(654f) Techno-Economic Analysis for CO2 Capture and Storage Processes With Concentrated PZ and Mdea/PZ Blends

Authors: 
Frailie, P. T. II, The University of Texas at Austin
Sachde, D. J., The University of Texas at Austin
Madan, T., The University of Texas at Austin
Rochelle, G. T., The University of Texas at Austin



This study proposes a method for quantifying the relationships between capital and operating costs for retrofit alkanolamine scrubbing systems for the removal of CO2 from the flue gas of coal-fired power plants.  This particular application was chosen because of its potential to significantly reduce CO2 emissions with relatively little energy penalty.  Coal-fired power plants generate forty-five percent of the electricity in the United States and about seventy-seven percent of the CO2 from the power sector, and the relatively high concentration of CO2 reduces the energy required to capture 90 % of the greenhouse gas emissions.  Concentrated piperazine (PZ) and methyldiethanolamine (MDEA)/PZ blends were chosen because they exhibit favorable rates and capacities over operationally significant conditions.  Using an Aspen Plus® thermodynamic, hydraulic, and kinetic model developed at The University of Texas at Austin, a base-case absorption/stripping/compression process was designed, tested, and optimized to minimize the cost of CO2capture and storage.

The Aspen Plus® thermodynamic model was developed using a sequential regression methodology to regress CO2 solubility, heat capacity (loaded and unloaded), CO2 activity coefficient, amine volatility (loaded and unloaded), amine pKa, and heat of absorption over operationally significant temperature, amine concentration, and loading ranges for MDEA, PZ, and MDEA/PZ systems using the electrolyte-NRTL method.  Hydraulic data was directly incorporated into Aspen Plus® using FORTRAN subroutines that calculate viscosity, density, and binary diffusivities for amine-derived species as a function of process parameters such as temperature, amine concentration, and CO2 loading. CO2 absorption rate data was incorporated by adjusting rate constants and binary diffusivities in an Aspen Plus® wetted wall column (WWC) simulation and a Microsoft® Excel regression to fit experimental data at various temperatures and loadings.  The final kinetics model predicts the CO2 absorption rate for 5 m PZ, 8 m PZ, 7 m MDEA/2 m PZ, and 5 m MDEA/5 m PZ from 10-100 oC over operationally significant loading ranges with an average error of 4.9 %.

The techno-economic analysis focused on four major cost centers: (1) absorber, (2) cross-exchangers, (3) reboiler, and (4) compressor.  These four unit operations can constitute as much as 85 % of the capital cost, and, thus, present the best opportunity for reducing the cost of CO2 capture.  Reducing the cost of one unit operation typically results in increasing the cost of another.  For example, increasing the liquid flow rate will decrease the required packing volume (lower CAPEX), but it will increase the size of the cross-exchangers (higher CAPEX) and the reboiler duty (higher OPEX).  It is possible that the savings in the absorber size will exceed the losses in cross-exchanger size and reboiler duty, but this cannot be quantified without a rigorous techno-economic analysis.  Expressions have been developed for the four aforementioned unit operations that calculate CAPEX and OPEX in the units of US dollars per metric ton of CO2 using Aspen Plus® predictions for stream properties.  These expressions are based on price quotes obtained from manufacturers, previous techno-economic analyses for alkanolamine CO2 capture processes, and Aspen Plus® ICARUS.  CAPEX predictions were converted to the units of US dollars per metric ton of CO2 using a factor that accounts for both the conversion of purchased equipment cost to total capital requirement and the annualizing of the initial investment over the lifetime of the project.  This factor was derived using pricing guidelines from non-specific estimation methods such as chemical engineering design texts, as well as CO2-capture specific guidelines from the Department of Energy and the IEAGHG. 

The base-case absorber design will employ either a pump-around or an in-and-out intercooler.  Pump-around intercooling effectively splits the column into three sections: (1) a top section into which lean solvent enters and scrubbed gas leaves, (2) a middle section containing 2-5 times more solvent than the top section, and (3) a bottom section containing the same amount of liquid as the top section from which the rich solution exits and the flue gas enters.  In-and-out intercooling simply removes all of the solvent from the top section, cools it to 40 oC, and feeds it to the top of the bottom section.  There are tradeoffs associated with lean loading, feed liquid flowrate, and intercooling configuration.  Decreasing lean loading increases capacity and decreases absorber height but increases reboiler duty.  Increasing the feed liquid flowrate from 1.1 to 1.2 times the minimum liquid flowrate decreases the capacity of the solvent by 8 to 11 % but decreases the amount of packing needed to remove 90 % of the CO2 by 25 to 30 %.  The base-case stripper contains both a cold rich bypass (CRB) and warm rich bypass (WRB) stream.  A fraction of the cold rich solvent exiting the bottom of the stripper is heated by the product gas before being fed into the top of the stripper.  The remaining rich solvent is heated by a warm solution from the bottom of the stripper in a cross exchanger.  Another portion of the warm rich solvent is bypassed and fed directly into the top of the stripper.  The remaining rich solution is heated by a steam heater and flashed into the bottom of the column.  

After developing the pricing method and designing a base-case configuration, operating conditions such as lean loading, solvent flow rate, and intercooling configuration were varied to minimize the cost of CO2 capture.  For each condition the percent bypass to the CRB and WRB was optimized to minimize the equivalent work, absorber section heights were adjusted to minimize packing area, heat-exchanger sizes were optimized for heat transfer area and pressure drop, and column diameters were adjusted to achieve a maximum of 70 % flood.  This analysis was performed for 8 m PZ, 7 m MDEA/2 m PZ, and 5 m MDEA/5 m PZ.