-
Júlia Guardiani
Departamento de Engenharia da Computação, UFRN, RN
-
Celi Lopes
Escola de Ciência e Tecnologia, UFRN, RN
-
Diego Medeiros
Departamento de Engenharia da Computação, UFRN, RN
-
Lucas Marques
Departamento de Engenharia Elétrica, UFRN, RN
-
Danilo Curvelo
Instituto Metrópole Digital, UFRN, RN
-
Diego Cabral
Escola de Ciência e Tecnologia, UFRN, RN
-
Gustavo Leitão
Instituto Metrópole Digital, UFRN, RN
-
Ivanovitch Silva
Departamento de Engenharia da Computação, UFRN, RN
-
Luiz Affonso Guedes
Departamento de Engenharia da Computação, UFRN, RN
Keywords:
flow inference, flow measurement, flare system, flare, mass of gas to flare, data-driven
Abstract
The entire industry has been striving to transition to a low carbon economy, and among these efforts are initiatives to reduce release of gas to the flare system. In order to achieve this reduction, it is necessary to measure and identify the sources of gas being released in the process. Due to the involved costs and technical difficulties, usually only the main flare lines have theirs flow measured in those systems. This article presents a flow inference system for major elements that release gas into the flare system, including safety relief valves (PSVs), control valves, and restriction orifices. In addition to estimating the flow, the proposed system groups gas delivery sources, making it possible to evaluate process units, areas and industrial plants, in addition to estimating the cost of this mass of gas burned in the flare system, considering even the type of involved gas. As a pilot application, the described solution comprises the monitoring of oil and gas refineries.