Acoustic Logging Research Articles

A Three-Dimensional Display Technology of Reflected Acoustic Logging and Its Application Example

Reflection acoustic logging is a new technology for identifying geological structures around the well. It can be used to describe the geological structure, such as formation interface, fracture, and limestone cave. Based on the acoustic logging tool composed of the source and array receiver, the method can be used to image the geological structure around the well. Through migration imaging, the section crossing the well can be obtained in any azimuth. In the current three-dimensional (3D) display technology of reflection acoustic logging, the reflection plane around the well is always displayed as a reflection arc. The arc surface correction method can correct the arc surface in 3D imaging into a plane. According to the similarity criterion in a physical experiment and the need for practical research, we designed a scale model well experiment of dipole reflection acoustic logging. We obtained the 3D display of the reflection structure around the well by the experiment and realized the arc surface correction of the 3D display in reflection acoustic logging. The application shows the method can qualitatively describe the lateral extension length of the geological structure around the well. After the arc surface correction, the geological structure can be more accurately described.

Electrical Interference Simulation and Prediction Model for Acoustoelectric Logging Detector.

Acoustic logging instruments generate high voltages in the order of thousands of volts. Electrical interferences are thus induced by high-voltage pulses that affect the logging tool and make it inoperable owing to damaged components in severe cases. High-voltage pulses from the acoustoelectric logging detector interfere with the electrode measurement loop through capacitive coupling, which has seriously affected the acoustoelectric signal measurements. In this paper, we simulate high voltage pulses, capacitive coupling and electrode measurement loops based on qualitative analysis of the causes of electrical interference. Based on the structure of the acoustoelectric logging detector and the logging environment, an electrical interference simulation and prediction model was developed to quantify the characteristics of the electrical interference signal.

Effective elastic properties and S-wave anisotropy for rocks containing any oriented penny-shaped cracks in transversely isotropic background

Fractures have a significant influence on rock elastic properties. The existing elastic models for cracked medium mainly consider the crack dip angle. In this work, based on the previous transversely isotropic (TI) background and arbitrary dip crack model, a theoretical model is derived using the Bond transform matrix to quantify the effects of any oriented penny-shaped cracks on the elastic properties of rocks with a TI background. Based on this model, the effects of crack orientation on rock elastic properties are studied. Particularly, a special case with random crack orientation is investigated by using the spherical averaging of the crack compliance matrix with any orientations. Furthermore, the S-wave anisotropy for rocks containing multiple crack sets also is analyzed. The results indicate that the crack orientation has a great influence on rock elastic properties. Especially, the rock elastic properties indicate nonlinear changes with crack azimuth angles when cracks are inclined with respect to the TI background isotropic plane. The randomly oriented cracks decrease the overall rock elastic moduli but have little effect on rock anisotropy. For rocks containing multiple crack sets, parallel crack sets have great effects on the S-wave anisotropy, whereas the effects of orthogonal crack sets are negligible. Using this model, a complete workflow for the inversion of fracture properties using field data measured by acoustic logging is introduced. To explain the effect of crack azimuth angle, the crack density is inverted without considering crack orientation, only considering crack dip angle, and considering both crack dip and azimuth angles. The results indicate that the inversion results considering the crack dip and azimuth angles are in good agreement with the resistivity imaging results. This work provides the theoretical basis for fracture properties inversion for rocks with a TI background.

Classification and well-log identification of lithofacies of continental shale oil reservoirs in Chang 7-3 submember of the Triassic Yanchang Formation in the Longdong area of the Ordos Basin, China

The division and identification of lithology and lithofacies are very important in exploration and development of oil and gas reservoirs. Information on lithofacies determines the development modes and provides the favorable “sweet spot” intervals of shales. The Chang 7-3 submember of the Triassic Yanchang Formation in the Longdong area of the Ordos Basin is an important shale formation in China. Core photos, cast thin sections, and log response characteristics were used to clarify the development and distribution characteristics of the various lithofacies in this formation. The lithofacies of target intervals are classified into fine-grained sandy debrite, silty to fine-grained sandy turbidite, fine-grained sandy slump, semideep to deep lacustrine mudstone, and deep lacustrine shale. Furthermore, the conventional and microresistivity imaging log response characteristics of different lithofacies are analyzed. Compared with other lithofacies, deep lacustrine shale facies are characterized by high gamma ray (GR), high compensated neutron log (CNL), high acoustic log (AC), and low-density (DEN) log values. Crossplots of GR-AC, DEN-CNL, and DEN-AC are proposed to identify most of the lithofacies, and a criterion was established for use of well logs to characterize the lithofacies. Finally, a method joining conventional logs and microresistivity imaging log data was applied to field wells, achieving the very high accuracy rate of 88%. The distribution maps of different lithofacies thickness in the study area are obtained. Deep lacustrine shale facies have the widest development area and many thickness centers. Accurate and effective identification of lithofacies indicates that within the Ordos Basin, there is an area of approximately 20,000 km2 with shale thickness of greater than 6 m, providing an important basis for shale oil exploration in the study area.

An Unsupervised Machine-Learning Workflow for Outlier Detection and Log Editing With Prediction Uncertainty

Recent advances in data science and machine learning (ML) have brought the benefits of these technologies closer to the mainstream of petrophysics. ML systems, where decisions and self-checks are made by carefully designed algorithms, in addition to executing typical tasks such as classification and regression, offer efficient and liberating solutions to the modern petrophysicist. The outline of such a system and its application in the form of a multilevel workflow to a 59-well multifield study are presented in this paper. The main objective of the workflow is to identify outliers in bulk density and compressional slowness logs and to reconstruct them using data-driven predictive models. A secondary objective of the project is to predict shear slowness in zones where such data do not exist. The system is fully automated, designed to optimize the use of all available data, and provide uncertainty estimates. It integrates modern concepts for outlier detection, predictive classification, and regression, as well as multidimensional scaling based on inter-well similarity. Benchmarking of ML results against those created by experienced petrophysicists shows that the ML workflow can provide high-quality answers that compare favorably to those produced by human experts. A second validation exercise, that compares acoustic impedance logs computed from ML answers to actual seismic data, provides further evidence for the accuracy of the ML-generated results. The ML system supports the petrophysicist by easing the burden on repetitive and burdensome quality control tasks. The efficiency gains and time savings created can be used for enhanced effective cross-discipline integration, collaboration, and further innovation.

Dolomite reservoir fluid discrimination method based on fluid identification factor

The complexity of the pore structure of dolomite leads to strong heterogeneity of the reservoir distribution and the inability to accurately describe the rock-electric parameters, which makes it difficult to distinguish the properties of reservoir fluids based on electrical logging. In this paper, fluid identification factors are applied to dolomite reservoirs for the first time, which enriches the methods of non-electric logging to identify reservoir fluid properties. Based on the inductive analysis of existing fluid identification factors, according to the number of times that logging data participates in the calculation, they are divided into three types of fluid identification factors: direct, indirect and composite; composite fluid identification factors can reflect changes in reservoir lithology, and reconstruct composite fluid identification factor H considering the difference between the lateral deformation coefficient of dolomite and other lithologies; the sensitivity of the composite fluid identification factor to gas saturation is tested, and fluid factors used in dolomite reservoirs were optimized. Select the array acoustic logging and density logging data from three wells in the subsalt dolomite in the Tarim Basin, China, calculate the parameters of the compressional wave and shear wave impedances (Ip, Is), λρ, μρ, ρf, F, H and etc., and use the two-dimensional chart method to analyze the effect of the fluid identification factors above, and λρ-μρ, F-σ, ρf-σ, H-μ and other charts are selected to establish logging fluid discriminant curves. The application shows that the fluid identification factor H is more suitable for distinguishing gas layers, and F is more suitable for distinguishing water layers.

A hybrid CNN-LSTM machine learning model for rock mechanical parameters evaluation

With the continuous deepening of hydrocarbon exploration and development, the exploration proportion of unconventional fields has gradually increased, and unconventional energy has become the main component of oil and gas production. Rock mechanical parameters are key to the efficient development of unconventional fields. However, rock mechanical parameters are mostly obtained from expensive measurements such as coring. In this study, a hybrid CNN-LSTM machine learning model is proposed to obtain rock mechanical parameters efficiently through predicting acoustic log in a highly accurate manner. The model incorporates a hybrid multi-channel CNN-LSTM network, which has the advantage of capturing the temporal and spatial features of logs. A group of feature enhancement methods are also proposed and incorporated into the model to improve the reliability of prediction. In this work, the model is utilized to predict acoustic log. Static rock mechanical parameters, which are typically hard to obtain, could be calculated based on dynamic parameters derived from predicted acoustic log. The performance of model is verified by the data of three wells in Jimsar field, and the results show that the introduced model has the highest prediction accuracy for acoustic logs. And the static rock mechanical parameters evaluated via proposed model are subsequently validated against coring data. A good match is achieved, which demonstrates the feasibility of using this model to guide the efficient development of unconventional fields in future.

A novel evaluation scheme of resistivity anisotropy in near-tight sandstones using conventional geophysical logs: A case study of the Triassic Chang 8 oil layer, Zhenjing area, Ordos Basin

The resistivity anisotropy characteristic of near-tight sandstone reservoirs is important for reservoir evaluation, but it is a challenge to obtain it by using conventional logging curves in vertical boreholes. Thus, a novel evaluation scheme is established. Firstly, in the Chang 8 Formation of Zhenjing area, Ordos Basin, China, horizontal rock samples (parallel to bedding) and vertical rock samples (perpendicular to bedding) are collected for designed rock electricity experiments and analyzed to obtain the corresponding horizontal and vertical Archie's parameters. Secondly, based on the detection characteristics of acoustic logging instrument and the transformation form of the Archie's equation, an estimation method of vertical resistivity is established. Combined with the horizontal resistivity obtained from conventional deep resistivity logs in vertical boreholes, an estimation method of resistivity anisotropy coefficient is proposed. Thirdly, the relationships between resistivity anisotropy coefficients and water saturations and porosities are analyzed using experiment results, and four adjacent vertical wells in the study area are processed using the estimated method. The processed results demonstrate the feasible of the estimation method. In addition, the cross well profile of resistivity anisotropy coefficient in the target formation is characterized, which shows the target formation is slightly resistivity anisotropic and slightly heterogeneous. The relationships between the resistivity anisotropy coefficient and commonly used logs are discussed, which contribute to rapid analyze the resistivity anisotropy of the target formation. Based on the above methods and processes, the novel evaluation scheme of resistivity anisotropy is established, which contributes to quantify and qualitatively analyze the resistivity anisotropy characteristics of formations in and between vertical boreholes. The application results show that it provides an alternative scheme for continuously evaluating the resistivity anisotropy in boreholes and cross well profiles by using conventional logs and experiments.

Fracture characterization of Asmari Formation carbonate reservoirs in G Oilfield, Zagros Basin, Middle East

The Asmari Formation in the G oilfield on the Iran-Iraq border is a fractured-porous multi-lithology mixed reservoir, for which fracture is an important factor affecting oil productivity and water cut. The characterization and modeling of fractures in the carbonate reservoir of G oilfield are challenging due to weak conventional well log responses of fractures and a lack of specific logs, such as image logs. This study proposes an integrated approach for characterizing and modeling fractures in the carbonate reservoir. The features, formation mechanism, influencing factors, and prediction methods of fractures in the Asmari Formation carbonate reservoirs of G oilfield were studied using core observation, thin section, image log, cross-dipole acoustic log (CDAL), geomechanics numerical simulation (GNS), and production data. According to CDAL-based fracture density interpretation, GNS-based fracture intensity prediction between wells, and DFN-based rock fracture properties modeling, the quantitative fracture characterization for G oilfield was realized. This research shows that the fractures in the Asamri Formation are mainly medium-to high-angle shear fractures. The substantial compression stress during the Miocene played a major role in the formation of the prominent fractures and determined their trend in the region, with primary trends of NNW-SSE and NNE-SSW. The fracture distribution has regularity, and the fractures in zone A dolomites are more highly developed than that in zone B limestones vertically. Horizontally, fractures intensity is mainly controlled by faults and structural location. The results of this study may benefit the optimization of well design during field development. From 2019 to 2021, three horizontal wells pilot tests were deployed in the fractures belt in zone A, and these fractures prominently increased the permeability of tight dolomite reservoirs. The initial production of the wells is four to five times the average production of other wells in the area, showing a good development effect. Meanwhile, the updated numerical simulation validates that the history match accuracy of water cut based on the dual-porosity model is significantly improved, proving the fracture evaluation and prediction results to be relatively reliable and applicable.

Using a Synthetic Borehole Model to Determine the Pore Aspect Ratio Dependence of Velocities from Acoustic Well-Logging Data

Revealing the elastic wave properties of carbonate rocks with complex pore structures and improving the reliability of carbonate reservoir descriptions have always been a global challenge in the field of carbonate geophysical exploration. In this study, we established a synthetic borehole model by selecting different particle sizes of cement, carbonate cuttings, and micro-silicon as the matrix, and silicon disks as the pores in carbonate rocks. We conducted four sets of low-porosity (0-3%) borehole models with different pore aspect ratios (ARs) and measured the P- and S-wave velocities ( V P and V S ) at the well-logging scale obtained from an acoustic logging system with one source and two receivers. The results indicate that the relationship between velocities and porosities in these borehole models follows a linear relation, with the pore AR significantly influencing the velocities at any given porosity. The velocity variation caused by pore AR reaches 560 m/s and 410 m/s at 3% porosity for the P-wave and S-wave within the AR range of 0.017-0.13. The theoretical DEM models provide a high and broad estimation of V P and V S at the well-logging scale in our measurement. They could perform better in fractured formation than in dissolved porous formation in carbonate reservoirs. The linear relation of V P and V S is independent of the pore AR and is effective for both fractured and dissolved porous formations. The change of V P / V S in different pore AR is more responsive to porosity and nonlinear dependent on the pore AR. The relationship between the defined normalized V P and V S indicates the pore AR has a more significant effect on V S than V P in our model. The constructed borehole models provide a unique opportunity for evaluating the availability of rock physics models at an acoustic logging scale. The study’s findings have significant implications for improving the reliability of carbonate reservoir descriptions and enhancing the accuracy of geophysical exploration in carbonate rocks with complex pore structures.

Experimental Study on the Azimuthal-Transmitting Transducer of Acoustic Logging

Azimuthal acoustic logging can survey the downhole formation more accurately, and the acoustic source is the crucial component of the downhole acoustic logging tool with azimuthal resolution characteristics. To realize downhole azimuthal detection, assembling multiple transmitting piezoelectric vibrators in the circumferential direction is necessary, and the performance of azimuthal-transmitting piezoelectric vibrators needs attention. However, effective heating test and matching methods are not yet developed for downhole multi-azimuth transmitting transducers. Therefore, this paper proposes an experimental method to comprehensively evaluate downhole azimuthal transmitters; furthermore, we analyze the azimuthal-transmitting piezoelectric vibrator parameters. This paper presents a heating test apparatus and studies the admittance and driving responses of the vibrator at different temperatures. The transmitting piezoelectric vibrators showing a good consistency in the heating test were selected, and an underwater acoustic experiment was performed. The main lobe angle of the radiation beam, horizontal directivity, and radiation energy of the azimuthal vibrators and azimuthal subarray are measured. The peak-to-peak amplitude radiated from the azimuthal vibrator and the static capacitance increase with an increase in temperature. The resonant frequency first increases and then decreases slightly with an increase in temperature. After cooling to room temperature, the parameters of the vibrator are consistent with those before heating. Hence, this experimental study can provide a foundation for the design and matching selection of azimuthal-transmitting piezoelectric vibrators.

Research on Improvement Measures of Cementing Acoustic Amplitude Logging

Cementing is one of the important aspects in the exploration and development of oil and gas fields, and the quality of cementing has a great impact on the life, production capacity and overall efficiency of a well. Due to the limitations of the evaluation technology and the complexity of the influencing factors, the conventional method of cementing quality evaluation currently used in China often leads to misjudgment of the quality of the cemented wells and can no longer meet the requirements of high accuracy research. By analyzing the factors affecting the solids quality testing and the advantages and disadvantages of several common solids quality evaluation methods, and organically combining them according to the measurement principles and their usage correlations, a system of evaluation methods applicable to different requirements is formed, and the intelligent evaluation of solids quality is further explored to provide a concrete and practical basis for solids quality evaluation in the development and production of oil fields.

Excess Anisotropy: A Method to Predict Frequency of Resistive Fractures in the Wolfcamp Shale

A horizontal well in the Wolfcamp Formation within the Delaware Basin exhibits cross-dipole sonic log anisotropy varying from 0.0% to 29.7% over the reservoir interval. Massive limestones exhibit little anisotropy while shales and intervals with high fracture counts have high anisotropy. Integrated analysis of acoustic, image and triple combo open-hole log data indicates that the computed volume of shale log is a strong predictor of anisotropy with a simple linear relationship. However, there is much anisotropy that is not accounted for by this lithologically related linear relationship. We refer to the residual obtained by subtracting the predicted anisotropy using this relationship from the measured anisotropy from a cross-dipole sonic tool as the excess anisotropy. The excess anisotropy is the remaining anisotropy that is not accounted for by linear regression between the volume of shale log and the measured anisotropy. We find that the excess anisotropy is linearly correlated to the resistive fracture density. Linear regression between the two variables yields a statistically significant relationship that enables a rough estimation of the resistive fracture frequency from the excess anisotropy. Validation in a second well produces a similar correlation to fracture frequency.

Typical misinterpretations and scientific concepts in well-logging geologic studies

The method and theoretical system of well logging geology have been widely used in the fields of basic geology, petroleum geology and engineering geology, but the different response sensitivity of different well logging series to geological information and the mismatching between geophysical properties of multiple well logs and geological genesis of rocks frequently result in misunderstandings in the research process of well logging geology. Therefore, it is in an urgent need to analyze the typical misunderstanding cases in the research of well logging geology and explore the corresponding scientific ideas and countermeasures. After analyzing the typical misunderstandings in the research of well logging geology, this paper investigates vertical resolution scale of various logging series and its contradiction with detection depth and illustrates the importance of the integration of different scales of data. In addition, the factor inducing “fake logging data” and its influence on interpretation evaluation are clarified and a set of ideas for well logging evaluation of geological interpretation is put forward. And the following research results are obtained. First, the typical misunderstandings in the research of well logging geology can be classified into two categories, namely geological body interpretation misunderstanding and reservoir property parameter calculation misunderstanding. Second, special geological phenomena, such as high-density and high-resistivity mudstone can lead to logging data ambiguity, so attention shall be paid to petrophysical response mechanisms during geological logging interpretation. Third, to carry out well logging evaluation of unconventional oil and gas, it is necessary to integrate new technologies of electric imaging logging, dipole acoustic logging and nuclear magnetic resonance logging, and the calibration of core data and the integration of geological ideas can improve the interpretation accuracy. Fourth, In the process of borehole structural logging analysis, sedimentary response, geostress evaluation and fracture identification, geological ideas shall be integrated to avoid the logging interpretation misunderstanding caused by the same response of different geological phenomena in well logs. In conclusion, the dialectical and systematic thinking from geology to logging and then to geology, from practice to recognition and then to practice and from “a narrow view” to “a broad view” can provide a scientific ideas for the comprehensive research of well logging geology.

Improving Prediction of Fracture Distribution Using Microseismic Data and Acoustic Logging Measurements

Summary The complex fracture network from hydraulic fracturing can significantly improve oilwell productivity, so it is widely used in the field of unconventional reservoir development. However, accurate evaluation of the fracture spatial distribution remains a challenge. As a result, how to combine a variety of data to avoid data islands and identify and predict the space of fracture zone is of great importance. In this paper, we present a method and workflow based on the microseismic (MS) data combined with shear wave velocity data to estimate the physical parameters of subsurface media and improve the description and prediction accuracy for hydraulic fractures. The method analyzes MS events to construct the fracture spatial distribution and uses acoustic logging measurements to correct the magnitude of MS events and enhance the resolution. The corrected MS magnitude is mapped to the MS event space for Kriging interpolation analysis to predict the improved spatial distribution of fractures, which is available in the format of a 3D cloud image.

Detailed Geoseismic Structure of the Riphean Deposits in Buolkalakh–Lena Area Based on Deep Drilling and Seismic Exploration Data

Abstract —The work considered an integrated approach to the study of deep-lying Riphean strata of the Buolkalakh–Lena region, which have certain prospects in terms of oil and gas potential. The method of revision and synthesis of acoustic and density logs is described in detail using the example of wells drilled in 80–90 for the purpose of correct reference to seismic sections. It has been shown that the Khaipakh, Debengdin, Arymas, Kyutingdinsky, and Sygynakhtakh formations of the Early and Middle Riphean eras overlook the pre-Permian surface in the middle part of the Lena–Anabar trough. The region also developed older Riphean deposits that are not exposed on the Olenek uplift and are not studied by deep drilling. The prospects for the oil and gas content of Riphean strata in the region are mainly associated with carbonate reservoirs formed in the top of the Riphean sequence and cropping out on the pre-Permian erosive surface, the most promising rocks are those constituting large carbonate platforms that were partially eroded and subjected to processes of hypergenesis and karst formation.

3D Viscoelastic Finite-Difference Analysis of the Monopole Acoustic Logs in Cylindrical Coordinates

In this paper, a model of the heterogeneous anelastic seismic wave problem is proposed in three-dimensional (3D) cylindrical coordinates. We use the velocity-stress formula to describe the realistic attenuation properties of viscoelastic materials, derived from a rheological model of the generalized standard linear solid (GSLS). The equation system is completed by additional equations for the anelastic functions including the strain history of the material. We apply the staggered grid finite-difference (FD) method in 3D cylindrical coordinates to solve the equations. Moreover, to avoid the effect of gradual expansion of the grid size as the radius increases, we use a variable grid method to achieve compensation. In real drilling operations, the mud injected in the borehole is a fluid with viscous properties. The actual formation is also not elastic. In the synthetic data of acoustic logging while drilling (LWD), we find that the drill collar wave is not affected by the viscoelastic parameters of the formation. In contrast, the Stoneley wave is more sensitive to the viscosity of the drilling fluid. The phase and amplitude of the received waveform are affected by the drilling fluid as well as the formation viscoelasticity. Therefore, the development of the cylindrical variable-grid FD method provides a flexible and efficient numerical technique to solve 3D viscoelastic wave propagation problems, including realistic attenuation and complex geometry.

A Reconstructed Method of Acoustic Logging Data and Its Application in Seismic Lithological Inversion for Uranium Reservoir

As a sedimentary mineral, most sandstone type uranium deposits are formed in petroliferous basins. Therefore, we can fully tap the residual economic value of historical logging and 3D seismic data measured for oil and gas to search for sandstone type uranium deposits. However, a large amount of acoustic logging data are missing in the target stratum of the uranium reservoir in that it is not the main stratum of oil and gas. A reconstructed method of acoustic logging data based on clustering analysis and with the low-frequency compensation of deterministic inversion is proposed to solve this problem. Secondly, we can use these logging data with seismic data to obtain the 3D inversion data volume representing the sand body of the uranium reservoir based on seismic lithological inversion. Then, we can also delimit the 3D spatial range of sandstone type uranium deposits in petroliferous basins based on the calibration of uranium anomaly and sub-body detection. Finally, a 3D field data example is given to test and analyze the effectiveness of the above research schemes.

Robust estimation of shear wave velocity in a carbonate oil reservoir from conventional well logging data using machine learning algorithms

In recent decades, various applications of shear wave velocity have been reported in oil and gas projects. These applications include the determination of lithology, estimation of geomechanical parameters, reservoir fluid detection, etc. The shear wave velocity is usually measured or recorded by DSI (Dipole Shear Sonic imager) instruments. This log is not usually available in all wells of a hydrocarbon field because of the high cost of running this log compared to other conventional logs. However, it is tried to estimate the shear wave velocity from other related logs.The literature contains many empirical rock physics models used to estimate shear wave velocity. Most of them were developed for sandstone lithologies and didn't applicable in carbonate reservoir because of existing some vertical and horizontally complex fracture networks in the carbonate reservoirs that cause a quite different behavior for the shear wave velocity.The main objective of this research is to concentrate on the effect of lithology in the estimation of the target log for a carbonate reservoir. At the earliest step, preprocessing is performed and 455 data points of several conventional well logs, i.g., acoustic, density, neutron porosity, resistivity, gamma-ray, calcite volume, dolomite volume, and water saturation logs have been applied to generate the synthetic shear wave velocity log in a carbonate oil reservoir.First, original data are divided into training, validation, and testing subsets. Different data-driven predictive models were built: Multiple Linear Regression (MLR), Ensemble Learning Method (ELM), and three architectures of Artificial Neural Networks (ANN). The performance of all techniques was compared with each other. Finally, the feed-forward neural network shows the highest accuracy with R-values of 0.99 and 0.96 for the training and testing datasets. Moreover, the network design and tuning parameters of the model are adjusted using the grid search optimization technique to obtain the optimal design of the artificial neural network for estimating the shear wave velocity log. Hence, a deep artificial neural network, i.e., [8–11–15-1] network with bayesian regularization training algorithm and hyperbolic tangent sigmoid transfer function, is proposed to predict the target response in other wells.

Assessment of reservoir stress state and its implications for Paleozoic tight oil reservoir development in the Oued Mya Basin, northeastern Algerian Sahara

• In-situ stress regime and pore pressure of the tight Paleozoic reservoirs. • A mean NW-SE S HMax orientation inferred from the B-D quality wellbore breakouts. • Multiple linear regression models to predict the S hmin and S HMax magnitudes. • Estimated pore pressure and in-situ stress behavior agree with observed data. • Optimum drilling and stimulation strategies are inferred. The Cambrian and Ordovician clastic reservoirs of the Oued Mya Basin exhibit significant vertical thickness and extensive lateral continuity, despite being tight. These reservoir intervals have not been properly understood yet in terms of in-situ stress distribution and pore pressure behaviour. The main objectives were to infer the reservoir stress state and draw implications for the tight oil reservoir development based on the geomechanical analyses. We interpreted breakouts from a cumulative 1485 m of acoustic image logs and interpreted a NW-SE S HMax orientation (N125°E-N147°E) in the Oued Mya Basin. The inferred breakouts were of B-D quality as per the World Stress Map ranking criteria. Both the reservoirs have a pore pressure gradient of 13.58-13.77 MPa/km, while the minifrac data infers a reservoir S hmin gradient of 17.3-19.2 MPa/km. Based on the breakout widths, we estimated the S HMax gradient as 23.8-26.5 MPa/km. Following the univariate regression analyses to identify various influencing parameters on horizontal stress magnitudes, we proposed multiple linear regression (MLR) models to predict the S hmin and S HMax based on pore pressure, S v , Poisson's ratio, and Young's modulus. Results indicate that S v influences the horizontal stress estimates significantly more as compared to the other influencing variables. The predicted S hmin and S HMax values are in good agreement (goodness of fit as R 2 = 0.976 and 0.994) with the measured data. The newly proposed MLR equations can be utilized in absence of subsurface validation data. A strike-slip faulting reservoir stress state is concluded from stress polygon analysis. An optimum drilling strategy is discussed based on the observed wellbore failures. We recommended the drilling fluid pressure to be increased by 8 MPa and 14 MPa to avoid breakouts against the Ordovician and Cambrian reservoirs respectively, however, that may incur tensile fractures which do not have a considerable effect on wellbore stability while drilling. Based on this work, horizontal well trajectory along NE-SW (i.e., parallel to S hmin ), together with oriented perforations aligned parallel to inferred S HMax direction is recommended. The potential fracture reactivation risks during reservoir pressurization are evaluated and discussed.