(80a) Parallel Column Model for Reactive Dividing Wall Column Simulation | AIChE

(80a) Parallel Column Model for Reactive Dividing Wall Column Simulation

Authors 

de Villiers, R. - Presenter, Clarkson University
Taylor, R., Clarkson University
Kooijman, H. A., Clarkson University
As the process industries face the challenge of digitally transforming their plants and assets, they need tools and technology to support this change. Integrating the use of process simulation, operator training systems (OTS) and digital twin technology is an essential first step on this journey. Plant Digital Twins have undergone significant transformation over the past decades driven by innovations in engineering best practices, catalyzed by chemical engineering, mathematical optimization, machine learning, computer visualization and software innovations.

Process simulation software such as Aspen HYSYS, the backbone for a plant digital twin, has come a long way from automating the design calculations and process modeling to an environment for process optimization accelerating collaboration across the disciplines and tackling optimization across multiple dimensions simultaneously. The use of Process Simulation has also expanded from the traditional design applications to operations and planning. This digital transformation has created significant value in the industry resulting in capital and energy savings combined with better yields and capacities, increased safety and dramatic improvements in engineering productivity and production. Recently published customer stories that will be presented include: how YPFB in Andina, Bolivia increased production capacity by 18% and increased revenue by $280 Million/Year and how BPCL in Kochi, India achieved sustainable operations by reducing refinery emissions using a system-wide digital twin; how BP reduced flaring in their refinery by 50% and how Petrobras achieved production and safety improvements of 16 Million $/Year and eliminated slop vessel vent related shutdowns in an Upstream asset using digital twins and dynamic simulation.

Rapid advances in machine learning, advanced multivariate analytics, high performance mobile computing, and equipment and sensor connectivity are playing a role in enabling a whole new generation of technological solutions. The paper will cover recent developments in Process Simulation including expansion from Process to Asset Optimization with Hybrid Models that combine the strengths of first principles-based models and data-based machine learning techniques for process engineering and production optimization. Recently published customer stories that will be presented include: how Petro Rabigh improved refinery margins by 11.3 cents/barrel, how Tupras saved up to $1M in annual energy use and how Process Ecology increased the NPV of an Upstream asset by 8% through the use Hybrid Models.

The paper will also discuss how process simulation can be combined with new innovations in operator training and virtual reality software from Emerson to provide a digital twin of plant automation systems, 3D physical assets, and first principles process responses with comprehensive operator training management and integrated operation capabilities that help process industries maximize the value of their investment by creating a digital twin from the plant design and using it across a plant lifecycle as a digital thread. Customers taking advantage of this see increased project certainty and design integrity through scenario-based simulation in design phases, an average of 18 days of reduced time to startup for new facilities through virtual commissioning and startup activities, an average of 2.2 days of faster implementation and adoption of major improvement projects as a part of management of change processes, and a 49% reduction in unplanned operational downtime through improved operator training and decision making.