Accurately quantifying individual flowrates of phases in an oil production stream allows operators to make informed decisions about oil well performance and to optimise transport operations, thus saving both time and cost. However, the complexity of the multiphase flow structure creates a challenge and is a serious barrier to obtaining this knowledge.
Simulation Technologies were approached to support the ongoing research and development of a prototype multiphase flow meter that aims to simultaneously measure the different phase flowrates through advanced measurement techniques [Faraj et al, 2015, Wang et al, 2015]. The application utilises new technology to measure rapid variations in multiphase flows, providing the capability to measure multiphase flow distributions. A large amount of experimental work had previously been carried out, which, proved to be quite costly and time consuming, therefore, the client had turned to advanced simulation technologies in order to further accelerate their R&D programme at reduced cost.
The previous experimental work facilitated the development of mathematical models, which greatly advanced the theoretical approach to the problem. However, discrepancies between the calculated results and the experiments were observed. In order to further calibrate and increase the accuracy of the mathematical models, additional understanding of the velocity differences between each of the phases was required. This information could only practically be provided by powerful multiphase simulation techniques, such as those used in the area of Computational Fluid Dynamics (CFD).
To tackle the problem, a bespoke simulation application (SimApp) was developed for the client that aimed to replicate the previous experimental test rig. We created a CFD simulation application (CFDApp) consisting of a back-end simulation package containing only the most appropriate underlying simulation software and determined the correct methodology for simulation of these multiphase flows. We then created of a front-end webapp, providing the user with a very simplified simulation experience and an easy to understand work flow. The CFDApp was hosted on our encrypted pre-configured virtual server, having the exact computational resources required for the problem at hand.
The result was that the client, a simulation non-expert, was able to easily run the much needed simulations and quickly understand and extract the simulation results, exactly in the way that they required.