Numerical simulation study of offshore heavy oil desanding by hydrocyclones

Abstract

Sand production often occurs in offshore heavy oil production. Sand particles enter offshore platform processing systems with heavy oil, easily causing pipeline sand deposition, valve and separator erosion and blockage, and pump and other equipment wear and damage, resulting in various oilfield production safety risks. Due to the high viscosity of heavy oil, it exerts a strong viscous drag on sand particles. They easily adhere to heavy oil and move with its flow, which also leads to the extremely difficult separation of heavy oil and sand. Due to the desanding difficulty of highly viscous oil, this work proposes a desanding scheme combining water blending for viscosity reduction with hydrocyclone separation and focuses on hydrocyclone separation in the heavy oil desanding process. Based on algebraic slip mixture, Reynolds stress and discrete-phase models, the hydrocyclone desanding characteristics of heavy oil with the water cut were studied by computational fluid dynamics (CFD) numerical simulations. Flow field, sand removal rate (SSR) and pressure drop differences under different oil viscosities, oil contents and hydrocyclone geometries were studied via single-factor sensitivity analysis, and the essence of these differences was comprehensively analysed. Through high-precision response surface prediction model establishment, the influence of the interaction between hydrocyclone structures on the pressure drop was analysed, and the influence of the pairwise interaction between the main hydrocyclone structures on the hydrocyclone pressure drop is mutual inhibition. The hydrocyclone structure is optimized with pressure drop minimization as the objective, and the pressure drop of the optimized hydrocyclone is 23.3% lower than that at the lowest pressure in response surface experiments, indicating that the optimization effect of the response surface method (RSM) is extremely significant. Moreover, the research methods and conclusions in this work provide a basis for subsequent experimental prototype development and experimental verification steps.