T. Larson, J. E. Sivil, Priyank Periwal, J. Melick
{"title":"将高分辨率岩心相集成到特拉华盆地Wolfcamp和Third Bone Spring Sand的盆地级地层模型中的机器学习工作流程","authors":"T. Larson, J. E. Sivil, Priyank Periwal, J. Melick","doi":"10.1190/int-2023-0009.1","DOIUrl":null,"url":null,"abstract":"Characterization of subsurface reservoirs that are dominated by mudrock facies is hindered by the inherent heterogeneity and high degree of spatial variability typical of mudrock depositional systems. Subsurface reservoir properties that include porosity and permeability, fluid saturations, stratigraphic thicknesses of reservoir units, and source rock potential are ultimately controlled by the spatial distribution of sedimentary rock facies, which supports efforts to improve subsurface characterization workflows. Although core-based data provide direct measurements of rock attributes that are used to inform static reservoir models, capturing high-resolution core-based rock facies and downscaling these observations to tie to lower-resolution wireline logs remains a challenge. The effort to integrate core-based facies to reservoir-scale models is especially difficult when trying to capture thin-bedded heterogeneity that is common to mudrock systems. Herein a workflow is developed and applied to visualize and integrate multivariate and spatially complex core-based datasets with wireline logs. Formation-specific core-based chemofacies training datasets are developed by integrating core descriptions with chemofacies clusters developed from high-resolution X-ray fluorescence core scanning. Core-based rock attribute data (e.g., X-ray diffraction mineralogy, total porosity, and total organic matter content) are used to describe the chemofacies, providing a means to upscale low-resolution rock attribute measurements to high-resolution core-based chemofacies. Supervised core-based chemofacies training datasets are then used with neural network multi-class classification machine learning tools to train triple combo wireline logs (gamma ray, deep resistivity, bulk density, and neutron porosity) to predict rock facies from wireline logs, providing a new approach to apply core-based facies classifications to wireline log studies. A basin-scale case study that applies this work flow is described for the Third Bone Spring Sand and units of the Wolfcamp Formation in the Delaware Basin of West Texas, United States.","PeriodicalId":51318,"journal":{"name":"Interpretation-A Journal of Subsurface Characterization","volume":" ","pages":""},"PeriodicalIF":1.1000,"publicationDate":"2023-08-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"A machine learning workflow to integrate high-resolution core-based facies into basin-scale stratigraphic models for the Wolfcamp and Third Bone Spring Sand, Delaware Basin\",\"authors\":\"T. Larson, J. E. Sivil, Priyank Periwal, J. Melick\",\"doi\":\"10.1190/int-2023-0009.1\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Characterization of subsurface reservoirs that are dominated by mudrock facies is hindered by the inherent heterogeneity and high degree of spatial variability typical of mudrock depositional systems. Subsurface reservoir properties that include porosity and permeability, fluid saturations, stratigraphic thicknesses of reservoir units, and source rock potential are ultimately controlled by the spatial distribution of sedimentary rock facies, which supports efforts to improve subsurface characterization workflows. Although core-based data provide direct measurements of rock attributes that are used to inform static reservoir models, capturing high-resolution core-based rock facies and downscaling these observations to tie to lower-resolution wireline logs remains a challenge. The effort to integrate core-based facies to reservoir-scale models is especially difficult when trying to capture thin-bedded heterogeneity that is common to mudrock systems. Herein a workflow is developed and applied to visualize and integrate multivariate and spatially complex core-based datasets with wireline logs. Formation-specific core-based chemofacies training datasets are developed by integrating core descriptions with chemofacies clusters developed from high-resolution X-ray fluorescence core scanning. Core-based rock attribute data (e.g., X-ray diffraction mineralogy, total porosity, and total organic matter content) are used to describe the chemofacies, providing a means to upscale low-resolution rock attribute measurements to high-resolution core-based chemofacies. Supervised core-based chemofacies training datasets are then used with neural network multi-class classification machine learning tools to train triple combo wireline logs (gamma ray, deep resistivity, bulk density, and neutron porosity) to predict rock facies from wireline logs, providing a new approach to apply core-based facies classifications to wireline log studies. A basin-scale case study that applies this work flow is described for the Third Bone Spring Sand and units of the Wolfcamp Formation in the Delaware Basin of West Texas, United States.\",\"PeriodicalId\":51318,\"journal\":{\"name\":\"Interpretation-A Journal of Subsurface Characterization\",\"volume\":\" \",\"pages\":\"\"},\"PeriodicalIF\":1.1000,\"publicationDate\":\"2023-08-17\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Interpretation-A Journal of Subsurface Characterization\",\"FirstCategoryId\":\"89\",\"ListUrlMain\":\"https://doi.org/10.1190/int-2023-0009.1\",\"RegionNum\":4,\"RegionCategory\":\"地球科学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q3\",\"JCRName\":\"GEOCHEMISTRY & GEOPHYSICS\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Interpretation-A Journal of Subsurface Characterization","FirstCategoryId":"89","ListUrlMain":"https://doi.org/10.1190/int-2023-0009.1","RegionNum":4,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"GEOCHEMISTRY & GEOPHYSICS","Score":null,"Total":0}
A machine learning workflow to integrate high-resolution core-based facies into basin-scale stratigraphic models for the Wolfcamp and Third Bone Spring Sand, Delaware Basin
Characterization of subsurface reservoirs that are dominated by mudrock facies is hindered by the inherent heterogeneity and high degree of spatial variability typical of mudrock depositional systems. Subsurface reservoir properties that include porosity and permeability, fluid saturations, stratigraphic thicknesses of reservoir units, and source rock potential are ultimately controlled by the spatial distribution of sedimentary rock facies, which supports efforts to improve subsurface characterization workflows. Although core-based data provide direct measurements of rock attributes that are used to inform static reservoir models, capturing high-resolution core-based rock facies and downscaling these observations to tie to lower-resolution wireline logs remains a challenge. The effort to integrate core-based facies to reservoir-scale models is especially difficult when trying to capture thin-bedded heterogeneity that is common to mudrock systems. Herein a workflow is developed and applied to visualize and integrate multivariate and spatially complex core-based datasets with wireline logs. Formation-specific core-based chemofacies training datasets are developed by integrating core descriptions with chemofacies clusters developed from high-resolution X-ray fluorescence core scanning. Core-based rock attribute data (e.g., X-ray diffraction mineralogy, total porosity, and total organic matter content) are used to describe the chemofacies, providing a means to upscale low-resolution rock attribute measurements to high-resolution core-based chemofacies. Supervised core-based chemofacies training datasets are then used with neural network multi-class classification machine learning tools to train triple combo wireline logs (gamma ray, deep resistivity, bulk density, and neutron porosity) to predict rock facies from wireline logs, providing a new approach to apply core-based facies classifications to wireline log studies. A basin-scale case study that applies this work flow is described for the Third Bone Spring Sand and units of the Wolfcamp Formation in the Delaware Basin of West Texas, United States.
期刊介绍:
***Jointly published by the American Association of Petroleum Geologists (AAPG) and the Society of Exploration Geophysicists (SEG)***
Interpretation is a new, peer-reviewed journal for advancing the practice of subsurface interpretation.