Method and practice of deep favorable shale reservoirs prediction based on machine learning

A set of methods for predicting the favorable reservoir of deep shale gas based on machine learning is proposed through research of parameter correlation feature analysis principle, intelligent prediction method based on convolution neural network(CNN), and integrated fusion characterization method based on kernel principal component analysis(KPCA) nonlinear dimension reduction principle.(1) High-dimensional correlation characteristics of core and logging data are analyzed based on the Pearson correlation coefficient.(2) The nonlinear dimension reduction method of KPCA is used to characterize complex high-dimensional data to efficiently and accurately understand the core and logging response laws to favorable reservoirs.(3) CNN and logging data are used to train and verify the model similar to the underground reservoir.(4) CNN and seismic data are used to intelligently predict favorable reservoir parameters such as organic carbon content, gas content, brittleness and in-situ stress to effectively solve the problem of nonlinear and complex feature extraction in reservoir prediction.(5) KPCA is used to eliminate complex redundant information, mine big data characteristics of favorable reservoirs, and integrate and characterize various parameters to comprehensively evaluate reservoirs. This method has been used to predict the spatial distribution of favorable shale reservoirs in the Ordovician Wufeng Formation to the Silurian Longmaxi Formation of the Weirong shale gas field in the Sichuan Basin, SW China. The predicted results are highly consistent with the actual core, logging and productivity data, proving that this method can provide effective support for the exploration and development of deep shale gas.

Forming Condition and Geology Prediction Techniques of Deep Clastic Reservoirs

1 Introduction As new exploration domain for oil and gas,reservoirs with low porosity and low permeability have become a hotspot in recent years(Li Daopin,1997).With the improvement of technology,low porosity and low

Application of Attributes Fusion Technology in Prediction of Deep Reservoirs in Paleogene of Bohai Sea

1 Introduction The Paleogene strata(with a depth of more than 2500m)in the Bohai sea is complex(Xu Changgui,2006),the reservoir buried deeply,the reservoir prediction is difficult(LAI Weicheng,XU Changgui,2012),and more

Prediction of Sedimentary Microfacies Distribution by Coupling Stochastic Modeling Method in Oil and Gas Energy Resource Exploitation

In view of the problem that a single modeling method cannot predict the distribution of microfacies, a new idea of coupling modeling method to comprehensively predict the distribution of sedimentary microfacies was proposed, breaking the tradition that different sedimentary microfacies used the same modeling method in the past. Because different sedimentary microfacies have different distribution characteristics and geometric shapes, it is more accurate to select different simulation methods for prediction. In this paper, the coupling modeling method was to establish the distribution of sedimentary microfacies with simple geometry through the point indicating process simulation, and then predict the microfacies with complex spatial distribution through the sequential indicator simulation method. Taking the DC block of Bohai basin as an example, a high-precision reservoir sedimentary microfacies model was established by the above coupling modeling method, and the model verification results showed that the sedimentary microfacies model had a high consistency with the underground. The coupling microfacies modeling method had higher accuracy and reliability than the traditional modeling method, which provided a new idea for the prediction of sedimentary microfacies.

Pre-stack inversion for caved carbonate reservoir prediction:A case study from Tarim Basin,China

The major storage space types in the carbonate reservoir in the Ordovician in the TZ45 area are secondary dissolution caves.For the prediction of caved carbonate reservoir,post-stack methods are commonly used in the oilfield at present since pre-stack inversion is always limited by poor seismic data quality and insufficient logging data.In this paper,based on amplitude preserved seismic data processing and rock-physics analysis,pre-stack inversion is employed to predict the caved carbonate reservoir in TZ45 area by seriously controlling the quality of inversion procedures.These procedures mainly include angle-gather conversion,partial stack,wavelet estimation,low-frequency model building and inversion residual analysis.The amplitude-preserved data processing method can achieve high quality data based on the principle that they are very consistent with the synthetics.Besides,the foundation of pre-stack inversion and reservoir prediction criterion can be established by the connection between reservoir property and seismic reflection through rock-physics analysis.Finally,the inversion result is consistent with drilling wells in most cases.It is concluded that integrated with amplitude-preserved processing and rock-physics,pre-stack inversion can be effectively applied in the caved carbonate reservoir prediction.

Machine learning seismic reservoir prediction method based on virtual sample generation

Seismic reservoir prediction plays an important role in oil exploration and development.With the progress of artificial intelligence,many achievements have been made in machine learning seismic reservoir prediction.However,due to the factors such as economic cost,exploration maturity,and technical limitations,it is often difficult to obtain a large number of training samples for machine learning.In this case,the prediction accuracy cannot meet the requirements.To overcome this shortcoming,we develop a new machine learning reservoir prediction method based on virtual sample generation.In this method,the virtual samples,which are generated in a high-dimensional hypersphere space,are more consistent with the original data characteristics.Furthermore,at the stage of model building after virtual sample generation,virtual samples screening and model iterative optimization are used to eliminate noise samples and ensure the rationality of virtual samples.The proposed method has been applied to standard function data and real seismic data.The results show that this method can improve the prediction accuracy of machine learning significantly.

Reservoir prediction using multi-wave seismic attributes

The main problems in seismic attribute technology are the redundancy of data and the uncertainty of attributes, and these problems become much more serious in multi-wave seismic exploration. Data redundancy will increase the burden on interpreters, occupy large computer memory, take much more computing time, conceal the effective information, and especially cause the ＂curse of dimension＂. Uncertainty of attributes will reduce the accuracy of rebuilding the relationship between attributes and geological significance. In order to solve these problems, we study methods of principal component analysis （PCA）, independent component analysis （ICA） for attribute optimization and support vector machine （SVM） for reservoir prediction. We propose a flow chart of multi-wave seismic attribute process and further apply it to multi-wave seismic reservoir prediction. The processing results of real seismic data demonstrate that reservoir prediction based on combination of PP- and PS-wave attributes, compared with that based on traditional PP-wave attributes, can improve the prediction accuracy.

Application of Pore Evolution and Fracture Development Coupled Models in the Prediction of Reservoir "Sweet Spots" in Tight Sandstones

The Chang-63 reservoir in the Huaqing area has widely developed tight sandstone ＂thick sand layers, but not reservoirs characterized by rich in oil＂, and it is thus necessary to further study its oil and gas enrichment law. This study builds porosity and fracture development and evolution models in different deposition environments, through core observation, casting thin section, SEM, porosity and permeability analysis, burial history analysis, and ＂four-property-relationships＂ analysis.

Sedimentary Characteristics and Reservoir Prediction of Paleogene in the East Part of Kuqa Foreland Basin

Most of the Mesozoic and Cenozoic large-scale hydrocarbon-bearing basins in western China were formed in a similar foreland setting. Hydrocarbon exploration of the Kuqa foreland basin requires research into the sedimentary characteristics and filling evolution of the depositional sequences and their response to the basin process. Based on an analysis of outcrops, well logs and high resolution seismic data, the sedimentary system types and distribution characteristics of the Paleogene in the east part of Kuqa foreland basin were systematically studied. The results show that： （ 1 ） Three types of sedimentary systems are developed in the area： an oxidative salty wide shallow lacustrine system, a fan delta system and an evaporitic bordersea system. （2） The configuration and evolution of the depositional systems of the Paleogene in the Kuqa foreland basin were predominantly determined by foreland tectonism. Vertically, the Paleogene sedimentary sequence can be divided into three parts： the lower, middle and upper depositional system tracts. The lower and upper tracts commonly consist of progradational or aggradational sequences, while the middle part is usually comprised of a set of aggradational to transgressive third-order sequences. Laterally, the sedimentary systems in the east part of the Kuqa foreland basin spread from east to west as a whole, and the sedimentary facies obviously vary from south to north. The sand bodies of the delta front facies are excellent gas reservoirs, characterized by rather thick, extensive and continuous distribution, high porosity and permeability, and just a few barrier beds.

Integrated Reservoir Prediction and Oil-Gas Evaluation in the Maoshan Area

The Maoshan area is an area with well-developed igneous rocks and complex structures. The thickness of the reservoirs is generally small. The study of the reservoirs is based on seismic data, logging data and geological data. Using techniques and software such as Voxelgeo, BCI, RM, DFM and AP, the authors have made a comprehensive analysis of the lateral variation of reservoir parameters in the Upper Shazu bed of the third member of the Palaeogene Funing Formation, and compiled the thickness map of the Shazu bed. Also, with the data from ANN, BCI and the abstracting method for seismic characteristic parameters in combination with the structural factors, the authors have tried the multi-parameter and multi-method prediction of petroleum, delineated the potential oil and gas areas and proposed two well sites. The prediction of oil and gas for Well JB2 turns out to be quite successful.

The Seismic-Geological Comprehensive Prediction Method of the Low Permeability Calcareous Sandstone Reservoir

Currently in Niu-zhuang sub-sag, the seismic reflection amplitude of the newly discovered turbidite sandstone is stronger in the third Segment. The main reason is that Calcareous components accounts for a large part and physical properties is relatively poor, which results in no corresponding relation between reservoir and seismic attributes, and effective reservoir is difficult to predict and describe. Therefore, using the method of geological statistics, we firstly study the distribution of calcareous matters, secondly study the contribution to seismic reflection amplitude made by Calcareous high impedance component;thirdly analyze its influence on actual seismic reflection amplitude and determine the lithology thickness of Calcareous via replacement forward modeling. At last, we characterize the reservoir using the amplitude of calcareous matters. It proves that the method of seismic-geological comprehensive prediction is reliable. It has good guidance for exploration and development of the calcareous sand lithologic reservoir in similar areas.

Prediction of Tight Sand Reservoir with Multi-Wavelet Decomposition and Reconstructing Method

Special reservoir or fluid has an abnormal response to some certain frequencies, so that seismic decomposition and reconstruction are used to highlight the seismic reflection at certain frequencies useful to identify special geological bodies. Because seismic wavelets are time-varying and spatial-variable in the propagation, synthetic traces based on single wavelet make some weak but useful information lost, and make artifacts form. However, Morlet wavelet aggregation with mathematical analytical expression is able to fully and correctly reflect the variations of wavelet in the propagation of underground medium. The matching pursuit algorithm on the basis of Morlet wavelet improves the calculating efficiency in decomposition and reconstruction greatly. This method is applied to the actual study area to do conjoint analysis of single well and well-tie multi-wavelet decomposition. It is found that frequencies sensitive to interest reservoirs range from 8 to 34 Hz. Reconstructing the wavelets at those special frequencies and analyzing the reconstructed seismic data, it is pointed out that interest reservoirs have abnormal characteristics with respectively strong RMS amplitude in the reconstructed data. Crossplot of gamma value at wells and reconstructed RMS amplitude suggests that anomalies caused by interest reservoirs are well separated from the background anomalies when the reconstructed RMS amplitude is greater than 3650. Quantitative prediction results of interest reservoirs distribution in the study area reveal that interest reservoirs of western and northern study area are distributed annularly and bandedly, while most contiguous sandstone in eastern regions appears sporadically.

Successful Lateral Predicting Technique for Reservoirs

ＳｕｃｃｅｓｓｆｕｌＬａｔｅｒａｌＰｒｅｄｉｃｔｉｎｇＴｅｃｈｎｉｑｕｅｆｏｒＲｅｓｅｒｖｏｉｒｓＴｈｅＢｕｒｅａｕｏｆＧｅｃｐｈｙｓｉｃａｌｐｒｏｓｐｅｃｔｉｎｇａｎｄｏｔｈｅｒｅｎｔｅｒｐｒｉｓｅｓｏｆＣＮＰＣｈａｖｅｓｕｃｃｅｓｓｆｕｌｌｙａｐ...

Dissolution mechanism of a deep-buried sandstone reservoir in a deep water area:A case study from Baiyun Sag,Zhujiang River(Pearl River)Mouth Basin

Dissolution mechanism and favorable reservoir distribution prediction are the key problems restricting oil and gas exploration in deep-buried layers.In this paper,the Enping Formation and Zhuhai Formation in Baiyun Sag of South China Sea was taken as a target.Based on the thin section,scanning electron microscopy,X-ray diffraction,porosity/permeability measurement,and mercury injection,influencing factors of dissolution were examined,and a dissolution model was established.Further,high-quality reservoirs were predicted temporally and spatially.The results show that dissolved pores constituted the main space of the Paleogene sandstone reservoir.Dissolution primarily occurred in the coarse-and medium-grained sandstones in the subaerial and subaqueous distributary channels,while dissolution was limited in fine-grained sandstones and inequigranular sandstones.The main dissolved minerals were feldspar,tuffaceous matrix,and diagenetic cement.Kaolinization of feldspar and illitization of kaolinite are the main dissolution pathways,but they occur at various depths and temperatures with different geothermal gradients.Dissolution is controlled by four factors,in terms of depositional facies,source rock evolution,overpressure,and fault activities,which co-acted at the period of 23.8–13.8 Ma,and resulted into strong dissolution.Additionally,based on these factors,high-quality reservoirs of the Enping and Zhuhai formations are predicted in the northern slope,southwestern step zone,and Liuhua uplift in the Baiyun Sag.

The applicability and underlying factors of frequency-dependent amplitude-versus-offset(AVO)inversion

Recently,the great potential of seismic dispersion attributes in oil and gas exploration has attracted extensive attention.The frequency-dependent amplitude versus offset(FAVO)technology,with dispersion gradient as a hydrocarbon indicator,has developed rapidly.Based on the classical AVO theory,the technology works on the assumption that elastic parameters are frequency-dependent,and implements FAVO inversion using spectral decomposition methods,so that it can take dispersive effects into account and effectively overcome the limitations of the classical AVO.However,the factors that affect FAVO are complicated.To this end,we construct a unified equation for FAVO inversion by combining several Zoeppritz approximations.We study and compare two strategies respectively with(strategy 1)and without(strategy 2)velocity as inversion input data.Using theoretical models,we investigate the influence of various factors,such as the Zoeppritz approximation used,P-and S-wave velocity dispersion,inversion input data,the strong reflection caused by non-reservoir interfaces,and the noise level of the seismic data.Our results show that FAVO inversion based on different Zoeppritz approximations gives similar results.In addition,the inversion results of strategy 2 are generally equivalent to that of strategy 1,which means that strategy 2 can be used to obtain dispersion attributes even if the velocity is not available.We also found that the existence of non-reservoir strong reflection interface may cause significant false dispersion.Therefore,logging,geological,and other relevant data should be fully used to prevent this undesirable consequence.Both the P-and S-wave related dispersion obtained from FAVO can be used as good indicators of a hydrocarbon reservoir,but the P-wave dispersion is more reliable.In fact,due to the mutual coupling of P-and S-wave dispersion terms,the P-wave dispersion gradient inverted from PP reflection seismic data has a stronger hydrocarbon detection ability than the S-wave dispersion gradient.Moreover,there is little difference in using post-stack data or pre-stack angle gathers as inversion input when only the P-wave dispersion is desired.The real application examples further demonstrate that dispersion attributes can not only indicate the location of a hydrocarbon reservoir,but also,to a certain extent,reveal the physical properties of reservoirs.

Petrophysical parameters inversion for heavy oil reservoir based on a laboratory-calibrated frequency-variant rock-physics model

Heavy oil has high density and viscosity, and exhibits viscoelasticity. Gassmann's theory is not suitable for materials saturated with viscoelastic fluids. Directly applying such model leads to unreliable results for seismic inversion of heavy oil reservoir. To describe the viscoelastic behavior of heavy oil, we modeled the elastic properties of heavy oil with varying viscosity and frequency using the Cole-Cole-Maxwell (CCM) model. Then, we used a CCoherent Potential Approximation (CPA) instead of the Gassmann equations to account for the fluid effect, by extending the single-phase fluid condition to two-phase fluid (heavy oil and water) condition, so that partial saturation of heavy oil can be considered. This rock physics model establishes the relationship between the elastic modulus of reservoir rock and viscosity, frequency and saturation. The viscosity of the heavy oil and the elastic moduli and porosity of typical reservoir rock samples were measured in laboratory, which were used for calibration of the rock physics model. The well-calibrated frequency-variant CPA model was applied to the prediction of the P- and S-wave velocities in the seismic frequency range (1–100 Hz) and the inversion of petrophysical parameters for a heavy oil reservoir. The pre-stack inversion results of elastic parameters are improved compared with those results using the CPA model in the sonic logging frequency (∼10 kHz), or conventional rock physics model such as the Xu-Payne model. In addition, the inversion of the porosity of the reservoir was conducted with the simulated annealing method, and the result fits reasonably well with the logging curve and depicts the location of the heavy oil reservoir on the time slice. The application of the laboratory-calibrated CPA model provides better results with the velocity dispersion correction, suggesting the important role of accurate frequency dependent rock physics models in the seismic prediction of heavy oil reservoirs.

Evaluation of reservoir environment by chemical properties of reservoir water‒A case study of Chang 6 reservoir in Ansai oilfield,Ordos Basin,China

The Ordos Basin is the largest continental multi-energy mineral basin in China,which is rich in coal,oil and gas,and uranium resources.The exploitation of mineral resources is closely related to reservoir water.The chemical properties of reservoir water are very important for reservoir evaluation and are significant indicators of the sealing of reservoir oil and gas resources.Therefore,the caprock of the Chang 6 reservoir in the Yanchang Formation was evaluated.The authors tested and analyzed the chemical characteristics of water samples selected from 30 wells in the Chang 6 reservoir of Ansai Oilfield in the Ordos Basin.The results show that the Chang 6 reservoir water in Ansai Oilfield is dominated by calcium-chloride water type with a sodium chloride coefficient of generally less than 0.5.The chloride magnesium coefficients are between 33.7 and 925.5,most of which are greater than 200.The desulfurization coefficients range from 0.21 to 13.4,with an average of 2.227.The carbonate balance coefficients are mainly concentrated below 0.01,with an average of 0.008.The calcium and magnesium coefficients are between 0.08 and 0.003,with an average of 0.01.Combined with the characteristics of the four-corner layout of the reservoir water,the above results show that the graphics are basically consistent.The study indicates that the Chang 6 reservoir in Ansai Oilfield in the Ordos Basin is a favorable block for oil and gas storage with good sealing properties,great preservation conditions of oil and gas,and high pore connectivity.

Theoretical and technical progress in exploration practice of the deep-water large oil fields, Santos Basin, Brazil

The history and results of petroleum exploration in the Santos Basin, Brazil are reviewed. The regularity of hydrocarbon enrichment and the key exploration technologies are summarized and analyzed using the seismic, gravity, magnetic and drilling data. It is proposed that the Santos Basin had a structural pattern of two uplifts and three depressions and the Aram-Uirapuru uplift belt controlled the hydrocarbon accumulation. It is believed that the main hydrocarbon source kitchen in the rift period controlled the hydrocarbon-enriched zones, paleo-structures controlled the scale and quality of lacustrine carbonate reservoirs, and continuous thick salt rocks controlled the hydrocarbon formation and preservation. The process and mechanism of reservoirs being transformed by CO_(2)charging were revealed. Five key exploration technologies were developed,including the variable-velocity mapping for layer-controlled facies-controlled pre-salt structures, the prediction of lacustrine carbonate reservoirs, the prediction of intrusive/effusive rock distribution, the detection of hydrocarbons in lacustrine carbonates, and the logging identification of supercritical CO_(2)fluid. These theoretical recognitions and exploration technologies have contributed to the discovery of deep-water super-large reservoirs under CNODC projects in Brazil, and will guide the further exploration of deep-water large reservoirs in the Santos Basin and other similar regions.

Lithofacies identi cation using support vector machine based on local deep multi-kernel learning

Lithofacies identification is a crucial work in reservoir characterization and modeling.The vast inter-well area can be supplemented by facies identification of seismic data.However,the relationship between lithofacies and seismic information that is affected by many factors is complicated.Machine learning has received extensive attention in recent years,among which support vector machine(SVM) is a potential method for lithofacies classification.Lithofacies classification involves identifying various types of lithofacies and is generally a nonlinear problem,which needs to be solved by means of the kernel function.Multi-kernel learning SVM is one of the main tools for solving the nonlinear problem about multi-classification.However,it is very difficult to determine the kernel function and the parameters,which is restricted by human factors.Besides,its computational efficiency is low.A lithofacies classification method based on local deep multi-kernel learning support vector machine(LDMKL-SVM) that can consider low-dimensional global features and high-dimensional local features is developed.The method can automatically learn parameters of kernel function and SVM to build a relationship between lithofacies and seismic elastic information.The calculation speed will be expedited at no cost with respect to discriminant accuracy for multi-class lithofacies identification.Both the model data test results and the field data application results certify advantages of the method.This contribution offers an effective method for lithofacies recognition and reservoir prediction by using SVM.

Lateral Contrast and Prediction of Carboniferous Reservoirs Using Logging Data in Tahe Oilfield,Xinjiang,China

Valuable industrial oil and gas were discovered in the formations of Ordovician, Carboniferous and Triassic of the Tahe （塔河） oilfield, Xinjiang （新疆）, China. The Carboniferous formations contain several oil- and gas-bearing layers. The lateral distribution of Carboniferous reservoir is unstable, and thin layers are crossbedded. This makes it difficult to do lateral formations＇ contrast and reservoir prediction, so it is necessary to develop a method that can achieve reservoir lateral contrast and prediction by using multi-well logging data and seismic data. To achieve reservoir lateral contrast and prediction at the Carboniferous formations of the Tahe oilfield, processing and interpretation of logging data from a single well were done first. The processing and interpretation include log pretreatment, en- vironmental correction and computation of reservoir＇s parameters （porosity, clay content, water saturation, etc.）. Based on the previous work, the data file of logging information of multi-well was formed, and then the lateral distribution pictures （2D and 3D pictures of log curves and reservoir parameters） can be drawn. Comparing multi-well＇s logging information, seismic profiles and geological information （sedimentary sign）, the reservoir of the Carboniferous in the Tahe oilfield can be contrasted and pre- dicted laterally. The sand formation of Carboniferous can be subdivided. The results of reservoir contrast and prediction of the Carboniferous formations show that 2D and 3D pictures of multi-weU reser- voir parameters make the lateral distribution of reservoir and oil-bearing sand very clear, the connectedness of the reservoir of neighboring wells can be analyzed, and five sand bodies can be identified based on the reservoir＇s lateral distribution, geological information and seismic data.