The Benefits of Using a Polarity-Swing-Assisted- Regeneration (PSAR) On CO2BOLs: An Experimental and Theoretical Study
- Type: Conference Presentation
- Conference Type:
AIChE Annual Meeting
- Presentation Date:
November 4, 2013
- Skill Level:
CO2-Binding Organic Liquids (CO2BOLs) are a class of second-generation non-aqueous chemically selective CO2-separating solvents. CO2BOLs are unique in that in the absence of a co-solvent they are the only solvents that remain a liquid and in the process undergo dramatic changes in polarity with and without CO2. The solvent’s polarity is dictated by the CO2 loading of the CO2BOL , and conversely we show here that solvent polarity can control the loading of CO2. We describe here the application of this unique polarity shift to assist in the release of CO2 from our CO2BOL solvents. This polarity-swing-assisted regeneration (PSAR) is conceptually simple , where the addition of a controlled amount of chemically inert non-polar “anti-solvent” is added to the CO2-rich CO2BOL with mild heating (>75 ˚C). Once the CO2 is released , the antisolvent is separated from the now CO2-lean CO2BOL by cooling below the solution’s miscibility temperature. The PSAR was found to benefit the CO2BOL solvents , notably decreasing the reboiler temperatures by as much as 72 ˚C , and as a consequence , reduce thermal and evaporative losses of the CO2BOL solvent. The key benefit of a lower reboiler temperature (in anhydrous solvents) is the potential for efficiency gains from the power plant’s steam cycle. Preliminary ASPEN plus simulations estimate the addition of the PSAR projects a 37% lower parasitic load compared to DOE’s case 10 baseline , with the possibility of 43% when maximizing the polarity assist at full strength. Alternatively , the higher thermal compression of CO2 can be achieved with the PSAR than conventional thermal regeneration at comparable temperatures; CO2 pressures as high as 6 ATM have been observed at 100 ˚C. We present thermodynamic evidence for the PSAR and its effect on VLE and LLE data , followed by ASPEN plus modeling of process configurations and implications. We conclude with projections of net power outputs and parasitic loads of the process as a function of PSAR strength.