xpertsystems/oil006-sample

OIL-006是一个合成岩心样本数据集（样本版），由XpertSystems.ai提供，用于岩石物理、特殊岩心分析（SCAL）、矿物学和地质力学的机器学习应用。该数据集是一个免费预览版，与企业版模式相同，覆盖10个全球油气盆地的500个岩心样本，包含37,398个塞子测量数据，通过11个表格关联。数据总计259,321行，约20.6 MB，包括岩心主表、塞子测量、常规岩心分析、特殊岩心分析、流体饱和度、岩性描述、XRD/XRF分析、薄片岩石学、汞注入、地质力学测试和岩心标签等CSV文件。数据集基于行业标准（如API RP-40、SCA、SPWLA、Archie方程等）校准，并通过10项指标验证，总体评分为A+。它支持多种机器学习用例，如孔隙度-渗透率回归、储层质量分类、产层识别、SCAL代理模型、烃类类型预测、多表关系学习和地质力学代理建模。数据生成具有确定性，使用种子参数确保可重复性，但作为样本版，存在渗透率重尾分布、岩性覆盖有限和注入异常等限制。完整版数据集包含25,000个岩心和约350万塞子测量数据，需商业许可。

OIL-006 is a synthetic core sample dataset (sample version) provided by XpertSystems.ai for petrophysics, SCAL, mineralogy, and geomechanics machine learning applications. It is a free, schema-identical preview of the enterprise core-analysis dataset, covering 500 cores across 10 global hydrocarbon basins with 37,398 plug measurements linked across 11 tables. The dataset totals 259,321 rows across 11 CSVs, approximately 20.6 MB on disk, including files such as cores master, plug measurements, routine core analysis, special core analysis, fluid saturations, lithology descriptions, XRD/XRF analysis, thin section petrography, mercury injection, geomechanical tests, and core labels. It is calibrated to industry standards (e.g., API RP-40, SCA, SPWLA, Archies law) and validated with a 10-metric scorecard, achieving an overall Grade A+ rating. The dataset supports various machine learning use cases, such as porosity-permeability regression, reservoir quality classification, pay zone identification, SCAL surrogate modeling, hydrocarbon type prediction, multi-table relational learning, and geomechanical surrogate modeling. Generation is deterministic via a seed parameter for reproducibility, but as a sample product, it has limitations like heavy-tailed permeability distributions, limited lithology coverage, and injected anomalies. The full product includes 25,000 cores with approximately 3.5 million plug measurements and is commercially licensed.