Boosting CO₂ capture efficiency of the exhausted RFCC flue gas by using intercooler exchangers: Leveraging ANN in MDEA-based approach

This study investigates the simultaneous capture of carbon dioxide (CO₂) and sulfur dioxide (SO₂) from residue fluid catalytic cracking (RFCC) flue gas using Methyldiethanolamine (MDEA) as an absorbent in a post-combustion capture system. The proposed system offers an effective solution for refineries aiming to reduce greenhouse gas emissions and comply with environmental regulations. The system captures approximately 97% of CO₂ and completely removes SO₂ from the RFCC flue gas. The integration of two inter-coolers significantly enhanced CO₂ capture efficiency by dissipating the heat generated during absorption, resulting in an 82% reduction in CO₂ emissions compared to systems without inter-coolers. A comprehensive analysis of absorbent operating parameters - including MDEA flow rate (1100-1300m³/h), temperature (40-50 °C), concentration (20-30wt%), and absorption pressure (25-28bar) - revealed that increasing all factors except temperature improved CO₂ capture performance. Notably, MDEA achieved complete SO₂ absorption under all tested conditions. An artificial neural network (ANN) model was developed to predict CO₂ emissions accurately, enabling real-time process control. The model demonstrated excellent performance, with an R²value of 0.9974 and a mean absolute error (MAE) of 0.0045 on the test dataset, indicating that operational conditions can reliably predict CO₂ emissions. This study contributes to enhancing the efficiency of practical post-combustion CO₂ capture systems.