(155a) Process Digital Twin As an Enabler for Improving Operational Excellence in Ethylene Production Unit

Global consumption of olefins and its derivative is anticipated to be surging in near future. Along with the new capital developments, chemical production giants are focusing on maximizing existing asset utilization and feedstock utilization with the help of latest technologies. In volatile market environment, key business data from various sources is spread across the organization in random fashion which makes decision making cycles even more challenging. This is where Process Digital Twins with advanced performance computing facilitates enables in-process performance analysis for non-linear, multi-dimensional trade-offs to maximize the profitability.

Conventionally process models and simulation tools have been utilized for design, engineering and offline analysis. Multiple models of the same facility, create model maintenance and quality assurance issues making them futile when needed the most for sensitivity analysis, optimization, or revamp studies. On the other hand Process Digital Twin, a result of streamlined collaboration provided by bridging virtual and physical environments coupled with cognitive data enables engineers and process designers to build an immersive virtual replica for first-hand information on what is happening across the entire plant. Using process digital twins Engineers can solve problems by reasoning alongside process models while they are in-operation, whether they are working on the field or in a remote location. Thus Process Digital Twin not only serve as the single source for design, optimization, process control, virtual testing, predictive maintenance, and on-stream estimation of a physical process plant.

This paper discusses the development of a Process Digital Twin tailored for an Ethylene Production Unit of SABIC that dynamically updates and adjusts in accordance with the real time operating data. Petro-SIM^® process modelling and simulation tool has unique functionality to model associated utility systems on a unified platform that makes model development, validation, and updating simple. Integrated process and utility model provided a holistic view of the entire plant in real time. Purpose built utilities enabled plant engineers and operators to configure desired process (yield, losses etc.,) as well as sustainability performance Indicators within the Digital Twin architecture to monitor plant performance against business targets. Key aspect of process digital twin is its flexibility to convert it for conducting various objectives functions like equipment health check, bottleneck identification and overall optimization considering variable feedstock quantity / quality, operating and specific constraints in individual unit operations.