(274a) Efficient Selection of Conventional and Phase-Change CO2 Capture Solvents Based on Nominal and Off-Design Process Operation

Solvent-based absorption/desorption is a promising and widely investigated technology for post-combustion CO₂ capture. Research efforts largely focus on the identification and selection of solvents that significantly reduce the regeneration energy requirements; the latter is the main reason for the very slow commercialization of the technology. Solvents are often evaluated and selected based on properties such as heat of absorption and loading achieved at absorption conditions. This property-based approach is useful, but it only provides indirect links to the expected solvent performance in the process where it is utilized. On the other hand, process-based solvent evaluation is a desirable approach in order to obtain a realistic evaluation of the economic and operating solvent performance. However, it is mostly performed through rigorous process and flowsheet models which provide a very detailed account of phenomena and process characteristics. The very non-ideal behavior of solvent-water-CO₂ mixtures makes the development of such reactive separation process models difficult, while their use is not without significant computational challenges. For example, numerical problems may appear due to discontinuities resulting from the appearance or disappearance of phases within a simulation. While such issues appear during simulations addressing solvent evaluation at nominal process conditions, they are amplified when off-design operation is considered to evaluate solvents at conditions other than the nominal design settings. The latter is very important for solvent evaluation because processes are essentially dynamic environments, susceptible to disturbances. A solvent that is economically desirable at nominal operation, may exhibit very high sensitivity to disturbances, resulting in the need for very expensive control approaches to maintain operation within desired set-points.

To attain process -based evaluation without numerical difficulties it is possible to use models of lower fidelity, such as shortcut models, which have been proposed in few occasions (e.g. Notz et al., 2011; Kim et al., 2015) as a means of fast, process-based solvent screening. Models such as the one proposed by Notz et al. (2011) have been shown to sufficiently capture the vapour-liquid equilibrium behavior of standard solvents (like Monoethanolamine). As they are Kremser-type models, linear approximations are required for the very non-linear equilibrium behavior of solvent-water-CO₂ mixtures which may be under- or over-estimated, while such approximations are difficult to derive. On the other hand, the model of Kim et al. (2015) focuses on the intermediate heat exchange and desorption part of the process, considering a stripper, while accounting entirely for the non-ideal behavior of the solvent in the presence of water and CO₂. The absorption operation is represented by accounting for equilibrium, whereas studying the stripper behavior permits the calculation of reboiler duty using vapor-liquid equilibrium data and energy balances around it. The approach further allows the calculation of flows, lean and rich loadings and temperatures at various points of the flowsheet, while it is only demonstrated for Monoethanolamine and Piperazine.

In this work, we adopt the model of Kim et al. (2015) and we extend and exploit it in various ways. We extend its use to additional, conventional amine solvents including solvent mixtures. We further extend the model itself to account for a new class of solvents, namely phase-change solvents. These are amines that exhibit liquid-liquid phase separation upon a change in processing conditions, e.g. upon reaction with CO₂ or upon temperature increase after reaction. The liquid-liquid phase-change enables non-thermal separation of a CO₂-lean phase prior to desorption, which is recycled to the absorber. This reduces the flow that enters the desorber, while desorption may take place at much lower temperature than with conventional solvents. As a result, it is possible to achieve very significant energetic reductions. We therefore propose a new model that is able to accommodate calculations for phase-change solvents. The model is used for evaluation of solvent performance at nominal operation, using equilibrium relations derived directly from vapour-liquid-liquid experimental data for selected phase-change solvents.

The proposed model captures directly the non-ideal solvent-water-CO₂ interactions hence it is also used to evaluate solvent performance in the presence of disturbances. This is approached through a systematic non-linear sensitivity analysis method, which investigates the static operability performance of each solvent in the process. It is based on the development of a sensitivity matrix which incorporates the derivatives of multiple process performance measures (e.g. reboiler duty, net energy penalty, cyclic capacity etc.) with respect to multiple operating parameters and solvents. The sensitivity matrix constitutes a measure of the process operating variation under the influence of infinitesimal changes imposed on the selected parameters. It is decomposed into major directions of variability associated with the eigenvectors corresponding to the larger in magnitude eigenvalues of the sensitivity matrix. The eigenvector of the largest eigenvalue represents the dominant direction of variability for the system, causing the largest change in the performance measures. The entries in the dominant eigenvector determine the major direction of variability in the multiparametric space and indicate the impact of each parameter in this direction. Having identified this direction, it is not necessary to explore all directions of variability (i.e. combinations of parameters) arbitrarily hence reducing the dimensionality of the sensitivity analysis problem. The dominant eigenvector direction is then utilised in the exploration of the system behaviour as indicated by the change of key performance indices under simultaneous, multiple and finite parameter variations.

The proposed developments are implemented for 11 solvents and mixtures, including 2 phase-change solvents. The thermodynamic equilibrium relations are derived from data available in the literature or from experiments performed internally. The process performance of few solvents is validated against results from literature sources, showing good agreement in indices like reboiler duty. The solvent-process evaluation criteria include indices such as reboiler duty, net efficiency energy penalty points, cyclic capacity, solvent mass flowrate, solvent purchase cost and lost revenue from parasitic electricity upon integration with power plants. Phase-change solvents and mixtures exhibit a reboiler duty between 2-2.3 GJ/ton CO₂, whereas the best conventional solvent mixture of 2-amino-2-methyl-1-propanol/Piperazine exhibits approximately 3.1 GJ/ton CO₂. The operability assessment highlights that certain economically desirable solvents may not be as attractive under off-design conditions.

Acknowledgements

This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the grant agreement 727503 - ROLINCAP – H2020-LCE-2016-2017/H2020-LCE-2016-RES-CCS-RIA.

Cited References

Kim, H., Hwang, S.J., Lee, K.S., 2015, Novel Shortcut Estimation Method for Regeneration Energy of Amine Solvents in an Absorption-Based Carbon Capture Process, Environmental Science and Technology, 49 (3), 1478–1485.

Notz, R., Tönnies, I., Mangalapally, H.P., Hoch, S., Hasse, H., 2011, A Short-Cut Method for Assessing Absorbents for Post-Combustion Carbon Dioxide Capture, International Journal of Greenhouse Gas Control, 5 (3), 413–421.