Enhancing seismic fault extraction through multiple ant tracking on cloud environment

Before this study, simple methods like seismic attribute analysis have often been used for reservoir characterization with successes however, there is still the need to reduce exploration uncertainty to a negligible level and boost investors’ confidence. This study integrated seismic inversion with seismic attribute analysis to better characterize the reservoirs in MTW Field in deep-offshore Niger Delta Basin. Five (5) wells with complete suite of petrophysical logs, three-Dimensional (3-D) seismic data, checkshot and other well information were used. The well data were thoroughly quality checked, reservoirs were litho-stratigraphically delineated and used for petrophysical analysis across the wells. This was followed by seismic-to- well-tie, seismic interpretation, and seismic attribute analysis (Root Mean Square-RMS) generated using depth surface maps. 3-D static reservoir model and volumetric evaluation were carried out. Petrophysical properties were derived and distributed across the 3-D static model using sequential gaussian simulation algorithm to ascertain shale volume spread across the model. To improve the seismic resolution and reduce interpretation uncertainty at greater depth, model- based post-stack seismic inversion was performed to obtain acoustic impedance cube. Litho-stratigraphic and petrophysical analysis result revealed five reservoir sands (A, B, C, D, and E). Reservoir-A was seen to be more viable with a thickness of 13.42 m, high effective porosity of 27%, permeability of 3187.53 mD, low water saturation of 34% and low shale volume of 11% which are indication good reservoir quality and producibility. The seismic interpretation revealed thirty-one (31) growth and antithetic faults oriented in the NE-SW and NW-SE directions, respectively. The RMS result revealed high amplitude reflectivity which is a measure of zone of interest. Based on this, seven (7) prospects and three (3) leads were identified. The seismic inversion result shows a high level of accuracy with a correlation coefficient of 0.997; 0.997; 0.995; and 0.996 in MTW-001, MTW-003ST1, MTW-004ST1 and MTW-005 wells, respectively. The acoustic impedance successfully resolved and improved on the resolution of the seismic stacking velocity especially at reservoir layers and at depth deeper than 3600 ms. Acoustic impedance as a layer property has improved on the lateral and vertical resolutions of the data beyond what the usual seismic interval velocity could image thus, validating some of the prospects and leads identified in this study. This demonstrated that uncertainty can be reduced by a blend of RMS and seismic inversion in identifying reservoir for accurate placement of wells. It is therefore recommended that E and P operators should adopt the technique in their future hydrocarbon exploration endeavor in frontier and matured basins.

Fault interpretation is a key link in seismic data interpretation. With the continuous increase in seismic data volume in original production, manual fault picking cannot efficiently handle massive seismic data. With the development of artificial intelligence technology, the automatic and rapid picking of faults has become a hot topic in the application of deep learning methods in the field of seismic data interpretation. Therefore, this paper proposes a seismic data fault identification method based on SEU-Net. The SEU-Net network introduces SE blocks on the basis of the original U-Net network, enhancing the weights of effective channels through adaptive feature recalibration, and combines a multi-scale feature fusion strategy to improve the model’s ability to identify fault edges and minor faults. The experimental results show that compared with the original U-Net network, the SEU-Net network exhibits higher fault identification accuracy and robustness both in synthetic seismic data and original work area data. This research provides an efficient and automated solution for fault detection in seismic data and it holds certain theoretical value and practical application potential.

The development of seismic technology has made seismic data to be widely used in the interpretation of stratigraphic sequence frames, reservoir identification, fluid detection, and other research fields involved in reservoir description. The 3D technology reservoirs have always been the focus, as well as difficulty, of research. With the rapid development of information technology and the continuous improvement of seismic exploration level, people have put forward higher requirements for the accuracy of seismic data interpretation results. Aiming at the large number of structural and unstructured data in seismic, logging, geology, and other disciplines involved in seismic interpretation, how to effectively organize and coordinate analysis to discover the hidden reservoir structure and oil and gas distribution information has always been a geological and important topic for information processing technicians. This thesis is aimed at the current high-water-phase development of Shengtuo Oilfield reservoir and the problems existing in geological research. Based on seismic structural interpretation and attribute analysis, this paper analyzes the reservoir structural characteristics, sedimentary characteristics, and reservoir physical parameter characteristics based on geology, logging interpretation, core analysis, drilling, and seismic interpretation. Using the kriging method with external drift can cooperate with seismic variables to establish a reservoir geological model to study the Shengtuo Oilfield reservoir. We combine artificial intelligence technology with geological modeling technology of seismic interpretation results to explore the best method for predicting earthquakes. The research results in this paper show that the relative error of the model established by the kriging method in the article is relatively small for thinning wells, mainly concentrated around 1%. Examination of the thinning wells of 45 wells shows that the model established is basically good and the example has high accuracy. The research results in this paper have a guiding study of distribution and tapping potentials in the study area, formulating reasonable development and adjustment plans and improving oil recovery.

Summary Three-dimensional (3D) seismic volumes make it possible to map subsurface structural features such as faults and folds, unravel tectonic history, in turn allowing us to predict spatial locations of sweet spots with higher fracture intensity while avoiding the potential migration pathways that decrease efficiency of hydrocarbon production from fractured reservoirs. Seismic attributes, such as variance, curvature, and ant tracking, may provide valuable insight into understanding the complex fault and fracture networks and overprinting relationships associated with the polyhistory of the north-central Appalachian Basin. Highquality 3D seismic data in Central Pennsylvania show vertical changes from rift-sag basin associated with the early Paleozoic extension along the Rome trough. This basin style changes upward to foreland basin dominated by regional folds and thrusts above the major detachment interval of the Salina Salt and Devonian shale. In addition to the regional detachment folds and thrusts we mapped cross-strike, northwest-trending lineaments, which may be co-eval to the previously-reported cross-strike discontinuities (CSD) that were found in other areas in the basin using different data sets. Our initial analyses suggest that these cross-strike lineaments may pose threat to hydrocarbon productivity and recovery efficiency due to faults and fractures acting as paths of least resistance for induced hydraulic stimulation fluids. We interpret that these seismically observed lineaments may represent major transfer faults that serve as the migration pathways for hydraulic fluid migration and affect stimulations by potentially absorbing and/or redirecting the energy. Detection and evaluation of fracture orientation and intensity and emphasis on the relationship between fractures and fluid migration pathways is of high interest in the study area as it entails significant time and cost implications for both conventional and unconventional hydrocarbon exploration and production.

Summary Deep learning has showed success in various industries for different processing tasks. However, it is still in the initial stage of the application of seismic exploration. In this paper, we propose to apply deep learning technology to detect seismic faults. The residual neural network (Resnet) is used to build the U-net model for seismic fault detection. And multi-scale and multi-level feature are extracted from seismic data. In the training stage, we use field dataset to train the U_net model. Finally, field data experiment demonstrates that our proposed U-net model is feasible for seismic fault detection. In the meantime, our research will help promote the intelligent development of seismic faults interpretation and provides a reference for further studies of seismic data processing and interpretation.

Seismic attributes analysis as a precursor for hydrocarbon indication in onshore Fuba Field, Niger Delta, Nigeria using Well-logs and 3D Seismic data are here presented. Well-to-seismic ties, faults and horizon mapping, time-surface generation, depth conversion and seismic attributes generation were carried out using Petrel software. The structural interpretation of seismic data reveal highly synthetic and antithetic faults which are in line with faults trends identified in the Niger Delta. Of the 36 interpreted faults, only synthetic and antithetic faults are regional, running from the top to bottom across the field. These faults play significant roles in trap formation at the upper, middle and lower sections of the field. Three distinct horizons were mapped. Reservoir M is found at a shallower depth from 10937 to 10997 ft, reservoir N is found at a depth ranging from 11213 to 11241 ft while reservoir O is found at a deeper depth ranging from 11681 to 11871 ft respectively. Analysis of the relevant seismic attributes such as variance edge, root mean square, maximum amplitude and relative acoustic impedance were applied to the seismic data. The variance values ranges from 0.0 to 1.0. The Variance edge analysis was used to delineate the prominent and subtle faults in the area. The RMS amplitude values ranges from 9,000 to 13,000 in the reservoirs. The root mean square amplitude and maximum amplitude analysis reveal bright spot anomaly. These amplitude anomalies served as direct hydrocarbon indicators (DHIs), unravelling the presence and possible hydrocarbon prospective zones. The relative acoustic impedance analysis was used for delineating lithology variation in the seismic sections. Results from this study have shown that away from the currently producing zone of the field, additional leads and prospects exist, which could be further evaluated for hydrocarbon production.

Seismic is one of the main methods used for the identification of structures and stratigraphic studies in sedimentary basins. In the Santos Basin, numerous 2D and 3D seismic surveys are being conducted in order to get a better ima-ge of the geological section to depths beyond the base of salt layer. Crustal modelling studies that make joint interpretation of seismic and gravity data are found in the literature, however there are few studies that relate gravity anomalies directly to salt structures. This work aims to associate gravity anomalies with salt structures from seismic and gravimetric interpretation. For studies aimed to model the crustal structure from gravity field data, the knowledge of two major discontinuities is required, the basement and the Moho. Such interfaces are often not easily seen by seismic and so, during this study, they were found by different methods involving analysis of gravity anomalies. The other interfaces involving density contrasts were analyzed based on seismic interpretation. The results showed that the obtained seismic geological interpretations may provide additional information when compared to gravity anomaly data. In all the modelled profiles, some geological information of the Santos Basin that are not visible in the seismic, could be interpreted according to the geological model and the adjustment of gravity anomaly curves. As a final conclusion of this work, it is suggested that the combined analysis of the two cited geophysical methods, can provide important information about the crustal structure and to assist in modelling the salt layer.

Seismic is one of the main methods used for the identification of structures and stratigraphic studies in sedimentary basins. In the Santos Basin, numerous 2D and 3D seismic surveys are being conducted in order to get a better ima-ge of the geological section to depths beyond the base of salt layer. Crustal modelling studies that make joint interpretation of seismic and gravity data are found in the literature, however there are few studies that relate gravity anomalies directly to salt structures. This work aims to associate gravity anomalies with salt structures from seismic and gravimetric interpretation. For studies aimed to model the crustal structure from gravity field data, the knowledge of two major discontinuities is required, the basement and the Moho. Such interfaces are often not easily seen by seismic and so, during this study, they were found by different methods involving analysis of gravity anomalies. The other interfaces involving density contrasts were analyzed based on seismic interpretation. The results showed that the obtained seismic geological interpretations may provide additional information when compared to gravity anomaly data. In all the modelled profiles, some geological information of the Santos Basin that are not visible in the seismic, could be interpreted according to the geological model and the adjustment of gravity anomaly curves. As a final conclusion of this work, it is suggested that the combined analysis of the two cited geophysical methods, can provide important information about the crustal structure and to assist in modelling the salt layer.

Chapter 13 - Understanding Seismic Interpretation Methodology

Interpretation of geologic structures entails ambiguity and uncertainties. It usually requires interpreter judgment and is time consuming. Deep exploitation of resources challenges the accuracy and efficiency of geologic structure interpretation. The application of machine-learning algorithms to seismic interpretation can effectively solve these problems. We analyzed the theory and applicability of five machine-learning algorithms. Seismic forward modeling is a key connection between the model and seismic response, and it can obtain seismic data of known geologic structures. Based on the modeling data, we first optimized the seismic attributes sensitive to the target geologic structure and then we verified the accuracy of the five machine-learning algorithms by the cross-checking method. In this case, the random forest algorithm had the highest accuracy. So we examined the structural interpretation method based on a random forest using the 3D seismic reflection data from coalfield exploration. The prediction effect of this interpretation workflow is verified by comparison with known geologic structures on the plane and profile. The results suggest that the random forest algorithm is feasible to indicate geologic structure interpretations in the case of collapsed column and fault structures and it can effectively improve the efficiency of seismic interpretation and its accuracy. The machine-learning-based workflow provides a new technique for seismic structure interpretation in coal mining.

Fault detection is an essential component of seismic interpretation and plays a crucial role in industrial processes. However, it is also one of the main challenges, especially in delineating faults in 3D seismic data. Recently, the rapidly developing technology, deep learning, has proven to be a powerful tool for this task. A number of neural networks have been proposed for this purpose by regarding 3D fault detection as a semantic segmentation task. To further enhance the effectiveness of the deep learning methods, we propose a novel network architecture, named R2SE-Unet, to solve the 3D segmentation problem. In the neural network, we design a recurrent residual-SE convolution unit (RRCU-SE) that integrates the residual learning and Squeeze-Excitation module to store information in 3D seismic data. This component promotes the spread of 3D volumetric information and aids in learning spatial dependencies in 3D images. In addition, to reduce the impact of insufficient spatial resolution resulting from the base architecture of U-net, we add an attention unit between skip connection operations. These two new units enable our R2SE-Unet to exploit semantic information more accurately in the feature maps. After many experiments on region-based loss functions and distribution-based loss functions, we also propose a novel loss function, which takes the advantage of generalized dice (GDice) loss and balanced binary cross entropy (b-BCE) loss, named Gdice-bce, to effectively train R2SE-Unet. Although only synthetic seismic data samples are used to train the network parameters, our R2SE-Unet could produce more reliable fault feature maps on field seismic data than two other conventional fault detection neural networks. Thus, the proposed neural network is easy to train and reliably works for seismic fault interpretation on field seismic data.

Deviation from exploration for conventional, Albian age, clastic plays towards carbonate plays of a stratigraphic nature has prompted the use of plan view - horizontal seismic displays in the interpretation process as an additional means of data visualization enhancement. Significant improvement in accuracy and speed have been attained with the use of horizontal displays in the mapping of minor lineaments and faults with no visible, or minor throw on vertical seismic sections. Mapping practices include work on horizontal slices and slabs, which can be either horizon or time consistent. This approach to seismic interpretation has allowed the inclusion of seismic Attributes, Spectral Decomposition, Edge detection, Coherence, Dip and Azimuth analysis, and other display enhancements into the interpretation process. Carbonate plays of the Lower Cretaceous are of emerging interest as exploration targets. This type of play relies on fracture porosity, which is developed in close association with faults. Generally poor seismic data quality prohibits accurate fault identification on vertical sections. Here it has often been necessary to employ circumspect methods including the use of horizon volumes. These may even be based on vertically offset, but well defined marker horizons that approximate the curvature of the studied event. Methods of this nature have been employed in the identification of stratigraphic features such as the edge of a Lower Cretaceous oolite play proximal to the Minagish field in southern Kuwait. Horizontal displays of seismic attributes were also used in the mapping of the extent of a flat spot in the Upper Cretaceous, and of a very shallow, Lower Miocene channel system in northern Kuwait. In the two latter cases it was extremely difficult to maintain lateral pick consistency among consecutive vertical sections. We conclude that for these types of play, the level of accuracy attained in the interpretation of faults, stratigraphic features, i. e. onlap edges, channels and direct hydrocarbon indicators on horizontal data volumes, widely surpasses what can be reasonably achieved on vertical data displays. We anticipate that interpretation and study of horizontal seismic attribute and other signal enhancement displays will play an increasingly important role in the future subsurface mapping in Kuwait. Introduction Large amounts of information embedded in seismic trace data are not always readily visible on traditional, vertical amplitude sections. The underlying problem is the level of resolution apparent in the seismic data. Whether this is due to low consistency of detail or high levels of cluttering varies among data sets. Sensitivity to detail and demands of the task and quality of data at hand will determine the level of quality standards imposed on acquisition and processing parameters to be used. Regardless of advances in data presentation and post-processing image enhancement, it is the underlying data quality that will ultimately determine the results of any interpretation. The data quality aspect extends from acquisition of raw data via signal processing to final display along with its various refinements. Early mistakes or bad practices can only rarely be repaired and never recovered in full. 3D seismic acquisition foot-prints can be softened at the processing stage, but not fully eradicated without compromising data content. This has particular relevance to fault resolution. Residual foot-prints are manifested as striations in the data that will interfere with real events of potential importance.

In some oilfield where 3D seismic survey has been done, the deeper structure cannot be discovered for poor deep seismic data. Layer stripping technique applied both seismic and gravity data is a solution, but it cannot get satisfied processing effect because the horizontal variation of formation density is ignored. To improve the gravity survey effect in 3D seismic survey area, this paper presents the formation separation technique. Based on 3D seismic depth data and the transformed density data from 3D seismic velocity data, the upper formation's gravity effects are calculated and removed from Bouguer gravity and then the formation‐separated gravity‐anomaly was obtained, which mainly reflects the deeper geological structure. In block XX, the seismic data of shallow formations are excellent but that of top of basement are poor. The formation‐separated gravity‐anomaly processed under the control of 3D seismic data fits well with the known seismic interpretation and wells. It makes the geological interpretation more reliable

Thin (<20 m), sand injectites, too small to resolve in conventional seismic reflection data, can constitute a large part of the net-to-gross volume and affect fluid flow in the reservoir. However, they may also cause challenges for well placement and reservoir development because they are too small to be reliably imaged in seismic reflection data. It is therefore important to better understand how small-scale injectites influence seismic images, and how they may be recognized and characterized above reservoirs. The Grane Field (North Sea) hosts numerous small-scale sand injectites above the main reservoir unit, challengin well placement, volume estimates and seismic interpretation. Here, we investigate how such small-scale sand injectites influence seismic images and may be characterized by (1) using well, 3D seismic and outcrop data to investigate geometries of small-scale sand injectites (0–15 m thick), (2) performing seismic convolution modelling to investigate how these would be imaged in seismic data, and (3) comparing these synthetic seismic images with actual 3D seismic data from the well-investigated Grane Field. Our results show that despite injectites being below seismic resolution, small-scale sand injectites can be detected in seismic data. They are more likely to be detected if they have high thickness (>5 m), steep dip (>30°), and when densely spaced. Furthermore, as the fraction of sand injectites increases the top reservoir amplitude will decrease. Moreover, comparison of the synthetic seismic images with real seismic data from the Grane Field indicates that the low-amplitude anomalies and irregularities observed above the reservoir may be a result of the overlying sand injectites. Additionally, the comparison strongly suggests that the Grane Field hosts sand injectites that are thicker and located further away from the top reservoir than is indicated by well observations. These results may be used to improve well planning and develop reservoirs with overlying sand injectites. Supplementary material: A PDF file containing all the seismic modelling results allowing the reader to flip back and forth between the models is available at https://www.doi.org/10.6084/m9.figshare.14333102 and well logs from well 25/11-18T2 are available at https://factpages.npd.no/pbl/wellbore_documents/2358_25_1_18_COMPLETION_REPORT_AND_LOG.pdf

Deep sand distributions are a vital component of reservoir characterization and the generation of successful reservoir prediction. Through the integration of selected seismic attributes using broadband seismic data. The reservoir distributions and architectures are accurately delineate the deep Paleozoic reservoir in IO gas field. Seismic interpretation on broadband data helping in making 3D structural modeling generate a regionally accurate image of the subsurface structures at low relief three way closure on the foot wall down thrown block. Seismic attributes like (second derivative and RMS Amplitude) could indicate that Paleozoic reservoir is characterize by the presence of discrete sand body (sand-filled channel). Seismic attribute provided an extensive vertical and lateral information set for the reservoir heterogeneity between wells; it was utilized as the information requirement in prediction of lithofacies distribution. The lithofacies obtained from integrating seismic attributes from seismic broadband data are calibrated with facies identified using well logs. The lithofacies model derived from this integrated approach provides an additional geologic insight into the stratigraphic architecture and reservoir distribution and lessens exploration/development risk in the IO gas field. The IO discovery gives us a good justification to follow this well with MAAT N-1X ST3 well to reveal the sand distribution in this area (recent extraordinary success gas/condensate discovery) and it will open new opportunities that have similar phenomena. The resulting workflow is an effective tool for highlighting potential areas of better quality reservoir development, thereby demonstrating the value of this integrated approach