Ongoing research on structural modeling and 3D visualization of the Junggar Basin
1. Project Overview
The Junggar Basin Whole Basin Structural Modeling and 3D Visualization Continuous Research (Phase II) project is of great significance in the field of petroleum geology. It aims to deeply explore the geological structure of the Junggar Basin through advanced technical means and scientific research methods, build an accurate 3D structural model, and provide indispensable basic data and powerful model support for oil exploration and development, etc., to help improve the efficiency and accuracy of related work and promote the further development of petroleum geological research.
2. Project Overview and Task Completion
(I) Work Area Overview
The work area is vast and covers the entire Junggar Basin, with an area of about 130,000 km², a north-south span of 100-390 km, and an east-west extension of 200-650 km. It contains rich and diverse geological data, with 196 3D survey lines and 5059 2D survey lines. In the process of research, the interpretation results of Phase I, Phase II and Phase III were fully integrated, and a three-dimensional structural model of the entire basin from the bottom of the Shawan Formation of the Neogene to the top of the Carboniferous was constructed based on this, which established a key framework for a comprehensive analysis of the geological structure of the basin.
(II) Research tasks and assessment indicators
1. Research tasks
The data collection work is comprehensive and in-depth, covering the structural interpretation results of Phase I, Phase II and Phase III of the entire basin and the relevant data of 3,278 drilling wells, of which the number of broken rods is as high as more than 130,000, laying a solid data foundation for subsequent research. The data collation link is crucial, and it is necessary to systematically sort out the complex structural interpretation results and make meticulous level divisions for a large number of faults in the entire basin. In view of the fact that most of the fault combinations in the original fault interpretation results lack unified naming and poor section consistency, it is urgent to carry out fault re-examination and section combination work, and at the same time, the stratigraphic interpretation scheme of the 19 layers is refined to meet the strict requirements of three-dimensional structural modeling. On this basis, we established a precise fault cutting relationship, constructed a time domain fault model of the whole basin's level III faults, and established a complete and unified time domain structural model in combination with the stratigraphic information. Then, we used the velocity field provided by Party A to convert it into a three-dimensional structural model of the whole basin in the depth domain. During the whole process, we worked closely with the project team to conduct strict inspection and quality control, continuously improve and optimize the model, and ensure the scientificity and reliability of the model.
2. Assessment indicators
The assessment indicators are clear and challenging. It is necessary to successfully complete the 19-layer time domain and depth domain three-dimensional structural model of the whole basin, draw the structural map of 19 layers and the thickness map of 18 layers, and train 10 technicians who can skillfully apply the three-dimensional structural model of the whole basin for Party A to ensure that they can deepen the application of the model in the exploration institute, geophysical institute and other departments, so as to effectively meet the needs of professional and technical personnel in petroleum geological research and exploration work.
(III) Project operation inspection
During the implementation of the project, we received strong support and careful guidance from leaders and experts, and conducted strict quality inspection and monitoring of various achievements through more than 20 important meetings. These meetings conducted in-depth discussions and strict reviews on the results of different stages. From basic data sorting to model construction, all aspects ensured that the project was carried out in a scientific and reasonable process, effectively guaranteed the quality of the project, and provided a solid guarantee for the smooth progress of the project.
(IV) Actual completion of work
The actual completion of work has achieved remarkable results. In terms of data collection, it is comprehensive and systematic. In terms of quality control, it has completed the fault naming inspection, fault rationality inspection, and stratum closure quality inspection of the entire basin and stratum closure difference inspection between work areas. In addition, more than 20,000 fault rods, 720 faults and 3 layers of strata have been added. In terms of model construction, the time domain model, velocity model and structural model of the entire basin have been successfully completed. In terms of map drawing, 20 top surface structural maps and 19 thickness maps of the entire basin have been completed. In addition, 10 structural geological profiles have been added, which provide rich and detailed data for in-depth research on the geological structure of the basin and strongly promote the progress of project research.
3. Seismic interpretation quality control
(I) Sorting out interpretation results of the whole basin
The scale of the collection and sorting work is huge. On the basis of more than 130,000 broken rods, more than 20,000 were added, and the total reached more than 150,000. At the same time, in the LandMark interpretation area, 19 layers of fault plane combination polygons, 22 layers of stratigraphic data and 19 layers of boundary polygons were carefully collected in the whole basin. The stratification data of key wells were also fully collected, covering 40 wells in Zhundong, 33 wells in Mahu slope, 49 wells in the abdomen, 20 wells at risk, 28 wells in the south margin, 167 wells around Kelameili, and 97 wells in Hongcheguai. The stratification data tables of 138 wells in the first phase project and the stratification data tables of key wells in the whole basin were sorted out. Finally, the stratification data of 537 wells were counted, which provided sufficient data support for subsequent analysis, just like building a solid data building to support subsequent research work.
(II) Quality control of fault data
1. Quality control standards
Strict quality control standards have been formulated, requiring that all faults must have unique names, the breakpoints of the same fault must be tightly closed, and the data should be on the same trend surface without any jump points. At the same time, the cross-section in the fault model must accurately coincide with the original data to ensure the accuracy and consistency of the fault data, laying the foundation for building a high-quality geological model.
2. Problems and difficulties
However, there are many difficulties in actual work. For example, the naming of faults is not uniform, which makes it difficult to combine the same section space; the inclination of the fault rods of the same fault changes significantly laterally, and the spatial morphology of the section is extremely unreasonable, making the optimization work extremely difficult and labor-intensive; the mismatch between the longitudinal fault rods and the plane polygons further increases the difficulty of spatial combination of sections; in some local areas, there are only plane polygons but no fault rods, which makes it impossible to establish sections or the section range is inaccurate; in some areas, due to the sparse measurement network, it is difficult to combine sections; the same inherited fault is interpreted as multiple sections, which seriously interferes with the combination of sections; there are also multiple fault interpretation results in some local areas, which causes data confusion. These problems are like obstacles that hinder the smooth progress of research work.
3. Solutions
In response to these problems, the project team has adopted a series of effective solutions. LandMark was combined with EarthVision to successfully complete the combination and naming of sections; the three-dimensional breakpoints were carefully edited and optimized to ensure that the breakpoints of the same fault were basically on the same trend surface and to eliminate obvious jump points; on the basis of following the original flat section interpretation results, the density of broken rod interpretation was appropriately increased; broken rods were supplemented and interpreted in Landmark, with more than 20,000 new ones added; using the three-dimensional spatial advantages of EarthVision, new control points were added according to the macro trend of the section; multiple sections of the same inherited fault were integrated into the same section; the most reasonable interpretation scheme was retained, and redundant or repeated broken rods were decisively deleted. Through these measures, the quality of fault data was effectively improved, just like removing the stumbling block on the road ahead, and promoting the research work to move forward.
(III) Quality control of layer data
1. Quality control standards
There are also strict quality control standards, requiring that the layers should not be crossed or jumped, the layers in the model should completely overlap with the original layer data, the layers after time-depth conversion must be consistent with the well layers, and there should be no closure difference between adjacent work areas in the same layer, so as to ensure the reliability and accuracy of the layer data and provide accurate stratigraphic information for the geological model.
2. Problems and difficulties
However, many problems were encountered in actual operation. Some 2D and 3D interpretation schemes are repeated and quite different, and each scheme has its own advantages and disadvantages, which makes it difficult to select and splice the layers. In the low-exploration area dominated by the 2D survey network, the interpretation schemes are scarce and partially unreasonable, and the closure error is large, which seriously affects the accuracy of the model. There is an interlacing phenomenon between the layers and the sections, which requires a lot of manpower to modify the interpretation scheme. The layers are missing in some areas because they are too far away from the sections, which in turn causes structural distortion and affects the quality of the model. In some areas with complex structures, due to multi-stage tectonic movements, the stratigraphic contact relationship is extremely complex, the stratigraphic contact relationship in some areas is unreasonable, and the stratigraphic deposition in some areas is relatively thin, so the interpretation schemes are scarce and the modeling is extremely difficult. There is a lack of stratigraphic constraints above the shallow Tn1s, which makes the stratigraphic contact relationship of some parts of the model chaotic. These problems have brought great challenges to the quality control of stratigraphic data.
3. Solutions
To solve these problems, the project team has formulated corresponding strategies. When selecting interpretation schemes, we followed the priority of III>II>I phases, and gave priority to the three-dimensional work area scheme; we carefully deleted unreasonable areas, and reasonably added control points in areas with fewer schemes; in the case of interlacing of layers and sections, we deleted the layer points beyond the section; we added control points near the fault to solve the problem of missing layers and structural distortion in local areas; in areas with thinner local layer deposition and lower control, we added control points in three-dimensional space; by adding three layers of data, Tn2, Tn2d, and RG, we solved the problem of shallow unconstrained layers. Through these efforts, we optimized and inspected 22 layers of layer data, and completed 3 layers in excess, which effectively guaranteed the quality of the layer data and cleared the obstacles for building an accurate geological model.
IV. 3D structural modeling of the whole basin
(I) Modeling technical process
The main technical process includes establishing the overall work area boundary polygon and the single work area boundary polygon, constructing the overall work area time fault file and the fusion area boundary, determining the calculation order of the single work area, reasonably setting the grid extrapolation factor, grid smoothing coefficient, grid tolerance and other key parameters, preparing the layer data, using polygons to obtain the single work area fault data and gridding, establishing the single work area fault model and layer model, and then establishing the time structure model, velocity model and depth model, and at the same time performing the time depth data, well layer correction, error analysis and industrial drawing, and finally adding the stitching boundary layer data to complete the entire modeling process. This series of processes are closely linked, like the operation of a precision machine, ensuring the scientificity and accuracy of the modeling work.
(II) Optimization of modeling parameters
1. Plane block division
The plane block division is based on structural units. Given the huge area, numerous layers, and huge amount of grid data in the whole basin model, the calculation time is too long (originally each calculation takes 70-80 hours), the model modification cycle is long and inefficient, which seriously affects the project progress. After optimization, it is divided into 21 blocks (14 blocks initially). After block calculation, the average calculation time is shortened to 5 hours per time, and the longest is no more than 10 hours, which greatly improves the work efficiency, just like splitting a huge project into multiple small modules, so that the work can be efficiently promoted.
2. Grid size
After multiple experimental comparisons, the grid size was finally determined to be 250m×250m, which not only ensures the accuracy of the model but also takes into account the calculation efficiency, just like finding the best balance between accuracy and efficiency, ensuring that the modeling work can accurately reflect the geological structure and will not fall into trouble due to excessive calculation.
3. Extrapolation factor
When the extrapolation factor is set to 0, abnormal values at the boundary or near the fault will interfere with the change of the stratigraphic trend surface, causing the stratigraphic model to be wrong. After repeated experiments, it is preferably 0.1, which ensures the accuracy of the stratigraphic model near the boundary and fault, and avoids the model deviation caused by improper extrapolation factors.
4. Accuracy coefficient
After multiple tests, the accuracy coefficient, when set to 4, can not only ensure a certain calculation speed, but also eliminate the influence of the fault block, so that the layer is more consistent with the original data. Compared with other values (such as 40 or more will appear scaly display, affecting the accuracy of the model), it is more reasonable and provides a suitable parameter setting for building a high-quality geological model.
5. Tolerance
After comparison, the tolerance is selected as 1000m in the XY direction and 500m in the Z direction. This parameter can effectively control the influence range of stratigraphic data at the section, avoid the loss of valid data due to too large a range or structural distortion due to too small a range, and ensure the rationality and reliability of stratigraphic data at the section.
(III) Establishment of time domain structural model
1. Modeling process
The establishment process of time domain structural model is complex and rigorous. Its modeling process includes extracting single work area modeling data files, carefully setting single work area modeling parameters, fusing boundaries, and generating single work area models. It involves overall fault scatter data, overall layer scatter data, 21 small work area boundary ranges, each fault scatter file, each fault polygon file, each fault grid file, large work area boundary range, 21 small work area modeling data files, small work area modeling calculation sequence settings and other links. Finally, the 21 small work area models are combined into an overall model. Each link is closely connected and indispensable to jointly build the building of the time domain structural model.
2. Key steps
In the block boundary fusion stage, by defining the fusion boundary, deleting the stratigraphic data points within the fusion range and performing data back-insertion, the model errors between adjacent blocks are effectively eliminated, and the seamless fusion of all blocks in the basin is achieved. The three-dimensional multi-angle stereoscopic display verification shows that the effect is good, just like perfectly stitching together the various parts of a puzzle to present a complete geological structure picture. In terms of fault model establishment and quality inspection, the section plane polygons were adjusted and key parameters were optimized. For some polygon files that could not correctly control the section grid extension range due to insufficient control points, manual editing was adopted, and a total of more than 3,200 fault polygons were manually edited; the section spatial morphology was optimized by adding three-dimensional control points; manual intervention and optimization were performed to address the problem of invalidation of some section scattered point data in the block model automatically generated by the software; manual adjustments were made to the fault cutting relationships that were incorrectly processed automatically by the software to make them more consistent with geological understanding; after multiple rounds of optimization, the fault polygons in the model are highly consistent with the interpretation area and more complete, the section model is basically consistent with the fault data, and the section spatial morphology is more reasonable and more in line with geological laws, ensuring the accuracy and reliability of the fault model. In terms of formation model establishment and quality inspection, the formation model is closely consistent with the interpretation scheme, the formation distribution and contact relationship are intuitive and clear, and formation trend control points are added in the unreasonable interpretation scheme area to accurately control the formation distribution range. The boundaries of each layer in the model are basically consistent with the interpretation scheme. After optimizing the local unreasonable areas, the model quality is significantly improved, and the formation and formation scattered point data are consistent. The model is highly reliable and provides accurate formation information for the time domain structural model. Finally, after testing 10 large 2D seismic regional profiles and 11 structural geological profiles, the model is reasonable and reliable. Through multiple combinations with the project team, the interpretation scheme and model of the seismic 2D survey line are compared, the seismic interpretation scheme and the modified model are continuously optimized, and finally a high-quality time domain model is obtained. After layer-by-layer testing and optimization, it is ensured that the model can truly reflect the geological structure.
(IV) Velocity model establishment
The velocity model is established based on the basin-wide velocity data provided by the ground objects, and the average velocity field of the basin is successfully established. Layered data is added to the workflow for model correction to ensure the accuracy of the velocity model, just like installing a precise clock for the model, so that it can accurately reflect the changes in geological structure in time and space.
(V) Depth domain
Establishment of structural model
The depth domain structural model is obtained by converting the time domain model through the velocity field of the whole basin, and correcting it in combination with the well stratification. After strict quality control, such as statistical analysis of the percentage of well-seismic errors of the Jurassic bottom boundary and the Triassic Baikouquan Formation, the depth domain structural model is more accurate, providing a reliable basis for geological research, just like lighting a bright light on the road of geological research, guiding the research direction.
V. Display and application of model results
(I) The results of three-dimensional model are more reasonable and comprehensive
The results of three-dimensional fault model have made a qualitative leap, the cross-section morphology is more reasonable, and the cross-section scheme is more refined, increasing from 963 in the first phase to 3256, providing more abundant information for the study of fault distribution and geological structure, just like opening a door to the depths of geology, allowing researchers to have a clearer insight into the mystery of geological structure. For the first time, a basin-level time-domain-depth-domain integrated three-dimensional structural model was realized. The spatial distribution of the strata is more accurate, highly consistent with the actual strata distribution, and the abnormal fault blocks in the model have been successfully eliminated, which improves the reliability of the model. It is like creating a precise scalpel for geological research, which can more accurately analyze the geological structure.
(II) The new structural results map is more accurate
20 structural maps and 19 thickness maps have been completed, and one structural map and one thickness map have been added. These maps can fully display the overall structural morphology of the Junggar Basin. Through arbitrary cutting, the structural change trend between the work areas can be clearly presented. Local details can also be displayed independently by block, providing important reference materials for geological research and exploration work, just like drawing a detailed geological map to guide researchers.
(III) Stronger later expansion applications
It can realize the rapid extraction of any structural profile in the whole basin, providing strong support for basin strategic selection, risk exploration and large-scale conferences; it can be optimized by combining satellite images and surface auxiliary well locations and seismic deployment plans to improve exploration efficiency; it provides key technical support for sedimentary simulation and basin simulation, and can quickly calculate the area, volume and resource volume of the target layer. Its reserve calculation module can perform comprehensive statistical analysis on model data, instantly calculate the volume currently displayed in the window, and perform statistical analysis and result output of all information according to the structural model, layer, block and calculation parameters, which can be output in ASCII text file or spreadsheet format; it can also quickly generate thickness maps between any target layers, providing strong support for subsequent research such as paleo-tectonic research; at the same time, it provides indispensable basic maps for sedimentary systems, reservoir analysis and geochemical research, and plays an important role in the field of petroleum geological research, as if injecting strong power into petroleum geological research and promoting the continuous deepening of research work.
VI. Understanding and Suggestions
(I) Understanding
The project team deeply realized that during the implementation of the project, a set of effective methods for modeling large-scale multi-fault structures in the whole basin was explored, that is, the whole basin was divided into multiple blocks according to the complexity of the fault system and the structural unit, and then the whole basin structural model was gradually integrated. At the same time, it was clarified that the quality of the structural interpretation results has a direct and critical impact on the time and accuracy of three-dimensional structural modeling. Only by continuously improving the quality of structural interpretation results can the quality of three-dimensional structural models be further improved, which provides valuable experience and lessons for future research work.
(II) Suggestions
Based on this, constructive suggestions were put forward. On the one hand, it is necessary to deepen and vigorously promote the application of the whole basin three-dimensional structural model in large-scale regional research such as risk and strategy, and give full play to the value of the model; on the other hand, on the basis of further improving the seismic interpretation results of the whole basin, the whole basin structural model should be updated in time to ensure the timeliness and accuracy of the model, provide better services for subsequent geological research and oil exploration work, and point out the direction for future research work. In summary, the continuous research on structural modeling and three-dimensional visualization of the whole basin in the Junggar Basin (Phase II)