(216c) A Dynamic Optimization Approach to the Design of Air Separation Plants

Cryogenic air separation units (ASUs) have traditionally been operated at relatively constant production rates owing to steady electricity prices and customer demands. The deregulation of the energy markets has imposed a shift in this operating paradigm: at present, frequent changes in operating conditions and/or startup/shutdown are required to gain economic benefits from electricity price fluctuations, given that energy consumption is the dominant source of ASU operating costs.

These considerations have led to an increased emphasis on the development of process designs and control strategies that can improve agility and controllability (Miller et al., 2008a, 2008b). Furthermore, strict constraints imposed by utility providers on the duration of transitions between steady states call for a design approach that integrates process and equipment design, and process control.

Systematic approaches to the design of dynamically operable plants have been the subject of many research studies over the past three decades (Mohideen et al., 1996; Baker and Swartz, 2004; Sakizlis et al., 2004). However, most studies have been limited to single process units, often represented by highly simplified models. Our work aims to address the integration of design and control in a more comprehensive manner.

The proposed strategy relies on the use of a detailed dynamic model of the plant in conjunction with a) a dynamic operating envelope, defined in terms of minimum and maximum product flows, rate-of-change requirements for production rate between different operating points, product quality specifications and steady-state and dynamic equipment constraints, and b) an economic performance criterion.

Based on the Real Time Optimization principle (Engell 2007), we formulate a two-layer optimization strategy, consisting of: 1. A nonlinear steady-state optimization aimed at identifying the optimal values (in terms of the economic performance criterion) of the manipulated inputs at each operating point 2. A nonlinear dynamic optimization seeking the optimal (in terms of minimum transition time) trajectories of the manipulated inputs between operating points. In this layer, the optimal values of the manipulated inputs are utilized as end-point constraints. The use of a dynamic model allows us to account for process and equipment constraints, thereby identifying both steady-state and dynamic limitations of the process, in a natural and straightforward manner. Another significant advantage of this approach is that it can be made independent of the configuration of the control system of the plant; the resulting transition trajectories thus represent a best-case scenario and can serve as a benchmark for any controller implementation.

Subsequently, the proposed strategy is used to investigate the feasibility of the re-design of a nitrogen plant. Using data collected from plant operations, we develop and validate a detailed dynamic representation of the plant, consisting of a feed compressor, a multi-stream heat exchanger for heat integration, an expansion turbine, a distillation column and a reboiler-condenser. Realistic thermodynamic models are used to describe fluid phase equilibria and the physical properties of the feed and products, and index-reduction techniques are used to ensure that the final model is an index-1 DAE system.

We carry out a comprehensive study of the possible transitions in the dynamic operating envelope of the plant. Based on the operating and equipment constraints, we identify the limitations posed by the hardware and potential economic opportunities afforded by equipment upgrades.

References:

Baker, R. and Swartz, C.L.E. (2004). Simultaneous solution strategies for inclusion of input saturation in the optimal design of dynamically operable plants. Optimization & Engineering, 5, 5-24.

Engell, S. (2007), Feedback control for optimal process operation, Journal of Process Control,17, 203--219

Miller, J., Luyben, W.L., Belanger, P., Blouin, S. and Megan, L. (2008a). Improving agility of cryogenic air separation plants. Industrial & Engineering Chemistry Research, 47, 394-404.

Miller, J., Luyben, W.L. and Blouin, S. (2008b). Economic incentive for intermittent operation of air separation plants with variable power costs. Industrial & Engineering Chemistry Research, 47, 1132-1139.

Mohideen, M.J., Perkins, J.D. and Pistikopoulos, E.N. (1996). Optimal design of dynamic systems under uncertainty. AIChE J., 42, 2251-2272.

Sakizlis, V., Perkins, J.D. and Pistikopoulos, E.N. (2004). Recent advances in optimization-based simultaneous process and control design. Computers & Chemical Engineering, 28, 2069-2086.