Search

In this research, the volume of fluid (VOF) method is used to study the hydrodynamics of rotating packed beds (RPBs). The model is validated, and grid independence analyses are performed for cases with different operating conditions. The droplet size distribution is investigated to characterize the hydrodynamics of RPBs. Droplet size distributions are compared in two-dimensional and three-dimensional simulations, and it is demonstrated that two-dimensional simulations can provide an accurate prediction while significantly reducing the significant computational cost. Radial distributions of droplet diameter in the packing region are studied, and different trends are observed at different rotational speeds (fluctuating at ω = 250 rpm, increasing–constant at ω = 500 rpm, and decreasing at higher rotational speeds). These trends are explained using the breakup and coalescence of droplets during droplet–packing and droplet–droplet collisions. Breakup, coalescence, and deposition regimes of droplets depend on the Weber, Ohnesorge, and impact parameters. We observed that with increasing rotational speed, the average droplet diameter and its standard deviation decreased, while changing the liquid flow rate did not significantly affect the average droplet diameter. It is also observed that there is a critical rotational speed (depending on the bed configuration), beyond which the average droplet size does not decrease with increasing rotational speed.

An apparatus consisting of an equilibrium cell connected to two vibrating tube densimeters and two syringe pumps was used to measure the saturated phase densities of (CO2 + H2O) at temperatures from (293 to 450) K and pressures up to 64 MPa, with estimated average standard uncertainties of 1.5 kg · m−3 for the CO2-rich phase and 1.0 kg · m−3 for the aqueous phase. The densimeters were housed in the same thermostat as the equilibrium cell and were calibrated in situ using pure water, CO2 and helium. Following mixing, samples of each saturated phase were displaced sequentially at constant pressure from the equilibrium cell into the vibrating tube densimeters connected to the top (CO2-rich phase) and bottom (aqueous phase) of the cell. The aqueous phase densities are predicted to within 3 kg · m−3 using empirical models for the phase compositions and partial molar volumes of each component. However, a recently developed multi-parameter equation of state (EOS) for this binary mixture, Gernert and Span [32] , was found to under predict the measured aqueous phase density by up to 13 kg · m−3. The density of the CO2-rich phase was always within about 8 kg · m−3 of the density for pure CO2 at the same pressure and temperature; the differences were most positive near the critical density, and became negative at temperatures above about 373 K and pressures below about 10 MPa. For this phase, the multi-parameter EOS of Gernert and Span describes the measured densities to within 5 kg · m−3, whereas the computationally-efficient cubic EOS model of Spycher and Pruess (2010), commonly used in simulations of subsurface CO2 sequestration, deviates from the experimental data by a maximum of about 8 kg · m−3.