CO₂ capture by absorption Experiment and modelling

A realistic strategy for reducing the concentration of carbon dioxide (CO₂) in the atmosphere should rely on various levels of carbon capture, storage, and utilization (CCUS). While the world should gradually shift away from using carbon-intensive processes to produce energy, CO₂ emissions from operational fossil-fuel-fired power plants must be minimized in the short term. Post-combustion absorption is a well-understood, mature process for removing acid gases from exhaust streams that were originated from power generation, steel making, and cement production. However, CCUS projects are still perceived as being capital intensive and highrisk.

Recent progress in CO₂ absorption technology has shown promising results for mixed electrolyte solutions. A Mixed-Salt Process (MSP) has been recently developed and commercialized by SRI International. The technology uses a solvent based on aqueous ammonium salts and potassium carbonate, which are low-cost, off-the-shelf, and chemically stable components. The technology overcomes problems with ammonia slip by introducing a series of process design innovations, such as operating the absorption tower with two distinct solvent compositions. These design choices also resulted in a reduced reboiler energy consumption of 2.0 to 2.3 GJ t⁻¹ CO2, compared to a consumption of nearly 3.6 GJ t⁻¹ CO₂ in monoethanolamine-based technology, which is the typical benchmark for CO₂ capture processes.

The goal of this thesis is to help improve thermodynamic modelling capabilities for designing CO2 absorption processes using mixed electrolyte solvents. Chemical absorption of CO₂ involves complex reactions between the species dissolved in the liquid. A comprehensive modelling of these mixtures’ properties must also be able to estimate the conditions that lead to vapor-liquid(-liquid)-solid equilibrium in these systems. In ammonia and potassium carbonate solutions, for example, there are at least 10 salts that could precipitate based on the reactions between the liquid phase species. The addition of other components into the solvent, either to enhance absorption rate or loading capacity, should increase the complexity of the combined chemical and phase equilibrium description of the system.

The first part of this work was dedicated to measuring the solid-liquid equilibrium of aqueous solutions containing various amounts of ammonia, potassium carbonate, and methyl-diethanolamine. These data will help improve the estimation of solubility limits of carbonate salts that might precipitate during the process. It also contributes to the calculation of water activity, which must be matched with high accuracy to ensure reliable salt solubility estimates.

Following these studies, an experimental procedure was designed and commissioned to measure bubble-point pressure data, which are fundamental to modelling the vapor-liquid equilibrium of these mixtures. The apparatus operates under static-synthetic conditions: precise amounts of each component were added to a reactor (synthetic) and left to reach equilibrium under constant agitation without any recirculation of the coexisting fluid phases (static). This methodology has been previously used to measure the solubility of gases in many different solvents, and it avoids uncertainties associated with sampling and analysis of the phases in equilibrium.

Finally, the experimental results obtained during this project were integrated into the Electrolyte Solutions Databank at CERE. These data were then used to perform the thermodynamic modelling of these mixed electrolyte systems. The Extended UNIQUAC model was used to calculate the excess Gibbs energy of the species in the liquid phase. The model parameters required to model the interactions between the solvent components were estimated based on more than 2000 experimental data on various physical and equilibrium properties, such as vaporliquid equilibrium, pure component saturation pressure, excess enthalpy, heat of absorption, molar heat capacity, apparent molar heat capacity, and solid-liquid equilibrium. The resulting model was validated in a temperature range between 253.6 K to 600 K and pressures up to 75.6 bar.