470127 Integration of Sustainability Assessment in the Early Design Phases of Amine Based Post-Combustion CO2 Capture
Monoethanolamine (MEA) is currently the industry standard for amine solvents in PCC applications. However, there is increasing pressure to develop alternative solvents that avoid some of the drawbacks encountered in the MEA process without posing additional health and environmental risks. A new solvent would have to reduce the energy penalty associated mainly with solvent regeneration as well as circumvent problems caused by solvent degradation while maintaining a high reactivity and capacity for CO2 absorption. Progress in development of more effective solvents and processes has been slow due to the need for extensive experimental research and pilot plant campaigns. There is a vast number of combinations of potential solvent candidates, including solvent blends, and process configurations; the experimental evaluation of each option is expensive and time-consuming.
To help speed up this effort, computer-aided design methods are used to guide the search for improved solvents and process schemes as well as to focus the experimental effort. The early integration of sustainability assessment in the process and solvent design is crucial to avoid costly choices and impacts that will be difficult to mitigate in later design and operation stages. Sustainability assessment is defined here as a combined hazard and life cycle assessment. The hazard assessment evaluates potential damage in accidental scenarios, while the life cycle assessment presents the environmental impact of normal process operation. A rigorous hazard and life cycle assessment is first conducted for the well-studied MEA process and compared to amine-based CO2 capture processes that have been proposed in literature. The findings of this assessment are used to guide the development of a robust sustainability assessment tool to be integrated into computer aided molecular design (CAMD) applications for initial solvent screening.
Typical solvent design applications use properties such as solubility parameters, pKa, vapour pressure, density and viscosity among others as proxy indicators for the technical feasibility of a solvent and its performance in a capture process. Such properties could represent a solvent’s capacity for CO2 capture, expected fugitive losses, pump work and heating loads for solvent regeneration. All these factors are also important as indicators for the life cycle assessment. QSAR models for estimating physical properties are largely available. The hazard assessment requires the estimation of additional properties such as solvent flammability, toxicity and persistence in the environment. In a smaller extent, the use of shortcut group contribution methods also allows for the integration of the impacts of solvent production to the life cycle assessment. Severe data gaps however exist for the prediction of a solvent’s potential to degrade and the impact of solvent degradation. Solvent degradation can lead to many operational problems such as reduced heat transfer efficiency, foaming, corrosion and the formation of hazardous degradation products in addition to loss of CO2 absorption capacity. Therefore, degradation potential is an important indicator for a solvent’s process, hazard and life cycle performance and should be included as early as possible.
A database of hazard and environmentally related properties is created comprising 80,000 chemical structures. Physical and chemical properties are collected mainly from available material safety datasheets. This database facilitates the automation of the online connection of the CAMD optimization routine and the complimentary sustainability assessment module. The connection can be established where an algorithm combines the proposed molecular groups into all potential structure formations and runs the proposed structures through the internal database for hazard related properties. If all proposed molecules are found the mean of the properties is taken, otherwise QSAR models are used wherever possible to predict the missing properties Experimental results are used in combination with literature data to develop a QSAR model for predicting solvent degradation rates. The model is developed using Principal Component Regression (PCR) and Multi Non-Linear Regression (MNLR) to correlate degradation rates with molecular structure, thermodynamic properties and operating conditions.
In conclusion, this work introduces a framework for the integration of the sustainability assessment at early stages of solvent and process design for post combustion CO2 capture. It shows the translation of rigorous hazard and life cycle assessment into indicators that can be easily integrated in a CAMD routine. It presents a database of hazard and environmental properties, which facilitates the online connection to an optimization routine. Finally, it proposes a QSAR model for the determination of solvent degradation rates under the PCC conditions, which enables the addition of an important and currently missing solvent feature at molecular design stages for a more comprehensive assessment.