Computational fluid dynamics modelling of benzene abatement using cryogenic condensation

Abstract

This thesis presents a computational fluid dynamics model of aerosol nucleation and growth using a Eulerian-Lagrangian approach. The research aimed to assess the applicability of cryogenic condensation to controlling benzene emissions from an industrial process operated by the industrial research sponsors. Cryogenic condensation is an attractive option for controlling vent emissions of volatile organic compounds (VOCs). In speciality chemicals industries such as pharmaceuticals, nitrogen is often used to create an inert atmosphere in vessel headspace. Cryogenic condensation can utilise the cooling potential of existing nitrogen infrastructure, making the process energy efficient in comparison to conventional alternatives such as adsorption and thermal oxidation. However, many pollutants freeze or desublimate at the low temperatures (ca. -100°C) used in cryogenic condensation. For these high melting point VOCs, a fine particulate could form under the temperature gradients inside the condenser. Through modelling the process, the research aimed to answer two main questions: will cryogenic condensation control benzene vapour emissions down to the limits set by the environmental regulators; and will it reach this limit without generating a benzene aerosol particulate that becomes entrained in the outlet gas. The research found that the cryogenic condensation alone would not reach the strict emissions limit required by the regulation, and that particle entrainment does make a contribution to this. The model showed roughly 97% of benzene is captured (compared to 99.978% removal that would be required to meet emissions limits) with around 1% escaping as particulate. This information is useful to the industrial sponsors of the research, and other industries considering using cryogenic condensation for benzene abatement. The modelling approach used is a novel contribution to the field with wider potential applications in other areas.