CO2 pipelines release and dispersion: A review

Carbon Capture, Utilization, and Storage (CCUS) is an emerging industry in response to climate change. With this, more pipelines are being built to transport CO2 to sequestration sites. CO2, though non-toxic and non-flammable, is an asphyxiant that can cause death when exposed to high concentrations. Hence, there is a need to evaluate the release and dispersion behavior of CO2 from pipelines. Industrial-scale experiments of CO2 dispersion have been performed to acquire quantitative data. However, the sheer number of scenarios that could possibly occur is difficult to perform one-by-one experimentally. Computational tools, such as Computational Fluid Dynamics (CFD), and risk assessment methodologies, such as Quantitative Risk Assessment (QRA), have then been utilized to predict CO2 dispersion behavior and hazards, respectively. In this review, the different industrial-scale experiments, CFD models for CO2 dispersion, and risk assessment of CO2 transport pipelines were discussed. Moreover, we emphasize the possibility of using machine learning as a tool to increase the accuracy and efficiency in predicting CO2 dispersion behavior. The future direction of predicting CO2 dispersion behavior through the use of machine learning was also explored.

Automated summary

Carbon Capture, Utilization, and Storage (CCUS) is an emerging industry that aims to address climate change. The construction of more pipelines for transporting CO2 to storage sites is a key aspect of this industry. Evaluating the behavior of CO2 release and dispersion from these pipelines is necessary due to the potential hazards of high concentration exposure. Industrial-scale experiments have provided quantitative data, but conducting experiments for every possible scenario is impractical. Computational tools such as Computational Fluid Dynamics (CFD) and risk assessment methodologies such as Quantitative Risk Assessment (QRA) have been used to predict CO2 dispersion behavior and hazards. This review discusses the different experiments, CFD models, and risk assessments related to CO2 transport pipelines. The potential of using machine learning to improve the accuracy and efficiency of predicting CO2 dispersion behavior is also explored.