Diseño y simulación de un sistema de control predictivo de perturbaciones hidráulicas y de proceso en el transporte de hidrocarburos

Abstract

Predictive control systems are essential to guarantee efficiency and safety in the transport of hydrocarbons, especially in complex environments such as Colombia, where factors such as orography, abrupt starts and stops, mixture of interfaces and variations in fluid density generate significant disturbances. The design and simulation of a predictive control system aimed at mitigating these hydraulic and process disturbances in pumping plants is proposed. For its development, historical operating data was analyzed, and machine learning models were implemented to identify the critical variables to be controlled. System simulation and predictive control integration were carried out using specialized software such as ASPEN Hysys and Rockwell Automation's RsLogix 5000. The main goal is to improve operational efficiency and minimize unscheduled downtime by real-time monitoring of suction and discharge pressures at each plant. This was achieved by dynamically adjusting the pump speed in the upstream plant based on the disturbance detection algorithm, ensuring that the system operates within safe limits. In addition, a proposal was presented for the efficient exchange of operational data when using the existing communication infrastructure. This system contributes to optimizing the management of hydrocarbon transport that increases the reliability and safety of the process.