Mineral trapping is pursued as a geological CO₂ sequestration (GCS) mechanism because it permanently stores CO₂ in solid phases or minerals. However, CO₂ mineral-trapping mechanisms are poorly understood due to (1) lack of sufficient field and laboratory data characterizing these complex processes, and (2) challenges to develop site-specific reactive-transport models coupling fluid flow and geochemical reactions occurring at various temporal (from milliseconds to years) and spatial (from pore (millimeters) to field (kilometers)) scales. Reactive transport with additional complexities such as heterogeneity can make the simulation outputs even more difficult to interpret because of complex nonlinearity and multi-scale interdependencies. Furthermore, the values of model outputs such as concentrations can vary by several orders of magnitude, making it harder to correlate and characterize the impact of the variables via traditional data interpretation techniques such as exploratory data analyses. Recently, machine learning (ML) has shown promise in feature discovery and in highlighting hidden mechanisms that cannot be obtained by existing data-analytics and statistical methods. In this study, we applied an unsupervised ML approach, non-negative matrix factorization with custom $k$-means clustering (NMF$k$) to the data generated by reactive-transport simulations of GCS. The reactive-transport data consisted of 19 attributes, including four physio-chemical variables (pH, porosity, aqueous CO₂, and sequestered CO₂), six chemical species (K⁺, Na⁺, HCO${}_3^{-}$, Ca²⁺, Mg²⁺, Fe²⁺), and four carbonate minerals (calcite, dolomite, siderite, and ankerite), a feldspar mineral (albite), and four clay minerals (illite, clinochlore, kaolinite, and smectite) over a period of 200 years of simulation time. The simulation data used was for Morrow B sandstone at the Farnsworth hydrocarbon unit in Texas. Data are sampled at two locations within the model domain: (1) at the injection well and (2) 200 m west of the injection well. The injection was performed for a period of 10 years. Using NMF$k$, we estimated the temporal interdependencies among the 19 attributes over a span of 200 years. We found that NMF$k$ was able to identify four reaction stages and their dominant attributes; these cannot be directly discerned through traditional visualization (e.g., line plots, Pareto analysis, Glyph-based visualization methods) or exploratory data analysis tools of the simulation data. The four stages were: reactions in the injection phase followed by short-, mid-, and long-term reactions. The NMF$k$ analysis also revealed that 10 among the 19 attributes are dominant. These dominant attributes for mineral trapping include calcite, dolomite at injection well, siderite at 200 m away from the injection well, clinochlore, kaolinite, Na⁺, K⁺, Ca²⁺, Mg²⁺, pH, and aqeuous CO₂. Finally, at late times (65–200 years), our results showed that calcite plays a major role in mineral trapping with insignificant contribution from siderite, ankerite, and clay minerals. These findings make the proposed unsupervised ML-model attractive for reactive-transport sensing towards real-time GCS monitoring.

矿物捕集作为一种地质 CO ₂封存 (GCS) 机制被追求，因为它永久地将 CO ₂储存在固相或矿物中。然而，CO ₂由于 (1) 缺乏表征这些复杂过程的足够的现场和实验室数据，以及 (2) 开发耦合流体流动和不同时间发生的地球化学反应的特定地点反应输运模型的挑战，对矿物捕集机制知之甚少。毫秒到年）和空间（从孔隙（毫米）到场（公里））尺度。由于复杂的非线性和多尺度的相互依赖性，具有额外复杂性（例如异质性）的反应输运会使模拟输出更难以解释。此外，模型输出的值（例如浓度）可能会发生几个数量级的变化，这使得通过探索性数据分析等传统数据解释技术关联和表征变量的影响变得更加困难。最近，机器学习 (ML) 在特征发现和突出现有数据分析和统计方法无法获得的隐藏机制方面显示出前景。在这项研究中，我们应用了一种无监督的 ML 方法，使用自定义的非负矩阵分解$克$-均值聚类（NMF$克$) 到 GCS 的反应传输模拟生成的数据。反应传输数据由 19 个属性组成，包括四个理化变量（pH、孔隙度、含水 CO ₂和封存的 CO ₂）、六个化学物种（K ⁺、Na ⁺、HCO${}_3^{-}$、Ca ²⁺、Mg ²⁺、Fe ²⁺ ) 和四种碳酸盐矿物（方解石、白云石、菱铁矿和铁长石）、一种长石矿物（钠长石）和四种粘土矿物（伊利石、斜绿石、高岭石和蒙脱石）超过 200 年的模拟时间。使用的模拟数据是德克萨斯州法恩斯沃斯油气装置的 Morrow B 砂岩。在模型域内的两个位置采样数据：(1) 在注入井和 (2) 在注入井以西 200 m。注射进行了10年。使用 NMF$克$，我们估计了 200 年跨度中 19 个属性之间的时间相互依赖性。我们发现 NMF$克$能够识别四个反应阶段及其主要属性；这些无法通过传统的可视化（例如，线图、帕累托分析、基于字形的可视化方法）或模拟数据的探索性数据分析工具直接识别。这四个阶段是：注入阶段的反应，然后是短期、中期和长期反应。国家气象局$克$分析还显示，19 个属性中有 10 个占优势。这些捕集矿物的主要属性包括方解石、注入井白云石、距注入井200m处菱铁矿、斜绿石、高岭石、Na ⁺、K ⁺、Ca ²⁺、Mg ²⁺、pH值和含水CO ₂。最后，在晚期（65-200 年），我们的研究结果表明方解石在矿物捕集中起主要作用，菱铁矿、铁橄榄石和粘土矿物的贡献微不足道。这些发现使所提出的无监督 ML 模型对实时 GCS 监测的反应传输传感具有吸引力。