(706e) Modelling and Experimentation on a Rotary Adsorber Applied to Direct Air Capture

Direct Air Capture (DAC) is a promising technology to aid the removal of CO₂ from the atmosphere and achieve net-negative emissions. However, it produces unique challenges due to the ultra-dilute nature of the ambient air feed (0.04 mol%). For comparison, CO₂ composition in post combustion gas streams ranges from 4-25 mol% [1] which is 100 to 600 times more concentrated than CO₂ in air. Therefore, capturing 1 kg of CO₂ in a DAC process requires at least 1,300 m³ of air compared to less than 13 m³ of post-combustion gas (at standard temperature and pressure). Although DAC proves more challenging, unlike post-combustion carbon capture, it can achieve net-negative emissions as well as offsets for non-point source emissions such as vehicles and planes. The requirements of high volumetric air flow and net-negative emissions dictate that feasible DAC separation processes must have minimal pressure drop and low specific energy. Today, research focused on process and contactor design for DAC in the field of adsorption remains limited yet crucial for the advancement of adsorption-based DAC. Beyond process design, the development of process models and experiments allows testing of promising sorbent materials for adsorption-based DAC.

In this work, we aim to assess the design and optimisation of a rotary adsorber by applying mathematical modelling and experimental validation. The rotary adsorber is a continually rotating thin cylindrical bed with a separated adsorption and desorption section, allowing the unit to both adsorb and desorb simultaneously. The pancake style design minimises pressure drop, while simultaneous adsorption and desorption allows a continuous process with one vessel. A rotary adsorber has been considered through equilibrium modelling for post-combustion carbon capture [2]. However, in the case of DAC, a kinetic model is necessary to account for the poor adsorption kinetics in ultra-dilute conditions and the dynamic nature of a rotating process contactor. We have considered temperature swing adsorption (TSA) operation, as adequate seals for pressure or vacuum swing cannot be achieved in the rotary unit. Initially TSA with an air purge has been applied, however, this will be extended to also assess steam and inert purge TSA.

To assess the rotary adsorber, we have created a one-dimensional dynamic adsorption model that describes the spatial and temporal evolution of gas and adsorbed phase concentration profiles in the unit. Initial results demonstrate that the model can assess the impact of the wheel speed, process flow rates, desorption conditions, material properties, adsorption kinetics, division of the wheel, and wheel dimensions. The model can be used to calculate key process performance indicators including CO₂ productivity, purity, working capacity and specific energy. We have commenced the design of a laboratory scale rotary unit connected to a continuous gas analyser for model validation. We aim to extend the experimental validation to assessing the impact of optimisation parameters through adjusting the operating conditions and layout of the rotary unit. The initial results highlight that that design and optimisation of a rotary adsorber is complicated due to the interconnected nature and significant number of parameters - ultimately requiring multi-variable optimisation.

References:

1. 1. E.S. Sanz-Perez et al. Chem rev. 5 (2020) 11840-11876
2. L. Herraiz at al., Front. Energy Res. 8 (2020) 482708