CO2 capture solvents; modeling and experimental characterization
MetadataVis full innførsel
With more than 80% of the world’s energy consumption coming from the burning of fossil fuel (BP, 2013; IEA, 2013), the mitigation of greenhouse-gas from the energy sector is crucial for avoiding climate change. Amine scrubbing is the most mature and well-understood process for post-combustion CO2 capture (Rochelle, 2009; Svendsen et al., 2011). In amine scrubbing, the regeneration of the solvent accounts for the most energy demanding part of the process (Aroonwilas & Veawab, 2007; Van Wagener & Rochelle, 2011). In a coal-fired power plant, the energy requirement for the stripper reboiler is calculated to be 15-30% of the net power production (Jassim & Rochelle, 2006). The energy demand of the process is directly correlated to the chosen solvent. Therefore, the development of new solvents or solvent blends has gained much attention during the past 15 years. The aqueous mixture of 5M DEEA/2M MAPA is a promising solvent for CO2 post combustion capture. This solvent has a particular characteristic as it forms two liquid phases upon CO2 loading. In this case the formation of two phases is desired as having one lean and one very CO2-rich phase (upper and lower phase, respectively) gives a potential for lowering the liquid circulation rate, thereby lowering the sensible heat needed for increasing the CO2 pressure in the desorber and thereby lower the stripping steam requirement, and finally, by using a mixture of a tertiary and primary/secondary amine the operative heat of absorption can be lowered thus lowering the heat needed for reversion of the absorption reaction. Moreover, the primary/secondary amine will enhance the capture rate while the tertiary amine will raise the capture capacity of the solvent. In addition, because of the very rich solution sent to regeneration, one may produce the CO2 at elevated pressures. In this thesis several experiments has been carried out in the DEEA/MAPA system in order to characterize this solvent. The screening experiments showed the potential of this solvent for CO2 capture. VLE experiments were carried out in the single amine systems both with and without CO2. In addition, density and viscosity measurements were performed. These experiments are described in Chapter 2 of this thesis. The VLE data were used to develop thermodynamic models. The eNRTL model was chosen to model the liquid phase while the Peng-Robinson EoS was used in the vapour phase modelling. The particle swarm optimization (PSO) algorithm was used to fit the eNRTL parameters. For the simulations of the pilot plant, a simplified soft model was used since very few data is available for the DEEA/MAPA system. The data from the screening experiments were used to fit the parameters of the soft model. An attempt to describe the oxidative degradation of MEA by a rigorous model is shown. The model is able to represent the development in time of the concentrations of most of the compounds. However, this model is only an initial step of the solvent degradation modelling and substantial work, both on the experimental and modelling side, is remaining in order to satisfactorily define a model for the oxidative degradation. The DEEA/MAPA system was tested in the Gløshaugen (NTNU/SINTEF) pilot plant. The solvent was shown to be relatively easy to operate within the pilot plant. Neither high viscosity nor foaming issues were noted. Absorption was reasonably fast and the CO2 stripping in this system was easily done due to the presence of a tertiary amine. Moreover, the system was able to operate at significantly lower specific reboiler duty and reboiler temperatures compared to the benchmark 30 wt.% MEA. The solvent was, however, more volatile than MEA and better control to avoid solvent losses is required in a real plant. The experimental pilot campaign for the DEEA/MAPA system was validated towards simulation models implemented in CO2SIM. A total of 18 runs from the campaign were validated and simulated with CO2SIM. The model of the DEEA/MAPA system was successfully implemented in CO2SIM and can be used for further simulations of large scale plants.