Modelagem e simulação de uma coluna de destilação usando linguagem python
Data
2023-11-23
Autores
Silva, Jéssica Brito da 
Orientadores
Tirelli, Marcelo Alexandre
Tipo
Trabalho de conclusão de curso
Título da Revista
ISSN da Revista
Título de Volume
Resumo
A destilação é uma operação unitária amplamente usada na indústria, cujo objetivo principal é separar ou purificar misturas por meio da aplicação de calor até o ponto de ebulição desta mistura, sendo assim um processo de grande custo energético. Para evitar desperdícios e otimizar a cadeia de produção, a modelagem e simulação são excelentes ferramentas para a predição do comportamento das colunas de destilação em determinadas condições, e cada vez mais aplicadas na indústria para a aceleração de projetos. Paralelamente ao crescente uso dos softwares de simulação comerciais, as linguagens de programação se desenvolvem para aumentar sua aplicabilidade em diversos segmentos do cotidiano. Nesse contexto, destaca-se a linguagem de programação Python, criada por Guido van Rossum nos anos 1990, atualmente aplicada em estudos científicos avançados devido a sua extensa coleção de bibliotecas e ambientes de desenvolvimento. Motivado por esses fatores, o presente trabalho teve como objetivo modelar e simular uma coluna de destilação fracionada de 11 pratos, a 1 atm e razão de refluxo 3,5, empregada na separação de uma mistura composta de benzeno e tolueno a 95°C, que opera em estado estacionário por meio de um algoritmo desenvolvida em linguagem Python. Foram aplicados dois métodos: modelagem por meio das equações MESH e pelo método FUGK. Para tal, foram usados os IDEs Jupyter Notebook e PyCharm, com a importação das bibliotecas math, NumPy e SciPy. Os resultados foram comparados com outra simulação do mesmo processo realizada no ambiente COCO Simulator por comparação direta e pelo cálculo do desvio relativo entre valores. Observou-se que o método FUGK, de equacionamento e aplicação mais simples, foi capaz de gerar os melhores resultados, com desvios menores do que 5%.
Distillation is a largely applied unit operation in industry, in which mixtures are purified or separated by the addition of heat until the mixture’s boiling point, being, therefore, a process that demands a high energy cost. To avoid waste and optimize the production chain, modelling and simulation are great tools to predict the behavior of distillation columns in determined conditions, having their use increased in industry to accelerate projects’ development. In parallel to the crescent use of commercial simulation software, programming languages are being developed to enhance its own applicability in many segments of daily routine. In this context, there is Python, a programming language created by Guido van Rossum in the 90’s, currently applied in several advanced scientific studies due to its extensive collection of development environments and libraries. Motivated by these factors, the present work had as objective to model and simulate a steady-state distillation column with 11 trays, reflux ratio of 3.5 and operating at 1 atm applied in separating a mixture of benzene and toluene at 95 °C using an application developed in Python. Two methods were applied: MESH equations and the FUGK method. For that, two IDEs were used: Jupyter Notebook and PyCharm, importing three libraries: math, NumPy and SciPy. The results obtained by both methods were compared to a simulation of the same process in COCO Simulator, using direct comparison and relative deviation calculus. After the end of the modeling and simulation process, it was observed that the FUGK method obtained the best results, with deviations lesser than 5%.
Distillation is a largely applied unit operation in industry, in which mixtures are purified or separated by the addition of heat until the mixture’s boiling point, being, therefore, a process that demands a high energy cost. To avoid waste and optimize the production chain, modelling and simulation are great tools to predict the behavior of distillation columns in determined conditions, having their use increased in industry to accelerate projects’ development. In parallel to the crescent use of commercial simulation software, programming languages are being developed to enhance its own applicability in many segments of daily routine. In this context, there is Python, a programming language created by Guido van Rossum in the 90’s, currently applied in several advanced scientific studies due to its extensive collection of development environments and libraries. Motivated by these factors, the present work had as objective to model and simulate a steady-state distillation column with 11 trays, reflux ratio of 3.5 and operating at 1 atm applied in separating a mixture of benzene and toluene at 95 °C using an application developed in Python. Two methods were applied: MESH equations and the FUGK method. For that, two IDEs were used: Jupyter Notebook and PyCharm, importing three libraries: math, NumPy and SciPy. The results obtained by both methods were compared to a simulation of the same process in COCO Simulator, using direct comparison and relative deviation calculus. After the end of the modeling and simulation process, it was observed that the FUGK method obtained the best results, with deviations lesser than 5%.