Hydrodynamic multiplicity in the trickle flow, or low interaction, regime is a well documented phenomenon. Multiple hydrodynamic states are often presented in the form of hysteresis loops where the hydrodynamic parameter studied are shown as a function of the operating history of the bed, i.e. liquid and gas flow rates. In extreme cases the lower leg, representing an increase in liquid flow rate on a pre-wetted and drained bed, is commonly referred to as the Levec mode. The upper extreme, referred to as the Kan-liquid mode, represents a decrease in liquid flow rate after operation in the high interaction regime. The many reported studies investigating liquid-solid mass transfer in trickle beds have generally used either the dissolution or electrochemical techniques. Numerous researchers have used their data to develop correlations predicting solid-liquid mass transfer coefficients. Most of these studies do not specify the multiplicity mode of operation. Only two studies (Sims et al. (1993) and Van der Merwe, Nicol&Al-Dahhan (2008)) use both the Levec and Kan-liquid operating modes. Both of these studies suggest that solid-liquid mass transfer also exhibit multiplicity behaviour although the trends suggested or speculated differ from each other. Sims et al. (1993) found that a Kan-liquid operated bed will outperform a Levec operated bed; however in contrast to this Van der Merwe et al. (2008) speculated that a Levec operated bed is better suited for liquid limited reactions due to enhanced liquid-solid mass transfer in the Levec mode as a result of faster interstitial velocity. This study showed that solid-liquid mass transfer coefficients, measured with both the dissolution and electrochemical technique, show multiplicity behaviour. Two distinct operating regions were found, which corresponds to the Levec and Kan-liquid modes. Measurements taken using the electrochemical technique yielded solid-liquid mass transfer coefficients larger than those measured using the dissolution method. The experimental results agree with the trend found by Sims et al. (1993) but the mass transfer coefficients in this study were significantly lower. Additionally it was shown that the difference in mass transfer coefficients, in the two modes, cannot be explained by merely compensating for the differences in wetting efficiency and interstitial velocity, suggesting that the Levec mode has a larger percentage of stagnant or poorly irrigated zones. It was also shown that mass transfer coefficients measured at the top of the column is higher than those measured at the bottom, suggesting that the flow structure is changing as a function of axial length. Lastly, with regards to electrochemical measurements of liquid-solid mass transfer, it was shown that measurements using a single particle electrode compared well to that of a multiple packing electrode.
Dissertation (MEng)--University of Pretoria, 2009.