Social Science Research Network | 2021

Development of a rate-based model for CO2 capture using a non-aqueous hydrophobic solvent

 
 
 
 
 
 
 
 
 
 
 

Abstract


Advanced water-lean solvents offer several advantages over aqueous amine solvents for post-combustion CO2 capture, particularly lower parasitic energy penalty, lower regeneration temperatures and lower corrosion. RTI and SINTEF have been working on developing eCO2SolTM, a non-aqueous solvent (NAS) that has shown stripper heat duties of 2.1 - 2.3 GJ/t-CO2 at pilot-scale testing at a 6-kW large bench-scale gas absorption system (BsGAS, RTI International, USA) and at the 60-kW Pilot testing unit (Tiller Plant, SINTEF, Norway). This technology will be demonstrated at the 12 MW scale at Technology Centre, Mongstad (TCM), Norway in early 2022. Development of novel solvent formulations is a time-consuming step, and models are needed to evaluate different process configurations and process conditions, to explore different energy integration strategies to minimize net heat duty and to size equipment and evaluate overall process economics. The non-aqueous solvent formulation includes novel components that have not been studied extensively in the literature. Most of the properties that are needed to develop rate-based kinetic-, thermodynamic and hydrodynamic models are not available and must be generated. This work provides details on the development of a detailed process model including the experimental measurement of properties required for the simulation model. The process model has been tested against quality pilot data gathered from the two pilot plants mentioned. \nThe average absolute deviation (\uf044AAD%) between model prediction and experimental data for the specific reboiler duty (SRD) is within 6% for the 6-kW BsGAS at RTI and 7% for 60-kW pilot testing at SINTEF. Mass and heat transfer of the developed rate-based model was validated by predicting the temperature, CO2 loadings (molCO2/molamine) and CO2 concentration profiles of the testing units and compared favourably to experimental data.

Volume None
Pages None
DOI 10.2139/SSRN.3812568
Language English
Journal Social Science Research Network

Full Text