Experiments conducted in 0.15 m diameter bubble columns using water and non‐aqueous liquids have shown that the gas velocity at which transition from the bubbly flow to the churn‐turbulent flow regime occurs is a function of gas density. The transition velocity increased with increasing gas density. The direct effect of gas density on gas holdup in the bubbly flow regime is small with only a slight increase in holdup being observed at higher densities (ϵG α ρg 0.04). In the churnturbulent region a much greater effect of gas density on gas holdup was observed. These differences were found to be a direct function of the differences in holdup values at the transition points. Gas holdup was found to be a function of the gas phase momentum. In the bubbly flow regime holdup was directly proportional to momentum while in the churn‐turbulent regime holdup was proportional to momentum to the one third power. Reasons for this behaviour are discussed, as well as the implied effects on liquid mixing in bubble column slurry reactors. The effects of gas density may offer an explanation for some apparently anomalous published results.

A review of existing literature for the purpose of comparing correlations which allow Lhe explicit predictim @ gas holdup in twoor three-phase bubble columns shows wide disagreement concerning the eTfegt. of various variables. Further, various correlations, apparently of equal validity, show equally wide disagreement in the Fiadicted absolute values of gas holdup among three test systems; water-air, hydrocarbon solvent-air, and trichloroethyrene air. Some possible reasons for these disagreements are discussed.Extensive gas holdup data obtained in a 0.30 m diameter column for the three test systems were correlated statisticany for the transition and the turbulent region. The resulting correlation gives agreement with experiment to within *20%, with most data showing * 10% agreement. The proposed correlation was also tested against thirteen data sets (from eight researchers) which corresponded to its region of validity and gave good agreement. Three additional relevant data sets were not satisfactorily predicted.On a prockdk Ir une ktude bibliographique dans le but de comparer des corrklations permettant de prkdire explicitement la ktention des gaz dans des colonnes a bulles deux ou trois phases et on a montrk que ces codlations ne s'accordent pas sur I'effet des diffkrentes variables. De plus, diverses codlations, apparernment de mkme validitk, montrent kgalement un dksaccord dans les valeurs absolues prkdites pour la ktention des gaz dans le cas de trois systbmes-tests: eau -air, solvant a base d'hydrocarburesair et trichlokthylbneair. On analyse quelques raisons pouvant expliquer I . Air-Water Air-Glycol aq. soh. Air -Methanol 02-Water 2. Air-Alcohols Air-Halogenated Hydrocarbons 3. Air-Water Air-Glycerol, aq. soln. Air-Kerosene 4. (Air, H2, C02, CH4, C3Hw, N2) -Water Air -Organic Liquids Air-Electrolyte soln. U,, m/s 0.003 -0.4 UL, m/s 0 -0.044 D, rn 0.152 -0.6 H , m 1.26 -3.5 Single nozzle sparger d,,, rn 0.005 C = 0.2 for pure liquids and non-electrolyte mixtures Akita and Yoshida (1973) U,. m/s 0 -0.2 D, rn 0.1 H , rn > 1.2 Perforated plate d,,, rn 0.0005, 0.001 U,, rn/s 0.0014 -0.14 D, m 0.05, 0.07, 0.10 Perforated plate d,,, m 0.00087 -0.00309 Bach and Pilhofer (1978) Ex -= 0.115[ I -E, E , = 0.72! U -0.485 U z + 0.0975 U 3 u = U G [ p L / u ( p L -PO)^]''^ Kurnar et al. (1976) M = 21-49 u, pL 0 5 7 8 U,, m/s 0.012 -0.38 D, rn 0.1

No abstract

NCUT Natural Resources Canada, Devon, AB T9G lA8, Canada ydrocracking and hydrotreating of heavy petroleum fractions or of hydrocarbon liquids that require desulfurization and/or H denitrification is one of today's major and now quite mature technologies. The reactions of hydrocracking, hydrodesulfurization (HDS) or hydrodenitrification (HDN) are usually carried out at relatively high temperatures (350'C to 45OOC) and high hydrogen pressures of up to 15 or 20 MPa. In most of these processes, a solid catalyst is present or is formed in the course of the reaction, so that the reaction system is always in a gas-liquid or gas-liquid-solid regime. In many cases, mass transfer can be the controlling rate mechanism in the reactor. In designing such reactors, or in the analysis of their performance, the ability to predict the hydrodynamic behaviour and the gas-liquid mass-transfer rate is necessary. One essential parameter for this analysis is a knowledge of gas-liquid interfacial tension values. The surface tension plays a major role in determining bubble sizes and bubble hydrodynamics, and therefore the interfacial transport area.There is another area in which knowledge of the gas-liquid interfacial tension is very important -the recovery of petroleum from reservoir formations. In the normal petroleum reservoir, oil trapping occurs at varying temperatures and at high pressures due to capillary forces resulting from surface tension. Particularly in secondary recovery, in which gas injection may be used, surface tension is a critical parameter. The variation of surface tension with temperature is well-known, and has been found to be essentially linear except in the vicinity of the critical point. The effects of pressure on surface tension are less easily predicted, and conflicting results have been reported. A significant reduction in surface tension has been found to occur at lower temperatures as gas pressure increases. However, the combined effect of high temperature and high pressure on the interfacial tension of hydrocarbon oils in contact with a gas with a high content of hydrogen or methane appears to require further investigation. The primary objective of this study was to make such surface tension measurements a t hydrocracking or hydrotreating conditions, if possible. Because no apparatus appeared to have been developed capable of such measurements a t these extreme 'Author to whom correspondence may be addressed. E-mail address: dscott@ eremail. uwaterlwxaAn apparatus for the measurement of surface tensions of organic liquids in contact with a gas has been developed which is capable of operation to 400°C and to 14 MPa. It is based on the maximum bubble pressure technique, modified for hydrocarbon oils at high pressures and temperatures. Accuracy of measurement is of the order of *3% for non-aqueous systems for values down to 5 mN/m. Only a 20 to 30 mL liquid sample is required, and small gas volumes. In practice, it was found that measurements with most organic liquids could only be made to a maximum of about ...