Application of Artificial Intelligent Methods in Identification of Two-Phase Flow Regimes

Multiphase flow plays an important role in oil and gas industry, the process industry, the nuclear industry, and many others. For example, accurate monitoring of multiphase flow is necessary to transport un-separated gas-oil-water mixtures over long distances, production of oil from deep wells, and in artificial lift. It is also encountered in power plants, geothermal plants and energy related industries for examples in boilers, evaporators, condensers, and many others. Monitoring the flow regime in the transmission pipe is important to maximize the rate of flow without driving the fluid to slug or froth mode, which creates enormous stress on the pipes, valves, pumps, and other operating instruments. Knowledge about flow regimes is also very essential for accurate predicting of the pressure gradient and the flow rate of each phase. In this research we will investigate methods for accurate identification of the flow regimes for gas liquid flow in pipes. Artificial Intelligent techniques as Fuzzy logic, and Neural Networks will be considered. Data will be generated using a well-known multiphase flow Unified model. Various input parameters will be evaluated for the best flow regime identification.