Experimental Study and Modeling of Sand Erosion in the Gas-Liquid Cylindrical Cyclone GLCC Separators

Abstract

AbstractGas-Liquid Cylindrical Cyclone (GLCC) separators have been widely accepted as an alternative to conventional vessel-type separator during recent years. The operators are using GLCCs at flow rates higher than the normal recommended operational envelope in applications where the increased liquid carry-over and gas carry-under can be tolerated. Experience gained from production of hydrocarbons with entrained sand has shown that severe degradation of production equipment may occur due to solid particle erosion at high productions rates. However, there are no data and no field feedback on erosion in the inlet region of a GLCC. Thus, the goal of this work is to evaluate the erosion condition in the Gas-Liquid Cylindrical Cyclone (GLCC) separator under gas production and low-liquid loading flow condition.Erosion experiments are conducted with gas-sand and gas-liquid-sand flow conditions in a laboratory scale GLCC with two particle sizes and gas velocities. An ultrasonic thickness measurement system is employed to monitor the erosion rates at the inlet region of the GLCC. Several cases are simulated with commercially available Computational Fluid Dynamics (CFD) software. Based on the experimental data and CFD simulations, a mechanistic one-dimensional model is proposed to calculate maximum thickness loss inside the GLCC.It is observed that the location of maximum erosion is not varying significantly with the flow condition with or without the presence of liquid and particle size, while the liquid entrainment can reduce the value of maximum erosion (in mm/kg) by one order of magnitude. The erosion increases with the flow velocity as expected but not with particle size for the particles used in current experiments. CFD simulation results that are obtained by utilizing the mechanistic erosion equations provided fair agreement with experimental measurements for gas-sand condition, and the simplified model showed consistency with the data in both gas-sand and gas-liquid-sand flow conditions.