Sand body are a kind of common reservoir unit, and their accurate identification and tracking are the key to discovering oil and gas fields and supporting the increase in oil and gas reserves and production. Existing methods such as attribute analysis and deep learning still face challenges such as low boundary identification accuracy, complex parameter selection, and poor noise resistance. To this end, this paper proposes a prompt prediction method based on the Segment Anything Model (SAM), a visual image segmentation foundation model. This method requires no model training, and by simply employing the boundary prompt points of target sand bodies, precise identification and tracking results of the boundaries of target sand bodies can be obtained. To address the prediction error of the prompt encoder in SAM when applied to seismic profiles, this paper proposes a KD-tree search method. By calculating the shortest distance from the prompt points to the potential sand body segmentation blocks, the optimal sand body prediction results are determined. After conducting verification with actual target area data and comparison with the customized training of a U-Net model on the target area data, it is demonstrated that the sand body tracking method based on SAM depicts more lateral changes of sand body boundaries and the boundaries are more consistent with seismic amplitude variations.
Accurate estimation of coalbed methane (CBM) content plays a crucial role in assessing and efficiently exploiting CBM resources. Deep CBM is influenced by multiple controlling factors and complex genetic mechanisms. Currently, machine-learning approaches for CBM content prediction typically rely on either seismic or logging data. As a result, the complex geological conditions of deep coal seam are not fully accounted for. This study proposes an intelligent prediction method for CBM content, which achieves multi-source data fusion through a multi-scale modeling and deep integration strategy. The approach first extracts multi-scale sensitive attributes or features relevant to CBM content from geological, logging, and seismic sources. For each dataset of the same scale, adaptive modeling is performed using a Bayesian hyperparameter-optimized random forest (RF) algorithm, which enhances model robustness and prevents overfitting. The prediction results from individual scales are subsequently integrated through the least squares method to construct a multi-scale RF composite model. The proposed method is validated using a field dataset and compare its performance with that of conventional approaches, including single-scale RF and linear regression. The results show that, compared with these baseline methods, the proposed method reduces the mean relative error of CBM content prediction on test wells by 3.01% and 4.94%, respectively. This demonstrates that the proposed approach achieves higher accuracy and stronger generalization capability, enabling precise characterization of the spatial distribution of CBM content.
Natural fracture systems serve as the storage space and seepage channel of shale gas reservoirs, and fracture identification of shale reservoirs is faced with such problems as high cost and low coverage of micro-resistivity imaging logging, as well as significant nonlinear characteristics of conventional logging data and strong subjectivity of manual interpretation. To this end, this paper screens four kinds of logging curves with high sensitivity to fractures of shale gas reservoirs, including the deep lateral resistivity, natural gamma ray, neutron porosity, and interval transit time via sensitivity analysis. Meanwhile, the first-order difference curve of resistivity and its product curve are introduced, and the complex temporal sequence relationship of conventional logging curves is transformed into a classifiable threshold discrimination problem. Subsequently, a BP neural network identification model for fractures in shale gas reservoirs is built based on the conjugate gradient descent optimization algorithm. The results demonstrate that this model can effectively eliminate the subjective bias in manual interpretation. Compared with the actual fractures, the recall ratio of fracture identification in shale gas reservoirs reaches 90%, with a precision rate of 87%. This research provides a novel approach for fracture identification of unconventional hydrocarbon reservoirs, effectively enhancing the identification efficiency of fractures in shale gas reservoirs.
By random sampling, compressive sensing seismic acquisition can significantly improve seismic acquisition efficiency and reduce seismic acquisition costs. Therefore, a design method and workflow are proposed for an offshore compressive sensing seismic acquisition geometry. Firstly, in response to the alias and difficulty in anti-alias reconstruction in the evaluation of offshore compressive sensing seismic acquisition geometry and field data processing, a frequency-wave number domain weighted anti-alias data reconstruction method is put forward. Then, the developed efficient offshore compressive sensing seismic acquisition technology and data reconstruction method are applied to a real survey of a volcanic buried hill in the South China Sea. A wide azimuth ocean bottom nodes compressive sensing acquisition geometry is also designed. The quality, feasibility, cost reduction and efficiency improvement of the compressive sensing acquisition geometry are evaluated by the methods including data reconstruction. The designed compressive sensing acquisition scheme is successfully implemented in the real survey in the South China Sea, obtaining the first independently designed and implemented offshore compressive sensing acquisition field data in China. The application results show that, under the premise of reducing costs and increasing efficiency by 25%, the compressive sensing acquisition in this survey achieves comparable results to the conventional acquisition, providing effective support for the subsequent research and promotion of efficient offshore compressive sensing acquisition technology.
The Kuqa foreland thrust belt is characterized by highly variable topography, complex near-surface structures, diverse surface lithologies, and intricate subsurface geometries. These factors collectively contribute to seismic data with low signal-to-noise ratio and poor imaging quality. To improve the quality of seismic acquisition data in this area, an optimization design method of the 3D seismic geometry is proposed for a complex area in the Kuqa foreland thrust belt. Firstly, a 3D geophysical model is constructed based on seismic, geological, and well-logging data from the region, followed by 3D wave-equation forward modeling and 3D seismic illumination. Secondly, the influence of geometry parameters on migration imaging and target-layer illumination energy is analyzed both quantitatively and qualitatively to identify sensitive parameters and optimization directions.The suggestions for geometry optimization in this area are also proposed.The study concludes that shot-receiver density is a key factor affecting structural imaging and illumination. A density higher than 1.48 million traces per km² is required in this area. Under a given shot-receiver density, a geometry with smaller receiver line spacing is more conducive to migration imaging. When receiver line spacing is less than 180 m, imaging accuracy for complex structures in the foreland thrust belt can be improved. Additionally, a wide-azimuth geometry is more favorable for target-layer illumination and thin-sand identification, with a recommended aspect ratio greater than 0.8. The findings provide valuable references for the geometry design in other similar areas.
Channel identification is crucial for predicting fluvial facies reservoirs. However, when the P-wave impedance contrast between channel sandstones and surrounding rocks is minimal, it is difficult to use only post-stack P-wave seismic data for channel identification. S-wave data can effectively enhance the reliability of predicting the spatial distribution of channels. However, the combined identification process of P-wave and S-wave involves challenges such as difficult parameter selection, high subjectivity, and extended working cycles, leading to inefficiencies and reduced reliability. This paper proposes an automatic channel identification methodbased on the joint P-wave and S-wave seismic data. First, to address the issue of insufficient sample data, it puts forward a method for automatically generating synthetic forward modeling samples of 3D channel geological models based on actual data interpretation and channel interpretation results, effectively expanding the sample data set. Subsequently, a new 3D automatic channel identification network structure is then designed, which effectively integrates P-wave and S-wave seismic data, enhancing the reliability of the identification results. Finally, the proposed method is applied to identify tight gas channel sandstones in a work area in southwestern China. Compared with traditional seismic attribute analysis and intelligent identification results relying on a single data type, the proposed method demonstrates higher efficiency and reliability, validating its applicability.
As a bridge linking surface seismic data and borehole information, Walkaway VSP technology has unique advantages in the exploration of structurally complex hydrocarbon reservoirs. Using actual Walkaway VSP data from a complex slope zone in the X Basin as an example, this paper systematically presents a complete technical workflow, covering raw data quality control, amplitude- and vector-preserving processing, and imaging. To address the challenges of strong topographic relief, complex wavefields, and severe signal loss during data acquisition in this area, a multi-stage, coupled amplitude-preserving strategy is innovatively proposed. The workflow integrates five techniques, including RT rotation correction, abnormal amplitude suppression, spherical divergence compensation, consistent amplitude compensation, and amplitude balancing, which effectively resolve the problem of uneven wavefield energy distribution. Depth-domain velocity modeling and imaging of Walkaway VSP data are carried out using tomography and prestack Kirchhoff depth migration. Processing results from real data demonstrate that the optimized preprocessing workflow significantly enhances deep effective signal energy and improves the signal-to-noise ratio, providing a high-quality data foundation for subsequent prestack depth migration imaging. The amplitude- and vector-preserving preprocessing and depth-domain migration imaging workflow proposed in this study offers a reference for Walkaway VSP data processing in complex geological areas.
Dispersion curve extraction is one of the key steps in microtremor investigation, and its accuracy directly affects the reliability of subsequent inversion results. Automatic picking of dispersion curves meets the requirements of large-scale data processing and therefore is of great significance for microtremor investigation. However, because microtremor is a passive-source signal, the energy in its dispersion spectrum is typically weak, which makes automatic picking challenging. To address this issue, this paper proposes an automatic extraction method for microtremor dispersion curve based on the maximum between-class variance method. The proposed approach first applies mean filtering to smooth the dispersion spectrum. Then, the dispersion curve region is determined from the smoothed dispersion spectrum through two successive applications of the maximum between-class variance method. Noise is further removed using a maximum connected-domain identification method. Finally, the algorithm automatically extracts the dispersion curve by selecting the energy peak velocities corresponding to all frequencies within the largest connected-domain. This method effectively removes minor energy interference in the dispersion spectrum and accurately locates and extracts the dispersion curve. Experiments on synthetic data show that the extracted dispersion curves are in good agreement with the theoretical dispersion curves, which demonstrates the accuracy of the algorithm. Experiments on real data, through 30 independent inversions and comparison with geological information, further verify the reliability of the method in practical applications.
Based on acoustic wave theory and petrophysical theory, the observed variations in the seismic wave amplitude and phase typically arise from the coupled effects of intrinsic attenuation associated with rock properties and thin-layer interference related to formation structures. Conventional Q value estimation methods generally overlook this coupling phenomenon, resulting in estimated Q values that represent the combined response of seismic apparent attenuation caused by intrinsic attenuation and thin-layer interference. Meanwhile, inaccurate Q values compromise the sparsity of inversion results for nonstationary reflection coefficients. To decouple the effects of formation intrinsic attenuation and seismic apparent attenuation and thus enhance seismic resolution, this study adopts a Q scanning method, in which an L2, p group sparse norm is introduced to quantify the relationship between reflection coefficient amplitude and intrinsic attenuation. Intrinsic attenuation and apparent attenuation are decoupled by measuring the sparsity of inversion results obtained with different Q values. Additionally, the accuracy and stability of Q value estimation are improved by multichannel sparse reflection coefficient inversion, thereby developing a multichannel Q value estimation technology based on the L2, p norm criterion. Finally, the effectiveness of the proposed method is validated by numerical experiments and actual data. The results demonstrate that this method can accurately estimate the interval Q field varying with the space and time while obtaining seismic data after resolution enhancement.
Seismic data generally lacks low-frequency information in a range of 0-5 Hz, so how to effectively compensate for such information constitutes a critical step in improving the objective accuracy of seismic inversion. To this end, this paper proposes a digital filtering-based impedance inversion method with controllable low-frequency constraints. First, an inversion objective function is established to match synthetic seismogram derived from inversion results with seismic data and impose regularization constraints on reflection coefficients. Second, digital filtering is integrated into the objective function. During inversion iterations, low-pass filters are separately applied to the broadband geological model and inversion results by digital filter, so as to force the low-frequency information of inversion results to conform to that of the geological model.Therefore, the model-driven constraints on inversion results are achieved. Both model test and field application demonstrate that, compared with the existing impedance inversion methods, the proposed method only uses the low-frequency information of the geological model to constrain the low-frequency information of inversion results without distorting the mid-frequency and high-frequency information. Moreover, by regulating the frequency components of low-frequency information via a digital filter, this method offers a novel low-frequency constraint strategy for seismic inversion, which effectively enhances the application flexibility of geological models and improves the resolution and accuracy of impedance inversion results.
Pure qP-wave equations in transversely isotropic (TI) media involve complex spatial-wavenumber coupling terms, but solving them by adopting low-rank approximation and pseudo-spectral methods comes at a high numerical simulation cost. Scalar operators are introduced to normalize the wavenumber vector into a unit vector aligned with the wavefield propagation direction, and the asymptotic approximation finite-difference method is combined for solution, which can transform the spatial-wavenumber coupling terms into the spatial domain for processing, thereby improving numerical simulation efficiency. However, serving as a rough estimate of the Poynting vector, asymptotic approximation yields direction vectors with alternating orientations and singularities. Additionally, the Poynting vector has obvious null points, which will affect the stability of wavefield extrapolation processes. Especially for some media with partial strong anisotropic characteristics, the results of asymptotic approximation wavefield simulation will produce numerical noise and amplitude instability. To this end, this paper adopts the stabilized Poynting vector approximation and optical flow vector approximation to replace the traditional asymptotic approximation for obtaining more stable and accurate wavefield directions. Meanwhile, scalar operators are combined to solve the pure qP-wave equation in TI media. Numerical experiments demonstrate that in some media with strong anisotropic characteristics, the method based on stable direction vectors can better suppress numerical noise in the wavefield, improve simulation accuracy, and yield higher-resolution reverse time migration (RTM) images. This study expands the application scope of this efficient finite-difference wavefield simulation method in strongly anisotropic media.
Least-squares migration (LSM) is an extension of classical migration methods within an inversion framework. Its core principle is to achieve high-resolution imaging of subsurface media by compensating for the effects of the Hessian matrix. However, the large size and sparse nature of the full Hessian matrix make it difficult to compute and characterize effectively, which constitutes an obstacle to the practical application of LSM. This paper first introduces an analytical representation of the Hessian matrix under the Born approximation. Given the sparsity of the Hessian matrix, an efficient approximation is performed using point spread functions (PSFs), and it is emphasized that traveltime calculation is key to constructing the Hessian matrix. Subsequently, a ray-beam-based traveltime calculation method is introduced to complete ray-beam-based PSF calculations. Finally, the calculated PSFs are applied to the conventional migration results via the data-driven, high-dimensional image-domain deconvolution, thereby realizing image-domain LSM. Ray-beam-based PSF calculation can effectively avoid the traveltime interpolation issues associated with central-ray methods, and its calculational efficiency is significantly higher than that of PSF calculation methods based on forward modeling and migration. Numerical experiments and field data processing results validate the effectiveness of the proposed method.
Array acoustic logging is an important method for acquiring information such as dispersion, absorption attenuation, and elastic parameters of subsurface reservoirs. To accurately simulate the propagation of the array acoustic logging wavefield in the borehole environment, it is usually necessary to solve the wave equation in cylindrical coordinates. However, different coordinate systems show significant differences in the representation of the spatial distribution and parameter characteristics of the exploration targets. In practical applications, it is necessary to select an appropriate coordinate system according to the characteristics of specific geological targets. Therefore, an adaptability study of rectangular and cylindrical coordinate systems in the dispersion response simulation of array acoustic logging is carried out. First, the problems existing in the two coordinate systems are theoretically analyzed. Then, the validity and accuracy of the algorithms are verified based on axisymmetric models, and the applicability of the two coordinate systems is analyzed using fracture elastic models and poroelastic models, respectively. Simulation results show that there is no obvious difference in waveforms between the two simulation methods, but the quality of dispersion curves obtained using the cylindrical coordinate system is significantly higher than that obtained using the rectangular coordinate system. Therefore, in practical simulations, the 2.5-dimensional cylindrical coordinate algorithm is the most economical and efficient for axisymmetric models. For non-axisymmetric near-well models, three-dimensional cylindrical coordinates provide higher accuracy. For non-axisymmetric far-well models, three-dimensional rectangular coordinates provide higher efficiency. This study provides an important reference for coordinate system selection in numerical simulation of array acoustic logging.
As overpressure in continental shale formations in the Fuxing area of Sichuan Basin seriously affects well control safety, timely and accurate prediction of formation overpressure can effectively prevent the occurrence of accidents such as well kicks and blowouts, contributing to safe and efficient drilling operations. To this end, this paper adopts methods including combined logging curve analysis, loading-unloading curves, and acoustic velocity-density cross plots to analyze the overpressure genesis of continental shale in the study area. In view of the genesis mechanism characteristics of continental shale overpressure, this paper proposes a new pore pressure prediction method that comprehensively considers the composite pressurization mechanism of hydrocarbon generation expansion and undercompaction. The research results show that the gas logging of multiple wells in the study area shows abnormal values, and two wells have repeated abnormal gas logging values and well kick even after increasing the density of drilling fluid, which can effectively determine the abnormal increase of pore pressure in the formation. By integrating three kinds of overpressure genesis mechanism, the overpressure of the Lianggaoshan Formation in the study area is attributed to undercompaction, and the overpressure of the Dongyuemiao Formation is due to the composite effect of hydrocarbon generation expansion and undercompaction. The relative error of formation pressure predicted by the new method is only 4.6%, with strong applicability and high prediction accuracy. The research results are of significance for the efficient exploration and development of continental shale oil and gas in the study area.
Located in the central Junggar Basin, the eastern ring of the West Sag of Well Pen-1 is a significant target for oil and gas exploration in China. However, controversies persist in terms of the sequence stratigraphic framework division of the Lower Jurassic Sangonghe Formation. Based on logging and core data, this paper adopts analytic methods including maximum entropy spectral analysis, prediction error filter analysis (PEFA), integrated prediction error filter analysis (INPEFA), cosine of instantaneous phase, and wavelet time-frequency analysis to process lithology-sensitive logging curves. As a result, curves such as the frequency trend, cosine of instantaneous phase, and wavelet coefficients at different scales as well as time-frequency chromatograms are obtained. Sequence boundaries of different orders are identified by analyzing trend inflection points of varying amplitudes, oscillation characteristics of the obtained curves, and energy variations in the time-frequency chromatograms. Single-well stratigraphic division is conducted on 53 drilled wells penetrating the Sangonghe Formation in the study area, and 28 of them are selected for cross-well sequence framework correlation analysis. Additionally, the sequence stratigraphic framework and sedimentary distribution characteristics of the second member of Sangonghe Formation (J1s2) and the third member of Sangonghe Formation (J1s3) are studied. The research results indicate that J1s2 and J1s3 in the study area collectively constitute a third-order sequence, internally developing three system tracts of the lowstand system tract (LST), transgressive system tract (TST), and highstand system tract (HST). Furthermore, J1s2 can be divided into four fourth-order sequences. The depocenter of the study area exhibits migratory patterns, primarily distributed in the northwestern and northeastern parts, with the most significant depocenters located near the Z11 and QS4 wells. Finally, a high-precision sequence stratigraphic correlation framework for J1s2 and J1s3 is established, various sequences and sedimentary boundaries are identified, and their spatial sedimentary characteristics are clarified, thus laying a foundation for the detailed oil and gas exploration in the study area.
CCS/CCUS technologies are essential for achieving zero and even negative carbon emissions, and fault sealing is a critical factor determining the effectiveness and safety of CO2 storage. Predominantly relying on the shale gouge ratio (SGR) on both sides of a fault for evaluating fault sealing, traditional methods are highly dependent on well data and feature oversimplified evaluation parameters, which results in low evaluation accuracy in areas with limited CCS/CCUS exploration data and thus fails to meet practical requirements. The method can be optimized from the following aspects to solve the above-mentioned problems. ①High-precision 3D modeling integrated with well and seismic data is carried out to characterize the heterogeneity of fault zones. As a result, this can solve the problem that traditional methods are difficult to apply to areas with few or no wells, and establish a framework for 3D sealing evaluation, improving the vertical evaluation accuracy from the 10-meter scale to the meter scale. ②A method for fault permeability (Kf) calculation is improved, and a model for fault transmissibility (TM) evaluation is introduced, thereby eliminating the influence of maximum burial depths and burial depths at the time of formation, and advancing the transformation from 2D fault planes to 3D fault bodies. Field application results show that compared with traditional methods, the proposed evaluation method improves the average accuracy of shale thickness by 20%. It has supported the implementation of a 127.5 km2 CCS storage geological body in Well SQ3 of the Zhundong area, with the key factors causing CO2 gas channeling in the H pilot test area identified. Additionally, the early warning accuracy is over 90%, which helps reduce the incidence of gas channeling and provides strong technical support for CCS/CCUS safety.
In the domains of geology and mineralogy, the precise identification of rock thin sections holds paramount significance for understanding the composition, structure, and formation process of rocks. However, traditional manual identification methods are cumbersome, time-consuming, highly subjective, and heavily dependent on experience. Deep learning enables rapid and accurate identification of rock thin sections. This study proposes a WTConvNeXt-Inception network for intelligent lithology identification, which significantly improves classification accuracy through multimodal feature fusion of plane-polarized and cross-polarized images. Aiming at the insufficiency of feature extraction in traditional methods when handling complex rock images, wavelet convolution is introduced to effectively capture the multi-scale features in the images. To address the issue of high memory consumption of the model, the Inception module is adopted to improve the operational speed and efficiency of the model. To tackle the problem of missing information in single images, a cross-attention fusion module is proposed, effectively exploiting the information between plane-polarized light and cross-polarized light images. Experimental results demonstrate that the proposed method achieves a classification accuracy, F1-score, and quadratic weighted Kappa values of 98.64%, 98.64%, and 98.30%, respectively, verifying its effectiveness in highly accurate and efficient identification of rock thin sections. This method shows strong potential for applications in geological research, petroleum exploration, and other related fields.
There are such problems as low recognition accuracy and insufficient reliability for existing fault identification methods in seismic data, and they cannot satisfy the demands of high-precision exploration. To explore the three-dimensional spatial structural information of seismic data, enhance the clarity of fuzzy and weak faults, and improve fault continuity, a fault identification method based on tensor sparse optimization analysis is proposed. First, based on the three-dimensional spatial distribution characteristics of fault seismic responses, tensor decomposition analysis is performed, combined with compressive sensing theory and matrix low-rank sparsity theory, to analyze the low-rank sparse decomposition characteristics of fault information, background information, and noise information. Second, vector sparse representation is combined with matrix sparse representation, and tensor decomposition theory is applied to achieve tensor dimensionality reduction, matricization, and vectorization. Finally, sparse wavelet decomposition orthogonal matching pursuit (OMP) reconstruction is used for vector optimization, and the matrix low-rank sparse method is used for matrix optimization, achieving noise removal and fault enhancement. Model tests and practical applications show that the proposed method has strong noise resistance and high identification accuracy, and demonstrates significant effectiveness in weak fault identification and fault continuity enhancement. This method has good reference significance for fault-developed areas.
Shear-wave velocity (VS) plays a crucial role in the prediction of clastic rock reservoirs, including the identification of "sweet spots". In the Shahezi Formation of the Xujiaweizi BSH9 well area, the differences in P-wave impedance (IP) among tight sandy conglomerate, tight siltstone, and mudstone are negligible, and the VP/VS exhibits substantial overlap. Rock physics analysis identifies VS as the most sensitive elastic parameter in the SQ4 member of this area. The relatively low accuracy of previous reservoir predictions leads to a drilling success rate below 80% for sandy conglomerate reservoirs in the SQ4 member, as well as poor prediction accuracy for high-porosity and high-permeability "sweet spots". This is a key reason for the low productivity of earlier wells such as BSH1 and BSH5. Based on single-interface model data at the top of the reservoir and well-log data (including VS) from the target interval, AVO forward modeling is performed to generate common reflection point (CRP) gathers. Two pre-stack inversion methods, full-offset-gather pre-stack inversion and partial-stack pre-stack inversion, are applied to these gathers to obtain relative VS values. Compared with the second method, the first method reduces the inversion error from 25% to 10%. Moreover, the relative VS values outperform the relative P-wave impedance in distinguishing lithology and identifying sweet-spot reservoirs. After residual multiple attenuation is applied to the actual quasi-pure wave gathers, the direct VS inversion results from full-offset gathers in the target interval (SQ4 member of the Shahezi Formation) show good consistency with the drilling results of Well BSH9H. Three subsequent wells drilled in predicted high-quality reservoir zones yield high industrial gas flow exceeding 1 million cubic meters per day, whereas one well located in a predicted low-quality zone produces only low gas flow.
Carbon dioxide enhanced oil recovery (CO2-EOR) technology is a crucial means of enhancing oil recovery and storing carbon, and its dynamic monitoring accuracy directly influences project performance. Conventional post-stack seismic data fail to preserve anisotropy information, which makes them unable to determine the main flow direction and dominant injection-production channels. To address this, using the beach-bar sand reservoir in the G89-1 well area of Shengli Oilfield as the research object, this paper proposes a CO2 monitoring approach based on pre-stack azimuthal information by utilizing the azimuthal data from the original offset-vector tile (OVT) gathers. By analyzing the relationship among azimuthal angles, offsets, and fold numbers, azimuthally stacked gathers suitable for the study area were generated. Meanwhile, by performing amplitude consistency processing on the top of the gas injection layer, the seismic anisotropic amplitude characteristics for CO2 monitoring are obtained. Well analysis results indicate that the extreme values of amplitudes in the azimuthal gathers are highly consistent with the main flow direction of CO2, providing a basis for predicting dominant injection-production channels. By conducting amplitude consistency processing on the seismic data of the entire area and extracting the maximum amplitude attribute, the sweep range of CO2 in the gas injection layer was predicted and four favorable distribution zones for remaining oil are evaluated, with the dominant directions for subsequent exploration and development identified. This study proposes a CO2 monitoring method based on OVT azimuthal information, offering a new technical approach for CO2 monitoring.
With the continuous advancement of unconventional oil and gas exploration and development in China, geosteering modeling technology for horizontal wells has become a critical method to ensure reservoir drilling rates. While seismic data provides essential parameters for predicting lateral stratigraphic variation trends, factors such as complex stratigraphic undulations and discrepancies between well and seismic scales severely limit the accuracy and efficiency of geosteering. Therefore, an integrated key well-seismic matching technology is proposed to enhance the precision and automation of geosteering modeling for horizontal wells. First, by an automated cross-section extraction technique based on the well trajectory-seismic grid relationship, the difficulty in obtaining cross-sections in complex geological conditions is resolved, thereby increasing modeling automation. Second, a trajectory-based cross-section reconstruction technique using neighborhood-weighted interpolation effectively overcomes reconstruction difficulties caused by stratigraphic undulations and scale differences, further improving modeling accuracy. Finally, bed thickness correction technology integrating well trajectory and seismic structural relationships addresses bed thickness calibration issues, significantly improving the accuracy of bed thickness estimation. According to practical application within a typical structurally complex block, this technology markedly enhances modeling automation and bed thickness estimation precision, providing effective technical support for geosteering modeling for horizontal wells.
The key of active geosteering is to obtain the pre-drilling formation information in realtime. Based on the field diffusion theory, transient electromagnetic (TEM) detection has advantages such as high detection efficiency, strong deep detection ability, and small influence from the spacing, and can effectively solve the limitations of frequency-domain ultra-deep look-ahead electromagnetic logging while-drilling. Thus, the TEM detection method is applied to downhole ultra-deep look-ahead detection. The generalized reflection-transmission theory and Talbot frequency-time conversion algorithm are adopted to efficiently simulate the TEM look-ahead logging responses. Meanwhile, the quantitative calculation of key pre-drilling logging parameters is completed via combining the Levenberg-Marquardt (L-M) inversion and all-time apparent resistivity algorithm. The results demonstrate that TEM logging can detect multi-boundary formation information of tens of meters ahead of the drill bit by utilizing short spacing under arbitrary well deviation. The pre-drilling boundary response characteristics are primarily related to the formation resistivity, which influences eddy current diffusion ahead of boundaries and the distance to boundary (DTB). The higher formation resistivity and shorter DTB mean earlier boundary response time. Although the apparent resistivity of logging can qualitatively determine pre-drilling formation resistivity distribution, the real-time and accurate acquisition of boundary distance and formation resistivity still necessitates refined logging inversion. The research on forward modeling and inversion of TEM logging for ultra-deep look-ahead detection validates the effectiveness of the technology in acquiring pre-drilling geological information, thereby providing theoretical guidance for optimizing logging tool design and realizing the applications of active geosteering.
Human noise (especially power-frequency interference) and random noise collectively constitute the primary interference sources of electromagnetic exploration signals. In severe cases, the noise may even obscure the electromagnetic response of geological structures. Therefore, the effective suppression or elimination of such noise is a pivotal step in electromagnetic data processing. This paper proposes a denoising method combining variational mode decomposition (VMD) and dynamic spectral kurtosis (SK) based on Bayesian optimization for borehole-acquired electromagnetic field data. Firstly, an adaptive VMD parameterization mechanism is constructed.The penalty factor α, mode number K, and noise tolerance τ are determined via Pareto-Bayesian joint optimization to achievethe efficient decomposition of borehole electromagnetic signals. Secondly, a dual-stage screening strategy coupling frequency-domain coherence with time-frequency SK is designed to suppress noise while preserving geological attenuation characteristics. The results of simulation and field data calculations demonstrate that compared with wavelet-based methods and other traditional approaches, the signals processed by the proposed method exhibit excellent performance within a wide noise range from -10 to 30 dB, with its maximum signal-to-noise ratio (SNR) improvement of 22.07 dB.The proposed method can effectively suppress power-frequency harmonics and random noise in data collected in practical scenarios, mitigating Gibbs oscillations induced by traditional notch filtering methods. It not only retains the integrity of geological response characteristics but also features strong robustness and high precision, thereby providing high-fidelity inputs for subsequent time-depth conversion and geological modeling.
Vibroseis in Sichuan seismic exploration faces challenges such as poor vibration energy output and significant signal distortion when operating on rural cement roads. To date, studies have yet to systematically investigate the influence of road conditions and surface layers on vibroseis excitation performance. Therefore, a vibrator plate-road-ground coupling excitation performance enhancement method is proposed. First, a longitudinal-wave vibroseis vibrator plate-road-ground coupling dynamic model considering the impact of rural cement roads is established, with its accuracy confirmed by experimental comparisons with existing models. Second, from an energy transmission perspective, the effect of hammer mass, plate mass, and plate area on excitation performance is analyzed, revealing that cement roads significantly attenuate energy transmission, particularly in the mid-to-high frequency range. Third, based on key influencing factors, an adaptive vibrator plate for cement roads is innovatively designed, improving excitation performance by 16.25%. Experimental results demonstrate that hammer mass and plate mass are critical factors for vibroseis excitation performance. Higher hammer mass and lower plate mass enhance excitation performance across both low- and mid-to-high frequency bands. The proposed method provides theoretical and practical guidance for vibroseis applications in complex terrains.
Ultrasonic Lamb wave logging plays a significant role in evaluating cementing quality and detecting casing damage. This paper takes China's first scientific exploration well with a depth exceeding 10, 000 meters as its research subject. The effects of casing thickness, mud velocity, excitation frequency, and other factors on the quality of ultrasonic Lamb wave logging data are systematically investigated. The results show that the excitation frequencies of ultrasonic resonance waves and flexural Lamb waves gradually decrease as the casing thickness increases. Under conditions of high mud velocity and thick-walled casing, increasing the incidence angle of flexural Lamb wave logging is more favorable for exciting single-mode waves, making the direct waves and reflected waves in the full-waveform easier to distinguish in the time domain. Based on theoretical simulations and experimental tests, the ultrasonic Lamb wave logging parameters suitable for 10000-meter-deep well operations are optimized. Field test results demonstrate that the domestically developed ultrasonic Lamb wave logging tool achieves an average signal-to-noise ratio greater than 30 dB at depths exceeding 10000 meters (measured bottom depth 10718.0 m), providing important technical support for engineering decision-making.
