Multiphysics modelling of geometrically-complex flow metering devices

Abstract

The simulation of complex multiphase flows has been the subject of extensive research for decades, owing to the wide range of industrial and natural processes involving the interaction of two or more fluids, particularly in the presence of variable geometries. This is essential for multiphase flow metering, which allows for production optimisation in oil fields for example. In this work, a novel approach for the numerical simulation of problems involving free-surface flows interacting with geometrically-complex rigid bodies is presented. This is implemented as an augmentation of a standard projection algorithm for the solution of the variable density incompressible Navier- Stokes equations. Additional physics, such as interfacial tension, are represented as body forces in the momentum equations. The procedure is simple to code and negates the use of specialised mesh generation or mesh-quality optimisation techniques. This method was implemented within an existing software framework (AMReX) which solves the governing partial differential equations for the flow while also including the capability for adaptive mesh refinement with optimal parallel scaling. The method is comprehensively assessed for accuracy and scalability by means of an extensive range of validation benchmarks. Subsequently, the code was used to simulate complex multiphase flow phenomena inside a prototype industrial flowmeter, demonstrating excellent agreement with experimental observations and showing that the framework can be used to inform the engineering design of flow control devices. Finally, the flexibility of the software was exploited to incorporate a computational model of an electrical capacitance tomography sensor for the purposes of multiphase flow measurement and monitoring. The approach was shown to accurately predict quantities of interest in a fully three-dimensional multiphase flow, allowing for investigations into the feasibility of using electrical sensors in real-time multiphase flow metering systems. This lays an excellent foundation for continually developing the software by incorporating additional physical models for flow phenomena of interest.