2008 Annual Meeting

(424f) Experiences in the Building of Dynamic Flowsheet Models for Embedded Nonlinear Control

Authors

Goodman, J., Honeywell
Renfro, J. G., Honeywell
Xie et al [1] were among the first to discuss nonlinear reactor modeling for gas phase polymerization processes with emphasis on control application. Subsequently controllers that directly incorporate an embedded first principles dynamic flowsheet model have been applied to a range of transition metal and free radical polymerization kinetics industrial processes. These controllers handle grade runs, transitions and constraint pushing. Conceptually, some of the benefits of an embedded first principles dynamic flow-sheet within the controller include (a) 'what if' studies in offline mode; (b) reduced engineering focus for regression evaluation due to the underlying nature of the control model; and (c) ready availabilty of inferred measurements, for example virtual gas composition analyzer. Realization of some of these benefits for an industrial gas phase polymerization process is discussed by Allsford et al [2].

This paper discusses lessons learned from these applications, with reference to industrial examples, in the contexts of (i) scoping the flowsheet model; (ii) offline parameter estimation to match the model to process steady state; and iii) online state estimation for continuous and discontinuous process measurements. An emphasis is placed on (ii). The discussion in this paper is of a general nature, although the experiences were gained with specific tools not in the public domain.

References

1. T. Xie, K. B. McAuley, J. C. C. Hsu and D. W. Bacon, “Gas Phase Ethylene Polymerization: Production Processes, Polymer Properties and Reactor Modeling”, Ind. Eng. Chem. Res., 33, pp. 449-479, 1994.

2. K. V. Allsford, J. Goodman, J. Ramlal and R. Beigley, "Nonlinear Multivariable Control and Optimization of a Polyethylene Process Based on Embedded Dynamic Chemical Engineering Model", Proc of 3rd International Symposium on Advanced Control of Industrial Processes (ADCONIP), Banff, Canada, May 4-7 2008.

3. J.D. Hedengren, K. V. Allsford and J. Ramlal, “Moving horizon estimation and control for an Industrial Gas Phase Polymerization Reactor”, Paper WeB19.4, Proc of American Control Conference (ACC), New York, NY, Jul. 2007.