The study explores dataset preparation, machine learning (ML) model training, interpretation, and life cycle assessment (LCA) to predict and enhance the performance and sustainability of rubberized concrete. A large dataset comprising 1209 collected samples with nine input features was used to train and evaluate six machine learning models. Among the six models, the Light gradient boosting machine (LightGBM) model achieved the highest prediction accuracy on the testing dataset, with an R² value exceeding 0.96, a MAPE of 8.31 %, a MAE of 2.36 MPa, and a RMSE of 3.25 MPa. The SHAP algorithm was used to interpret predictions and identify key factors influencing compressive strength. Rubber content and water-to-cement ratio reduced strength, while longer curing time, more superplasticizer, higher fine aggregate content, and a greater silica fume-to-cement ratio improved it. Coarse aggregate, crumb rubber size, and cement content had minimal impact. Optimal performance was achieved with: Rubber content <55.92 kg/m³, w/c ratio <0.4, curing time >26 days, superplasticizer >3.7 kg/m³, fine aggregate >642 kg/m³, and silica fume/cement ratio >2.5 %. LCA results show that, although rubberized concrete offers no clear advantage over conventional concrete in cost, carbon, or energy, the lower strength and higher superplasticizer use of rubberized concrete lead to greater strength-normalized impacts, and therefore its value lies more in waste recycling and toughness than in strength-based sustainability. The unique advantage of this research lies in the development of a ML-LCA integration framework that synthetically balances performance prediction and sustainability assessment of rubberized concrete, with delivering actionable mix design recommendations, identifying key and low-impact variables, and revealing trade-offs among strength, cost, and environmental performance.

Foam is utilized in enhanced oil recovery and CO₂ sequestration. Surfactant-alternating-gas (SAG) is a preferred approach for placing foam into reservoirs, due to it enhances gas injection and minimizes corrosion in facilities. Our previous studies with similar permeability cores show that during SAG injection, several banks occupy the area near the well where fluid exhibits distinct behaviour. However, underground reservoirs are heterogeneous, often layered. It is crucial to understand the effect of permeability on fluid behaviour and injectivity in a SAG process. In this work, coreflood experiments are conducted in cores with permeabilities ranging from 16 to 2300 mD. We observe the same sequence of banks in cores with different permeabilities. However, the speed at which banks propagate and their overall mobility can vary depending on permeability. At higher permeabilities, the gas-dissolution bank and the forced-imbibition bank progress more rapidly during liquid injection. The total mobilities of both banks decrease with permeability. By utilizing a bank-propagation model, we scale up our experimental findings and compare them to results obtained using the Peaceman equation. Our findings reveal that the liquid injectivity in a SAG foam process is misestimated by conventional simulators based on the Peaceman equation. The lower the formation permeability, the greater the error.

Aqueous foam is a dispersion of gas in liquid, where the liquid acts as the continuous phase and the gas is separated by thin liquid films stabilized by a surfactant. Foam injection is a widely used technique in various applications, including CO2 sequestration, enhanced oil recovery, soil remediation, etc. Surfactant-alternating-gas (SAG) is a preferred approach for foam injection, and injectivity plays a vital role in determining the efficiency of the SAG process. Different gases can be applied depending on the process requirements and availability. However, the underlying mechanisms by which gas composition impacts injectivity are not yet fully understood. In this work, the effect of gas composition on fluid behavior and injectivity in a SAG process was investigated using three gases: N2, CO2, and Kr. Our observations revealed that gas solubility in liquid was key for the formation and evolution of liquid fingers, and therefore was very important for liquid injectivity. A lower gas solubility in liquid led to a slower increase in surfactant solution injectivity. In addition, the development of surfactant solution injectivity took significantly longer when the surfactant solution was partially pre-saturated compared to when it was unsaturated. Additionally, the propagation of the collapsed-foam bank during gas injection was accelerated when the gas had a greater solubility in water.

Large volumes of carbon dioxide (CO₂) captured from carbon emission source can be stored in deep saline aquifers as a mean of mitigating climate change. The deep saline aquifers are naturally heterogeneous at multiple scales. It is important to generate representative multiscale heterogeneous fields of various hydrogeologic properties and understand storage safety by studying CO₂ migration and distribution in such fields. In this work, a new multiscale heterogeneous model with partly fine multi-facies heterogeneous domain is proposed. A method based on transition probability theory is referred to establish a multi-facies model. A new multiscale heterogeneous model with partly fine multi-facies heterogeneous domain is built up according to the categorized permeability data obtained from the Geological Carbon Storage Frio site in USA. TOUGH2/ECO2N is applied to simulate CO₂ migration and distribution in such a multiscale heterogeneous model. The CO₂ plume shows obvious viscous fingering and non-uniform migration both in layered and vertical directions, implying vertical and horizontal heterogeneity which cannot be represented by a single-scale model or simulated with the assumption of homogeneous formation. The profile of CO₂ migration shown in the numerical simulation at a time of 10 days is in a good accordance with the seismic data of Frio situ in qualitative and quantitative aspects.