(128f) Robust Insights into CO2 Hydrate Kinetics in Brine-Saturated Sediments Using Experimental and Machine Learning Approaches for Ccus Application

One of the potential Carbon Capture and Sequestration (CCS) methods is capturing CO₂ and injecting it into the oceanic sediments where CO₂ is converted into CO₂ gas hydrates. However, the presence of brine [NaCl] in sediments likely affects the kinetics of CO₂ hydrate formation significantly. To tackle this challenge, a real-world scenario was mimicked in this work, by creating artificial brine-rich different-sized sediment beds in a laboratory-scale reactor. Afterward, CO₂ was injected multiple times via injection tube directly into the sediment, and the CO₂ hydrate formation, dissociation, and morphological changes were recorded.

The experimental results show that the water-to-hydrate conversion [%] was estimated to be about highest in smaller size sediments [~60%] > both (smaller + larger sediments) dual [~35 %] > larger sediments [~20 %]. The visual observations show the smaller-size sediments show the haphazard isotropic advancement of whitish CO₂ hydrate whereas in the presence of large-sized sediments, various hydrate capping which trapped brine or CO₂ was observed. Moreover, using ~94,000 experimental data (% water to hydrate conversion, pressure, and temperature) a newly proposed CO₂ hydrate formation kinetic model was trained using a supervised ML-based algorithm to predict best-optimized parameters having the capability of forecasting CO₂ hydrate formation kinetics with %AARD of < 9%. This study’s experimental and modelling results offer valuable insights that can play a pivotal role in advancing the CCS via hydrate technologies.

Acknowledgments

PL acknowledges the funding support from the Agency of Science, Technology and Research (A*STAR) under the low carbon energy research (LCER) funding initiatives (project ID: U2102d2010).