Amine and amino acid absorbents for CO₂ capture
MetadataVis full innførsel
World energy demand is on a continuous increase and fossil fuels will remain the dominant energy source for the foreseeable future. Fossil fuel combustion, however, results in greenhouse gas emissions, and in particular of large quantities of CO2. This contributes significantly to global warming and the subsequent adverse health and environmental effects. CO2 capture by absorption remains the most mature technology to reduce global anthropogenic CO2 emissions. The cost of this technology and the potential secondary environmental impacts remain some of the issues that need to be reduced to facilitate full scale global deployment in industrial processes and power plants. These issues are directly linked to the solvent used in the process. Amines are the most used and studied absorbents and significant progress has been made during the last 10-15 years, in particular with regard to heat needed for the CO2 capture regeneration, reducing this from typical figures of 3.8-4.2 GJ/ton CO2 captured to today 2.6-2.8 GJ/ton CO2. This is, however, still high, and combined with pressing environmental issues like e.g. degradation and process and atmospheric nitrosamine and nitramines formation, there is a definite need for developing better amines or other new absorbents to improve both efficiency and environmental characteristics. The work in this thesis focuses on finding and developing new absorbents for CO2 capture. It started by an evaluation of different solvent characterization factors. These are often related and also work against each other in many cases. The work then proceeded with a solvent preselection based on theoretical studies using structure-pKa relationships where solvents with similar structures were arranged according to their pKa values. Relationships were found and explained by electron donation and withdrawal through bonding and resulting solvation effects. Further screening experiments were conducted on selected amine absorbents and a method of combined absorption-desorption analyses was developed for selecting the most promising solvents for CO2 capture based on the rapid screening experiments. It pointed at TMBPA to be good amine candidate solvent and that when TMBPA was promoted with PZ it showed a good potential to reduce the costs of the CO2 capture process. This work, however, emphasizes the environmental impact of the new solvents. Thus, also amino acid salts were screened and investigated for CO2 absorption potentials. In addition, a new absorbent class, amine amino acid salt solutions, was investigated. The amino acid absorbents are benign solvents and they have highly reduced volatility but often do not give the reduction in regeneration heat needed. It was found, however, that the new absorbent class, the amine amino acid salt solution, could have the potential of combining benign solvents, reduced volatility and highly reduced regeneration heat demand, thus providing an overall superior performance for CO2 absorption than both the conventional amines and amino acid salt systems. For the solvents to be of industrial use they must be characterized. The physicochemical properties, density, viscosity and N2O solubility of both the amino acid salt, KSAR and the amine amino acid salt, SARMAPA, necessary for kinetics and mass transfer studies, were measured. The amino acid absorbents were found to have very low physical solubility of CO2. This was attributed to the salting out effect of the ionic amino acid absorbent. Several different models were tested and developed to represent the experimental results and the agreement between model and experimental results were in general good. Kinetic and mass transfer properties of the absorbent necessary for the absorber design and process modeling were measured using a string of discs contactor, SDC, apparatus. The results obtained were used to re-interpret the termolecular and zwitterion mechanisms by introducing an ionic correction factor to take into account the effect of ionic strength on the pseudo first order kinetic rate constant of an aqueous amino acid absorbent. The reaction rate constant for the amino acid salt was found to be high but still comparable to the amines. The rate constant for amine amino acid salt, SARMAPA, was found to be lower than for KSAR but higher than for MEA. KSAR had the highest overall mass transfer coefficient, KGov, followed by MEA and then SARMAPA. The developed kinetic model for KSAR, when combined with its vapor liquid equilibrium model, adequately predicted both laboratory SDC and pilot plant mass transfer results well. Vapor liquid equilibrium for the unloaded and loaded amino acid absorbents, KSAR and SARMAPA, as well as for MEA was measured using two low and high temperature VLE apparatuses. VLE for the H2O-MEA-CO2 system were measured for four different MEA concentrations: 15, 30, 45, and 60wt% and for the temperature range 40-120oC. The extended UNIQUAC thermodynamic framework was developed and applied to calculate the nonidealities in the systems represented by the activity coefficients of the species. Non-idealities in the gas phase were calculated using the Soave-Redlich-Kwong equation of state. Measurements with different MEA concentrations enabled the detailed study of the effect of concentration on the CO2 partial pressure. The physical solubility of CO2 in aqueous MEA solutions was determined using the N2O analogy and was applied in calculating the true activity coefficients of CO2. Thermodynamic model for the amino acid salt, KSAR and a simplified one for amine amino acid salt, SARMAPA were developed to predict speciation and VLE. The amino acid salt, KSAR, has a highly reduced volatility, and the volatility of the amine in the amine amino acid salt, SARMAPA, was reduced by 50 %. The behavior of the amino acid absorbents was studied further in two laboratory pilot plant campaigns. The results confirmed that the mass transfer rates of the amino acid salt, 3.5M KSAR, were higher than those of 30wt % MEA and 5M SARMAPA. However, the regeneration energy requirement was found to be much higher than for both SARMAPA and 30wt% MEA. SARMAPA was found to exhibit a much lower desorption energy requirement than KSAR, and the stripping stream requirement for the SARMAPA system was about half the values for MEA. The amino acid absorbents, KSAR and SARMAPA, showed practically no corrosive effects on process lines during the tests. Amino acid absorbents studied in this work practically reduces or/eliminates volatility problems found in amine systems and the amine amino acid salt studied, SARMAPA, has shown that the amine amino acid salts may be good candidates to reduce the energy requirements of the CO2 capture processes combined with minimal environmental impact.