Accurate predictions of flow patterns for multiphase flow in pipes remain unresolved problems in petroleum engineering. Many efforts have been directed towards improved understanding of the phenomena and the development of physical models that allow more accurate analytical predictions of the flow pattern transition boundaries in steady-state multiphase flow. The primary goal of such models is to provide a truly general method that predicts existing flow patterns, once the flow rates, pipe geometry, and the fluid properties are specified. A unified approach, in which the same transition mechanisms can be applied over the whole range of inclination angle, has also been proposed.

The most essential part of solving engineering problems that depend on flow pattern predictions is the interpretation of results. The ability to explore and manipulate the results using a computerized flow pattern simulator would greatly enhance an engineer’s capability to optimize the design of multiphase transportation systems. This would also permit eliminating undesirable operating conditions; e.g. severe slugging at the riser in offshore operations. Progress in the area of Artificial Intelligence (AI) has made possible the partial realization of human decision making ability in a computer program. An intelligent program, supported by a knowledge data base, an inference engine, and human interaction, is an effective and economical solution for the present needs of advanced production optimization.

A computer simulator is undergoing development, using a Knowledge Based System approach. The system attempts to imitate, and eventually replace, a multiphase flow engineering expert. The Expert System utilizes an inference engine which is based on analytical mechanistic models and the expertise of many multiphase flow engineering experts and consultants. It is capable of predicting the existing flow pattern based on the current design, and will offer advice on alternatives and the direction to proceed if the existing flow pattern is unacceptable, or potentially problematic.

You can access this article if you purchase or spend a download.