Multiphase flow and mixing quantification using computational fluid dynamics and magnetic resonance imaging

This paper investigates the current challenges and practices of quality measurement in the oil & gas industry. It particularly focuses on automatic pipeline sampling of petroleum liquids according to ISO 3171. The problem is tackled using innovative diagnostic techniques, advanced design optimisation tools and a new mixing system that uses a Liquid Jet In Cross Flow (LJICF) configuration. First, a 2.5 '' diameter small multiphase flow loop (SMPFL) was developed and magnetic resonance (MR) was utilised to characterise the mechanistic behaviour of mixing and the mixing efficiency of various nozzles. Second, a computational fluid dynamics (CFD) model was developed and validated using MR measurements on the SMPFL. The CFD model was then used to optimise the nozzle design as well as the design of a 10 '' nominal diameter large multiphase flow loop (LMPFL). The LMPFL is a well instrumented facility and was used to conduct mixing experiments on low velocity, low density and low viscosity fluids flowing in a horizontal pipe, which constitute challenging conditions for a mixing device to create homogeneous mixture. To quantify the homogeneity of the mixture created by the new mixing system on the LMPFL, a multiport profile proving (MPP) technique was developed and used to conduct water injection testing in compliance with ISO 3171 and API 8.2 standards. The water volume fraction (WVF) determined by the MPP had low relative error when compared to the mean WVF measured by the water cut meters and samples analysed using Coulometric Karl-Fischer (KF). Additionally, in an earlier study [1], the MPP measurement was able to detect a density gradient across the diameter of the pipe, making it an appropriate method to judge the homogeneity of the mixture. Therefore, the new mixing system together with the MPP technology shows real promise as an effective sampling and proving system for the petrochemical industry.