(733e) Enabling Model Based Decision Making by Sharing Consistent Equation Oriented Dynamic Models Between Multiple Simulation and Optimization Environments

Decision making in process industries relies on the use of models that represent operation of the equipment, process sites, and the entire supply chain. Different decision-support tools have been developed for specific business processes, such as process modeling, design, simulation and optimization, production planning, and production scheduling. Even within the domain of process design and simulation, distinct tools have been developed for dynamic modeling versus steady state modeling. Typically, process modelers used one of these tools to dynamically model a section of the process while modeling the rest of the process in a steady state environment. Simplified versions of the dynamic models or extensions to built-in models were then used within the steady state environment in an attempt to model the overall process. In order to accurately model certain processes, however, it is required to enable the sharing of rigorous dynamic models between multiple simulation and optimization environments to handle the diversity of mathematical formulations required to represent various aspects of the process system's operation. This problem is addressed by this research and development work.

Over the last several decades academic research institutions have developed generalized frameworks for process modeling, simulation, and optimization. Some of these frameworks have evolved into well-known commercially available mathematical modeling systems. Classic examples of such systems include Aspen Plus family of products for steady state modeling, Aspen Custom Modeler family of products for dynamic modeling and the HYSYS family of products for steady state and dynamic modeling. Since the late 90's there has been continued research and development on these products and the underlying Open Object Model Framework ([1], [2]) to enable the sharing of models between these applications and facilitate model based decision making.

This paper outlines significant algorithmic insights and contributions that enable the sharing of open equation oriented dynamic models authored in a high level modeling language between multiple modeling environments. It also details a successful implementation of this algorithm that facilitates simulation and optimization of large scale process models with dynamic batch distillation sequences and continuous steady state unit operation models executing seamlessly within a single environment.

The Common Model Environment [3] with the Open Object Model Framework as its backbone enables use of consistent models in Engineering, Planning and Scheduling, Advanced Control, Optimization, and Operations. Sample implementations of sharing of dynamic models in engineering are described in this paper. A second paper that is being prepared shows how similar models are used across Engineering, Planning and Scheduling, and Operations in the refining industry.

References

[1] Lakshmanan, A., Paules, G., Mahalec, V. (2006). A Modeling Framework That Enables Process Synthesis, Design, Analysis, Optimization, and Planning. Annual American Institute of Chemical Engineers Meeting. San Francisco, CA, (60c).

[2] Lakshmanan, A., Paules, G., Mahalec, V. (2007). Unified Architecture For Process Optimization. Mathematical Programming Society International Conference on Continuous Optimization. Hamilton, Canada, (ICCOPT II & MOPT A).

[3] Joffe, B. L., Kunt, T., Paules, G. (2008). The Common Model Environment – A new paradigm for model sharing between process engineering, planning and scheduling. Foundations Of Computer-Aided Process Operations. Cambridge, MA.