Key points are not available for this paper at this time.
Abstract The Oil and Gas industry is facing complex challenges including spiralling exploration and development costs, aging assets, long lead times on equipment, shortage of skilled people, among others. As such, it is imperative existing assets are operated efficiently and properly maintained to maximise production. A typical approach to achieve this objective considers industry best practices, international standards, manufacturer recommendations and condition monitoring techniques such as oil analysis, vibration, thermography, ultrasound, etc. However, the advent of Industry 4.0 has promoted the development and adoption of advanced techniques: a prominent example being the Digital-Twin concept. A Digital-Twin is a virtual model developed using advanced modelling and data handling techniques to replicate and predict the behaviour of individual equipment or entire production systems. A Digital-Twin approach can deliver predictive monitoring capabilities; analysis of complex systems; define solutions to operational issues; virtual testing and confirmation of proposed modifications; etc. This paper presents the development of a Digital-Twin for troubleshooting and operations optimisation of a motor-driven three-stage gas-lift compression system with a long history of unstable operation, recurrent emergency shutdown events and low availability. The multidisciplinary approach includes analysis of the control system architecture, assessment of field instrumentation, data validation, performance benchmarking, dynamic modelling and off-line control system testing. A cost-effective solution with only minor software modifications was designed, tested and demonstrated before its implementation. The successful outcome of the project reduced the need for constant operator intervention, reduced emergency shutdown events, increased availability of the gas compression system from 86% to 98.2%, reduced gas imports, increased gas exports and enhanced oil production.
Mucino et al. (Mon,) studied this question.