Optimal control of heat exchanger networks
Nenhuma Miniatura disponível
Citações na Scopus
1
Tipo de produção
Artigo de evento
Data
2009-01-05
Autores
Luis Novazzi
ZEMP, R. J.
ZEMP, R. J.
Orientador
Periódico
Computer Aided Chemical Engineering
Título da Revista
ISSN da Revista
Título de Volume
Citação
NOVAZZI, L.; ZEMP, R. J. Optimal control of heat exchanger networks. Computer Aided Chemical Engineering, v. 27, n. C, 1647-1652, 2009.
NOVAZZI, L.; ZEMP. R. J. Optimal control of heat exchanger networks. Computer Aided Chemical Engineering, v. 27, n. C, p. 1647-1652, 2009.
NOVAZZI, L.; ZEMP. R. J. Optimal control of heat exchanger networks. Computer Aided Chemical Engineering, v. 27, n. C, p. 1647-1652, 2009.
Texto completo (DOI)
Palavras-chave
Resumo
Energy integration among chemical process streams can lead to quite complex heat exchanger networks (HEN) with difficulties in terms of control. In the HEN, the control system primary objective is to keep outlet stream temperatures in a specified range. This objective can be achieved by manipulating bypasses in the exchangers or heat loads in coolers or heaters, which results in a positive degree of freedom, since the number of possible manipulated variables is greater than the number of controlled variables. Thus, a secondary control objective can be set: minimization of utility consumption. In this work HEN dynamics and steady state aspects were addressed with the aim of proposing a control strategy that minimizes utility consumption and satisfies imposed restrictions. With the tools developed in this work it was proposed a feedforward optimal control strategy, which consisted in the minimization of a steady state objective function, connected to utility consumption and subjected to constraints in outlet stream temperatures. The suggested approach is an interesting one since rigid or flexible control objectives can be set. The solution of the minimization problem resulted in optimal bypasses positions in the HEN, which were dynamically implemented using a ramp function and a step function. It was verified that ramp implementation was better, with tolerable dynamic violations. © 2009 Elsevier B.V. All rights reserved.