@article { author = {Samadi, Iman and Kordi, Masoumeh and Soleimani Monfared, Mehrdad and Ahmadi, Amir}, title = {Introducing an integrated strategy in fault modelling with multi-attributes in 3D seismic data in a field from Persian Gulf}, journal = {Journal of the Earth and Space Physics}, volume = {46}, number = {1}, pages = {81-96}, year = {2020}, publisher = {Institute of Geophysics, University of Tehran}, issn = {2538-371X}, eissn = {2538-3906}, doi = {10.22059/jesphys.2020.286837.1007147}, abstract = {Fault and fracture modelling is an important step in reservoir engineering which is required for any reservoir characterization and production management. There are various types of methods and strategies for building such models, however, each has its own advantages and drawbacks. The most important issue that should be considered is the ability to model both large- and small-scale faults, simultaneously. It is important, as large faults define geological frameworks of the reservoir, while small faults influence fluid movement in the reservoir. In this study, we introduce an integrated strategy for modelling small- and large-scale faults by seismic data, using multi-attributes. Large faults are defined by hand picking from seismic data using attributes, and small faults are modelled by an automatic ant tracking algorithm. Then, two separated models are integrated to build a unique, but multi-scale fault model. Result of each step of modelling is evaluated by well data. The methodology is applied on a hydrocarbon reservoir from the Persian Gulf. Results show that the multi scale fault model is accurate when evaluated by well data. Integrated modelling of faults of fractures to obtain a unique multi-scale model is an interesting topic in reservoir engineering. Normally fractured reservoirs are divided into several production zones based on division made by large faults, while fluid movement in each zone is controlled by small fracturs and faults. Thus, obtaining a unique model which contain information of faults in several scale is under investigation. However, conventional methods use separate sources of information for modelling faults in various scales. Large scale faults are normally modelled by seismic data while well data are used for modelling small faults. Ozkaya (2019) stated that modelling of faults both with seismic and well data would reduce uncertainty in reservoir fracture modelling. Cao et al. (2019) introduced an integrated strategy for modeling faults with two scales in 2D seismic data, but using seismic and well data. Kurison et al. (2019) have modelled faults and fractures in reservoir with 3D seismic data and well data, but in separate manners. But their final interpretation has shown that using both types of model would result in better reservoir modelling. Xu et al. (2019) introduced an integrated strategy for modelling faults and fractures in two scales simultaneously using seismic and well data. In this study, we introduce an integrated strategy for multi-scale fault modelling using only seismic data, which could be used in reservoirs which lack of well data. The proposed strategy introduced here, initiates with a geological model building. Subsequently, large faults can be defined on seismic data and related attributes. Simultaneously, small scale faults can be modelled by an ant tracking algorithm in an automatic manner, then it would be refined by interpreter to remove other lineaments than fault that was modelled by the algorithm. Each model then would be evaluated by well data and in case of any error in the model, they would be removed by more ant tracking parameter optimizations and also deeper investigation by the interpreter. In the final step, both fault model would be integrated to build a unique informative multi-scale fault model which contains information of all faults in various sizes. Other characteristics of faults in the integrated model would be investigated for further analysis. Large scale fault model showed major faults with northwest-southeast trending acting in the center of the reservoir, which has a dome shaped structure, and some minor faults with various trending around the major one. Through this modeling curvature, chaos and variance attributes were used for better fault detection. Small faults obtained by ant tracking distributed around the center of the field. Ant tracking algorithm parameter were optimized through sensitivity analysis prior to application. Afterwards, fault model was refined to remove non-fault lineament. Both models were evaluated by a fullbore formation microimager (FMI) log which proved fractures and faults that were obtained by seismic data. One fault that was detected by the proposed strategy were also captured by well. Then both fault models were integrated to a unique model and faults were modeled by deterministic method. The integrated fault model obtained by the proposed strategy revealed the importance of a multi-scale fault model in reservoir engineering. Large faults of the study reservoir showed different zones of fractures in the formation reservoir, while small faults in the same model built a discrete network of fractures which provides canals for fluid movement. The integrated model shows that large faults in the study field are mostly in the center of the reservoir, while small faults are distributed through the edges of the formation reservoir, which could be used for further investigation of locating for production and/or injection wells.}, keywords = {multiscale modelling,seismic attributes,curvatures,Chaos,Ant tracking algorithm}, title_fa = {ارائه استراتژی ترکیبی مدل‌سازی گسل‌ها به‌روش چند نشانگری در داده‌های لرزه‌ای سه‌بعدی در یکی از میادین خلیج فارس}, abstract_fa = {شناسایی و مطالعه گسل­ها در مخازن هیدروکربنی، اهمیت ویژه­ای در مراحل ازدیاد برداشت و توسعه میدان دارد. در بررسی ساختارهای با زمین‌شناسی پیچیده، تفسیر گسل­ها با عدم‌قطعیت بالایی همراه خواهد بود. روش­های متعارف تفسیر و مدل­سازی گسل­ها در داده­های لرزه­ای علاوه بر نیاز به دانش زمین­شناسی مفسر­ که خود می­تواند به‌عنوان منشأ عدم ‌قطعیت باشد، فرآیندی بسیار دشوار و وقت­گیر است. بدین‌منظور در این مطالعه یک استراتژی خودکار و ترکیبی به‌منظور افزایش دقت و سرعت مدل­سازی گسل­ها و شکستگی­ها در داده لرزه­ای معرفی می­شود. گسل­ها به­طور معمول با استفاده از نشانگرهای لرزه­ای تفسیر می­شوند. به‌منظور تفسیر گسل­های میدان مورد نظر در این تحقیق، ابتدا نشانگرهای آشفتگی، واریانس، انحنا و الگوریتم ردیابی مورچه از داده‌های لرزه­ای استخراج شد. از بین نشانگرهای موجود، نشانگرهای آشفتگی، واریانس و انحنا به‌طور واضح گسل‌های بزرگ‌مقیاس را مشخص کردند. گسل­های کوچک‌مقیاس که شناسایی آنها در داده­های لرزه­ای دشوار است، به‌کمک الگوریتم ردیابی مورچه مدل‌سازی شدند. به‌کارگیری روش‌های بیان‌شده در تفسیر ساختاری مخزن در کنار مدل­سازی قطعی گسل­ها به‌روش ترکیبی بر روی داده­های لرزه­ای، نشان‌دهنده شناسایی و تفسیر بهتر گسل­ها با استفاده از استراتژی پیشنهادی و رویکرد ترکیب روش­های موجود بود. نتایج حاصل از تفسیر چند نشانگری و همچنین مدل­سازی گسل­ها در میدان مورد مطالعه، انطباق خوبی با اطلاعات زمین‌شناسی نشان داد. لذا می‌توان پیشنهاد داد استراتژی به‌کار گرفته‌شده در مدل­سازی و استفاده از یافته‌های مطالعات چند نشانگری می‌توانند به‌منظور افزایش دقت در مطالعات ساختاری مخزن، مورد استفاده قرار‌گیرند.}, keywords_fa = {مدل‌سازی چند‌مقیاسی,نشانگر لرزه‌ای,انحنا,آشفتگی,الگوریتم ردیابی مورچه}, url = {https://jesphys.ut.ac.ir/article_74736.html}, eprint = {https://jesphys.ut.ac.ir/article_74736_7918ff3d4864abd2d35820d6b3c5e9b2.pdf} }