Digital Twins and Augmented Observability: An Architecture Model for Experimental Oil & Gas Separator Vessel Digitization

Nome: Raphael Almeida Guimarães dos Santos
Tipo: Dissertação de mestrado acadêmico
Data de publicação: 22/07/2021

Nomeordem decrescente Papel
Arnaldo Gomes Leal Júnior Co-orientador
Moisés Renato Nunes Ribeiro Orientador


Nomeordem decrescente Papel
Arnaldo Gomes Leal Júnior Coorientador
Celso Jose Munaro Examinador Interno
Humberto Salazar Amorim Varum Examinador Externo
Moisés Renato Nunes Ribeiro Orientador

Resumo: Digital industrial transformation has leveraged the concept of Digital Twins (DTs) andCyber-Physical Systems (CPSs) in the Industry 4.0 context. While CPS are multididimen-sional and complex systems that integrate computation, communication and control ofdynamic physical systems, DTs are related to high-fidelity models of physical elementsin a virtual space. Its goal is to simulate the physical world and provide near real-timefeedbacks to "what if" scenarios in order to assist operational decision-making. Separatorvessels are key elements in water-oil separation process in oil & gas industry. Basically, they are pressurized vessels that can be subjected to structural failures like fatigue if notadequately monitored, maintained and operated. Conventional sensing alone cannot matchthe diverse needs of DTs, and computer vision and machine learning come in hand toprovide a modern way to estimate multiple parameters. Thus, this work aims at proposingan architectural model to an experimental separator vessel DT encompassing modern sens-ing and data processing techniques. The experimental separator vessel is a multi-materialprototype built with carbon steel and polycarbonate monitored by a sensing techniquefor density profiling the multiphase crude oil based on a new computer vision techniquethat is proposed for non-contact density measurement. This information is consumed bythe DT’s structural digital models, which are custom built for this experimental separatorvessel. This way, digital models enable an improved sensing, i.e., extrapolating data fromfew sensors, to compose a rich information set to be provided to operators. This unleashes"what if" scenarios to be quickly tested and maintenance indicators to be inferred. In orderto illustrate these functionalities, a pressurized scenario is considered for this experimentalseparator vessel originally designed for 1 atm operation. The work contributions are pre-sented through a simple structural failure prediction approach under static and dynamicloads as well as through critical points traceability. Also, it is proposed a new methodfor estimate level interfaces in multiphase liquids by using clustering techniques. Thus, itis concluded that, at operation time, DTs has the potential of aggregating to separatorvessel process control new dimensions of observability.

Acesso à informação
Transparência Pública

© 2013 Universidade Federal do Espírito Santo. Todos os direitos reservados.
Av. Fernando Ferrari, 514 - Goiabeiras, Vitória - ES | CEP 29075-910