes to improving the efficiency of the program,but also solve the problem of linear inseparability by combining PCA and SVM.By this method,the ability of LIBS to classify rock is validated.

Research on the Application of the Light-Energy Spectrum Fusion Technique to Land and Resources Survey

This paper introduces how to use remote sensing images including Landsat (MSS and TM) and airborne radioactivity images to identify the type of rocks in the areas covered by vegetation. The relationship between light spectrum (Landsat MSS and TM) and energy spectrum (U, Th and K) is discussed on the basis of correlation analysis, and it is proven that there are correlations between the Landsat MSS or TM data and the U, Th and K data. By using the fusion technique, new images were generated, which contain both the light spectrum and the energy spectrum information.

Hard-rock tunnel lithology identification using multiscale dilated convolutional attention network based on tunnel face images

For real-time classification of rock-masses in hard-rock tunnels,quick determination of the rock lithology on the tunnel face during construction is essential.Motivated by current breakthroughs in artificial intelligence technology in machine vision,a new automatic detection approach for classifying tunnel lithology based on tunnel face images was developed.The method benefits from residual learning for training a deep convolutional neural network(DCNN),and a multi-scale dilated convolutional attention block is proposed.The block with different dilation rates can provide various receptive fields,and thus it can extract multi-scale features.Moreover,the attention mechanism is utilized to select the salient features adaptively and further improve the performance of the model.In this study,an initial image data set made up of photographs of tunnel faces consisting of basalt,granite,siltstone,and tuff was first collected.After classifying and enhancing the training,validation,and testing data sets,a new image data set was generated.A comparison of the experimental findings demonstrated that the suggested approach outperforms previous classifiers in terms of various indicators,including accuracy,precision,recall,F1-score,and computing time.Finally,a visualization analysis was performed to explain the process of the network in the classification of tunnel lithology through feature extraction.Overall,this study demonstrates the potential of using artificial intelligence methods for in situ rock lithology classification utilizing geological images of the tunnel face.

Applicability of deep neural networks for lithofacies classification from conventional well logs: An integrated approach

Parametric understanding for specifying formation characteristics can be perceived through conven-tional approaches.Significantly,attributes of reservoir lithology are practiced for hydrocarbon explora-tion.Well logging is conventional approach which is applicable to predict lithology efficiently as compared to geophysical modeling and petrophysical analysis due to cost effectiveness and suitable interpretation time.However,manual interpretation of lithology identification through well logging data requires domain expertise with an extended length of time for measurement.Therefore,in this study,Deep Neural Network(DNN)has been deployed to automate the lithology identification process from well logging data which would provide support by increasing time-effective for monitoring lithology.DNN model has been developed for predicting formation lithology leading to the optimization of the model through the thorough evaluation of the best parameters and hyperparameters including the number of neurons,number of layers,optimizer,learning rate,dropout values,and activation functions.Accuracy of the model is examined by utilizing different evaluation metrics through the division of the dataset into the subdomains of training,validation and testing.Additionally,an attempt is contributed to remove interception for formation lithology prediction while addressing the imbalanced nature of the associated dataset as well in the training process using class weight.It is assessed that accuracy is not a true and only reliable metric to evaluate the lithology classification model.The model with class weight recognizes all the classes but has low accuracy as well as a low F1-score while LSTM based model has high accuracyas well as a high F1-score.