New Method Predicts Hazard Zone Post-CO₂ Leak

Carbon capture, storage and utilization (CCUS) is an important technology for meeting global carbon emission reduction targets. The development of CO₂ transportation, as a link in the CCUS industry chain, is crucial for CCUS projects.

The supercritical or dense phase is widely recognized as the optimal phase state for carbon dioxide (CO₂) transport. Therefore, it is of great value and significance to ensure the safe and efficient transportation of CO₂ in this phase state.

In a study published in the KeAi journal Journal of Pipeline Science and Engineering, the PipeChina Group from China conducted the first full-size CO₂ pipeline burst fracture test in China to evaluate the pipeline's fracture arrest performance.

"CO₂ leaks caused by pipeline breaks can have more serious consequences than property damage," says lead author Prof. Yuxing Li from the Key Laboratory of Oil and Gas Storage and Transportation Safety in Shandong Province, China University of Petroleum (East China)" Due to the positive throttling effect of CO₂ and the toxicity of high concentrations of CO₂, it can frostbite or even cause asphyxiation of plants and animals near the leakage area. Therefore, it is meaningful to study the leakage characteristics of supercritical/concentrated-phase CO₂ and predict its potential hazard distance."

The team first carried out four sets of full-size burst tests with different initial conditions to clarify the effect of initial conditions on the CO₂ concentration in the near and far field of leakage. The researchers then verified the CO₂ concentration diffusion model through the measured concentration data, on the basis of which the CO₂ hazard distance calculation model was proposed.

"There are large temperature and pressure differences between the start and end points of industrial-grade CO₂ pipelines, and leakage at any location of the pipeline will lead to different leakage consequences," shares Li. "Meanwhile, the relative distance between the leakage point and the cut-off valve will affect the CO₂ leakage characteristics and thus the delineation of the hazard distance."

Taking into account these factors, it is therefore difficult to predict the hazard distance due to leakage at different locations. To that end, the team proposed a PSO-BP neural network to predict the hazard distance for leaks at any location, which is consistent with the results of the CO₂ concentration diffusion model but with greatly reduced computational demands.