ARTIFICIAL INTELLIGENCE SIMULATION OF MULTIPHASE FLUID FLOW THROUGH EXTRACTION PIPES

Abstract

The flow of petroleum fluids through extraction pipes is a phenomenon difficult to describe in mathematical terms. Complex numerical relationships are usually used, most of which are derived from observation of wells in operation and analysis of collected production data, and others are determined from laboratory experiments. But a model that uses artificial intelligence in this field has not been created to date, even if this concept is widely used in the oil industry. This is precisely why this article proposes a numerical model based on artificial intelligence that determines the flow equations and their limits according to the type of flow, determines the relative velocities of gas and oil liquids during the ascent through the extraction pipes and also statistically analyzes the data obtained as a result of applying the described model with real data collected from a group of wells in production.

Keywords

• oil and gas
• fluid modelling
• artificial intelligence
• well production