The online use of first-principles models in process operations: Review, current status and future needs

The online use of first-principles models (FPMs) to support process operations has been practised in the chemical and petroleum industry for over 40 years. FPMs can encapsulate a large amount of process knowledge and many companies have realized significant value from the use of these models in online model based applications (OMBAs). Such applications include real-time optimization, model predictive control, data reconciliation, virtual sensors, and process performance monitoring to name a few. The sophistication of both the FPM models and applications based on them has increased over time. At some points in the evolution certain applications were not successful due to issues related to sustainability, which includes model complexity, solvability, maintainability and tractability. Also, model development cost can be a factor in considering the type of model used in these applications. Hence many simplified and empirical model-based online applications became preferred in some domains, even though the overall prediction quality of the FPM may be superior. This paper will review the past experiences, current status and future challenges related to FPM based online modeling applications. There are many areas where the issues related to FPMs can be addressed through proper model management, better software tools and improved technical approaches and work processes. It is hoped that this paper can serve as a basis to promote an understanding of the issues for researchers, modeling software vendors, modeling engineers, and application engineers and help to stimulate improvements in this area leading to increased usage and value of FPMs in supporting process operations.

► A unified framework for online applications based on first-principles models (FPMs).
► A historical perspective highlighting past successes and discussing the reasons for past failures, and future research needs.
► The step-wise process of migrating a FPM from offline to online use.
► “Master” FPMs used to generate automatically via “on-the-fly” model transformations models used by online applications.
► A case study of the application of a FPM for nonlinear control of a polymerization process.