John Vincent Ergina
Chief Technology Engineer - LNG & Gas Processing
14:00 - 14:30
Thursday, 19 September 2019
T2.9 InteLNG: Predictive Analytical Modelling of an LNG Plant
Automation Digitalization and AI in Energy
Machine learning, artificial intelligence, and the Industrial Internet of industrial things (IIoT) opens up a whole new frontier in the energy industry where engineers can now develop predictive models to prevent the onset of upsets in plant operation and improve real-time performance. EPC contractors who configured and constructed the facility are well positioned to be a major contributor to this revolution. Typically, EPC contractors use self-developed rigorous analytical models that are used in the engineering phase of a project to design a plant to react to various operating scenarios. Leveraging these models with the operating company at the LNG facility creates a rigorous replicate of the plant that is embedded with design know-how that is not obvious to an owner/operator. This know-how is particularly important to an operating plant because the conditions during operation are ever changing from the design basis and the critical equipment can be operated outside the intended design premise. Therefore, custom analytical models are not only beneficial in early design stages but they can also be the foundations of an LNG plant’s digital twin during operation.
Technological developments in recent years allow real-time data transfer and model initialization to become a reality. Thus, a digital twin, based on rigorous analytical validation, can be used for the optimization of plant capacity and early detection of unfavorable operating conditions that are not visible from the operating panel. Commercially, plants that employ these strategies will have shorter down-time and deliver more product annually. Again, EPC contractors can play an important role in maximizing plant rate as they are equipped with unique knowledge on where the design margins exist and how a piece of critical equipment has performed in similar facilities.
This paper describes KBR’s work process and experience in creating a digital twin for an operating plant. This paper will expand on how the digital twin can improve performance, monitor critical operating parameters and prevent unscheduled process upsets.