Síntesede de redes flexíveis de trocadores de calor via programação matemática usando um método sequencial

• Ano de defesa: 2014
• Autor(a) principal: Miranda, Camila de Brito
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Estadual de Maringá; Brasil; Departamento de Engenharia Química; Programa de Pós-Graduação em Engenharia Química; UEM; Maringá, PR; Centro de Tecnologia
• Palavras-chave em Português: Calor; Redes flexíveis de trocadores; Otimização; Programação linear (PL); Método sequencial; Programação não linear (PNL) Programação matemática; Brasil.; Engenharias; Engenharia Química
• Link de acesso: http://repositorio.uem.br:8080/jspui/handle/1/3826

One of the most important fields in the study of industrial processes is the synthesis of heat exchanger networks (HEN). By using a HEN it is possible to maximize the energy recovery, reducing the use of cold and hot utilities. In this paper it is presented models using mathematical programming for the synthesis of flexible heat exchanger networks, which are HEN able to support changes in flowrates, inlet and outlet temperatures in a finite sequence of periods of time. The models were developed according to a sequential method based on Floudas and Grossmann (1986, 1987a). This work is decomposed in three goals: the minimum utility cost for each period of operation formulated as a linear programming (LP) problem; the fewest number of units of heat exchange formulated as a mixed integer linear programming (MILP) problem; and the minimum investment cost, formulated as a nonlinear programming (NLP) problem. In the last case the original superstructure is improved. A bypass coming out of the entrance of a heat exchanger and into the output of another heat exchanger is added. Also two sets and a parameter are created to reduce the number of variables and constraints involved in this mathematical formulation. Two examples were used to test the models applicability and results achieved by the models are compared with the results of Floudas and Grossmann (1987). Total costs are better than the original paper, indicating that the authors result was trapped in local minima. The last example was also applied for Isafiade and Fraser (2010), who had a total annualized cost higher than that found for models developed in this work. The results found demonstrate the viability of the proposed models, reaching the goals of this work.