The investigation of local bubble properties in multi-component liquids subject to foaming and elevated pressures can support the operation and optimization of industrial fluidized beds, such as the LC-Finer^SM hydroprocessor. Elevated pressures increases gas density which in turn enhances the bubble break-up rate resulting in smaller bubbles, lower bubble rise velocities and greater gas holdups.  Natural surface-active agents further reduce the rise velocity of these small bubbles by creating surface tension gradients and immobilizing the surface [Fan, L-S., et al. "Some aspects of high-pressure phenomena of bubbles in liquids and liquid–solid suspensions." Chemical engineering science 54.21 (1999) 4681-4709]. These two synergistic effects greatly increase the gas holdup in the hydroprocessor, which in turn decreases the liquid residence time resulting in lower product yields. In the hydroprocessor of interest, bitumen and hydrogen are heated separately, then mixed and introduced into the reactor plenum.  The reactor contains a bed of cylindrical catalyst particles fluidized by the upward flow of hydrogen, fresh bitumen and internally recycled liquid products.  The liquid is the continuous phase, while the hydrogen and catalyst constitute dispersed phases. Based on liquid and gas physical properties estimated by a commercial simulation package, previous simulations identified a large gas pocket at the top of the reactor, extending down into the recycle line [McKnight, Craig A., et al. "Fluid Dynamic Studies in Support of an Industrial Three-Phase Fluidized Bed Hydroprocessor." The Canadian Journal of Chemical Engineering 81.3-4 (2003) 338-350]. This suggested that the upper region of the reactor and recycle pan was not operating in a liquid-flooded regime and hence presented non-optimal conditions for gas/liquid separation. The gas phase separation efficiency of the recycle pan is critical, as gas entrained with the recycled liquid increases the gas throughput at the expense of the liquid holdup and product yield.

The objective of this experimental study is thus to investigate the effect of liquid phase properties, pressure, superficial gas velocity, and superficial liquid velocity on local and mean bubble characteristics, including gas holdup, bubble rise velocity and chord length, as well as liquid flow patterns in the reactor’s freeboard. This information will be used for boundary conditions of the recycle pan simulation, which will then in turn provide information on the gas and liquid throughputs in the reactor. This study is significant industrially as it contributes to the general understanding of high pressure multiphase reactors used in the chemical and petrochemical industries and provide understanding of gas/liquid interactions at higher gas holdup conditions in pilot-scale systems.

Local bubble characteristics were measured using a monofibre optical probe designed for high gas holdup conditions and to withstand high pressures. Results were compared to global phase holdups and mean bubble rise velocities obtained using the Dynamic Gas Disengagement (DGD) technique, adapted to gas-liquid flow, to verify the optical probe’s applicability range. It has been demonstrated that bubble column experiments can be used to simulate gas-liquid-solid fluidized bed’s freeboard region at high gas holdup conditions [Pjontek, Dominic, Valois Parisien, and Arturo Macchi. "Bubble characteristics measured using a monofibre optical probe in a bubble column and freeboard region under high gas holdup conditions." Chemical Engineering Science 111 (2014) 153-169]. Therefore, a 0.1 m inner diameter and 1.8 m in height bubble column operating a continuous circulation of gas and liquid was used. Several sight windows were installed along the column to allow for visualization and the gas was sparged in the plenum chamber of the column via a porous pipe with openings of 10 μm in diameter. The gas-liquid mixture then flows through a perforated distributor plate with 23 holes with diameters of 3.175 mm. This creates high shear conditions, producing small bubbles that do not tend to coalesce due to the high pressure and presence of surfactants. The operating pressures of the system was increased up to 6.5 MPa. Two liquid phases: tap water and 0.5 wt.% aqueous ethanol were employed to simulate coalescing and non-coalescing systems respectively. Various gas and liquid superficial velocities, both ranging between 0-120 mm/s, were used in order to respectively simulate gas entrainment in the liquid recycle line and varying liquid recycle pump speeds observed in the industrial reactor.

The optical probe was validated in water using local gas holdup profiles and superficial gas velocity measurements at atmospheric and elevated pressures. The probe struggled in the aqueous ethanol solution due to significant bubble size reduction (chord lengths below 0.5 mm), and that these smaller bubbles no longer rose rectilinearly (i.e., more affected by liquid flow field). Beyond the applicability range of the optical probe, the DGD technique was thus used to obtain mean bubble rise velocities at conditions of industrial interest. The evolution of mean and local bubble characteristics as well as gas and liquid flow patterns at high gas holdup conditions are correlated to operating conditions of the experimental system.