CONICET | Buscador de Institutos y Recursos Humanos

We present experimental data on droplet-wire interactions and compare with CFD simulation of idealized droplet-wire interactions. Since it is time consuming and difficult to perform accurate experiments due to the size of the droplets involved and the high pressures aimed for in our applications we have used CFD to investigate the effects of different process parameters on the droplet wire interaction, before continuing with detailed experiments. We investigate the effects of; droplet velocity, off and on centre collisions, wetted or dry wires, and surface tension (pressure normally lowers the surface tension in hydrocarbon systems). An important unit operation in many process industries is the removal of small dispersed liquid droplets (mist) from a gas stream. This is usually done to ensure that the gas; is within specifications set by the customer, does not cause adverse effects in the downstream process (e.g. foaming, erosion or corrosion), or lead to unwanted/unacceptable emissions to the environment. Typical processes where wire meshes are used to collect and facilitate removal of unwanted dispersed liquid include; amine absorbers for natural gas sweetening, glycol contactors used for dehydration of natural gas and gas-liquid scrubbers in hydrocarbon processing. A typical gas scrubber consist of (a) an inlet arrangement that distributes the two phase mixture as evenly as possible into the scrubber cross section, and if possible performs some degree of gas liquid separation, (b) a mesh pad / mist pad that coalesce droplets either into (i) larger droplets that are large enough settle under gravity into the liquid sump, (ii) collected and diverted by some means towards the vessel walls where they can settle under gravity due to lower gas velocity, or (iii) simply coalesced in the mesh to a droplet size that can be captured by the last main component of a scrubber (c) a vane pack or cyclone deck that is situated above the mesh pad / mist pad. The mesh pad is designed to suit the application in mind. Generally this entails optimizing free volume, surface area per volume, material compatibility, wire diameter and knitting of mesh. The complex flow inside a scrubber unit combined with a wide range of droplet dynamics as these interact with the flow, internals and each other makes these units hard to characterize and even harder to model. One would expect that there is a wide range of experimental data available on these systems, and that it would only be a matter of meticulous labour to piece together a robust model of the system. However, this is only true at the coarse correlation level. At the detailed level investigators are challenged to characterizing the dynamics of small droplets (e.g. 10-300µm) interacting with surfaces (walls and wires) and with each other. A rather large body of work has been performed on colliding droplets, and droplets colliding with walls where the walls are either dry or wet. However, most of this work has been performed using air-water systems and this data cannot really be used for high pressure gas-liquid separation. The reason for this is that at high pressures (at least in the processing of natural gas) the density and viscosity ratios between the two phases may be much smaller and last but not least, the all important surface tension drops dramatically. Actually, it turns out that there is only a limited amount of reliable data on the surface tension itself. All this is due to the difficulty of performing measurements of physical properties of hydrocarbon mixtures at elevated pressures and to conduct detailed experiments on them under process conditions (e.g. 80-150bar). This work forms part of an effort to improve current understanding the dynamics of high pressure gas-liquid separation processes. We report results from using CFD together with introductory experiments to plan high pressure experiments on hydrocarbon systems