On the Comparison of Different Multiphase Flow Kernels for Gas Pipeline Real Time Advanced Functions

The nature of real time advanced functions or systems for gas pipelines carrying two phase mixtures is complex, especially when the dominant phase is gaseous. In this case, the flow quite often requires periodic pig runs, among several other specific operational procedures. Thus, a robust, accurate, reliable and fast kernel is mandatory for good performance on leak detection, location and quantification, as well as any other real time function such as virtual instrumentation or inventory management and reconciliation. This work aims at comparing different transient two-phase one-dimensional kernels under the perspective highlighted above. Specifically, one may cite: (i) a two-fluid model, implemented in 20 years consolidated commercial flow simulator, bringing a kernel of first order in time and second-order in space, three continuity equations, and three momentum equations, where 2 different pressures are possible to be calculated; (ii) a simple two-fluid model to serve as the basis of a different way to approach the same problem. The main focus of this paper is on real time applications, then its main goal is to provide considerations and comparisons on accuracy, reliability and processing speed (CPU performance) coming from six different numerical methods and schemes. The commercial flow simulator is then being used as the basis of comparison, as it is considered highly reliable and robust by the market.