Mostrar registro simples

dc.contributor.authorLima, Tiagopt_BR
dc.contributor.otherLuz Junior, Luiz Fernando de Lima, 1966-pt_BR
dc.contributor.otherMitchell, David Alexander, 1952-pt_BR
dc.contributor.otherUniversidade Federal do Paraná. Setor de Tecnologia. Programa de Pós-Graduação em Engenharia Químicapt_BR
dc.date.accessioned2021-06-09T19:21:46Z
dc.date.available2021-06-09T19:21:46Z
dc.date.issued2009pt_BR
dc.identifier.urihttps://hdl.handle.net/1884/24023
dc.descriptionOrientador: Prof. Dr. David Alexander Mitchellpt_BR
dc.descriptionCo-Orientador: Luiz Fernando Lima Luz Jrpt_BR
dc.descriptionDissertação (mestrado) - Universidade Federal do Paraná, Setor de Tecnologia, Programa de Pós-Graduação em Engenharia Química. Defesa: Curitiba, 31/08/2009pt_BR
dc.descriptionBibliografia: fls.136-142pt_BR
dc.description.abstractResumo: A fermentação no estado sólido é um processo capaz de produzir diversos compostos de interesse comercial. Uma das dificuldades para a aplicação industrial deste método é a falta de sensores economicamente viáveis capazes de realizar medidas on-line da umidade da fase sólida, que se for reduzida demasiadamente pode limitar o crescimento do microrganismo e a produção dos metabólitos. O objetivo deste trabalho foi de desenvolver um modelo matemático determinístico para inferir a umidade da fase sólida em um biorreator com aeração forçada e agitação intermitente. A combinação de duas características distingue este modelo de inferência de modelos anteriores: a primeira, é que o modelo descreve os gradientes axiais da umidade do leito ao invés de assumir que o leito é homogêneo; a segunda, é que o modelo descreve as transferências de massa e energia entre a fase gasosa e sólida dentro do leito ao invés de assumir que estas fases estão em equilíbrio. As medidas do processo fornecidas ao modelo de inferência eram as temperaturas da fase gasosa em diferentes alturas do leito. Inicialmente, o modelo de inferência foi testado com um modelo preditivo do biorreator: quando as temperaturas da fase gasosa do modelo preditivo foram fornecidas ao modelo de inferência, o modelo de inferência calculou corretamente a umidade da fase sólida. Na seqüência, motivado pelo fato que um biorreator real possui um número limitado de sensores de temperatura, o modelo preditivo foi utilizado para gerar curvas de temperatura da fase gasosa em função da altura do leito em diferentes tempos de cultivo. Equações empíricas foram ajustadas a estes perfis de temperatura, com a intenção de usar estas equações no modelo de inferência para estimar temperaturas dentro do leito do biorreator de escala piloto nos pontos intermediários aos sensores, que estão localizados nas alturas de 0, 5, 18, 33 e 50 cm, representando a base, três alturas intermediárias e o topo do leito. Entretanto, em um teste no qual o modelo preditivo foi utilizado para gerar temperaturas da fase gasosa nestas alturas do leito, o modelo de inferência estimou melhor a umidade da fase sólida quando temperaturas nas regiões intermediárias dos sensores foram obtidas por interpolação linear. No terceiro passo, o modelo de inferência, com interpolação linear das temperaturas, foi empregado para investigar o melhor posicionamento dos três sensores internos no leito. Foi observado que as melhores estimativas da umidade do leito eram obtidas quando os sensores estavam concentrados na região inferior do leito. Na última parte do trabalho, o modelo de inferência foi testado em cultivos de Aspergillus oryzae sobre uma mistura de soja e farelo de trigo no biorreator de escala piloto. Durante estes cultivos as temperaturas do gás de entrada e saída do biorreator e nas alturas internas do leito de 5,18 e 33 cm foram monitoradas. Amostras do leito foram removidas nas mesmas alturas dos sensores para a determinação da umidade do leito. Na altura de 5 cm o modelo inferiu o ressecamento do leito mais acentuado do que as medidas experimentais. Nas alturas do leito de 18 e 33 cm a umidade real estava abaixo da umidade inferida. Foi concluído que o modelo de inferência tem potencial para fornecer estimativas adequadas da umidade da fase sólida, mas em trabalhos futuros seria necessário obter melhores estimativas dos parâmetros chaves, especialmente dos coeficientes de transferência de massa e calor.pt_BR
dc.description.abstractAbstract: Various products of commercial interest can be produced by solid-state fermentation. One of the difficulties faced in this cultivation method is the lack of simple sensors that can be used to measure the water content of the solids on-line; this is an important state variable since, if it falls to low values during the process, the growth of the microorganism and the production of metabolites will be adversely affected. The aim of the current work was to develop a deterministic mathematical model to infer the water content of the solids in a bioreactor with forced aeration and intermittent agitation. The combination of two features distinguishes this inference model from previous models: Firstly, the model describes axial gradients in water content, rather than assuming that the bed of solids is well mixed and, secondly, the model describes heat and mass transfer between the solids and gas phases within the bed, rather than assuming that these two phases are in equilibrium with one another. The input for the inference model is comprised of gas phase temperatures measured at different heights within the bed of solids. Initially the inference model was tested against a predictive model of the bioreactor: when the gas phase temperatures from this predictive model were supplied to the inference model, the inference model correctly calculated the solid phase water content. In a second step, motivated by the fact that in a real bioreactor there is a limited number of temperature sensors, the predictive model was used to generate gas phase temperature profiles for different cultivation times. Empirical equations were adjusted to these temperature profiles, with the intention of using these equations within the inference model to estimate temperatures within the bed of a pilot-scale bioreactor, at points between the temperature sensors, which are located at the bed heights of 0, 5 18, 33 and 50 cm, representing the bottom of the bed, three internal heights, and the top of the bed. However, in a test in which the predictive model was used to generate gas phase temperatures at these bed heights, the inference model estimated the solid phase water content best when temperatures in the intervening regions were obtained simply by linear interpolation. In a third step, the inference model, with linear interpolation of temperatures, was used to investigate the optimum positioning of the three internal bed sensors. It was shown that the best estimates of bed moisture content will be obtained when the internal sensors are concentrated in the bottom part of the bed. In the last part of the work, the inference model was tested in cultivations carried out in the pilot-scale bioreactor and involving the growth of Aspergillus oryzae on a mixture of wheat bran and soybeans. During these cultivations gas phase temperatures were monitored at the air inlet and outlet and at the internal bed heights of 5, 18 and 33 cm. Samples also were removed from these bed heights for the determination of the solid phase water content. At a bed height of 5 cm the model inferred a rate of drying of the solids that was higher than the experimentally measured rate. On the other hand, at the bed heights of 18 and 33 cm, the model overestimated the solids phase water content. It was concluded that the inference model has the potential to give adequate estimations of the solids phase water content, but that further work will be needed to obtain better estimates of key parameters, especially the mass and heat transfer coefficients.pt_BR
dc.format.extent165f. : il. [algumas color.], grafs., tabs.pt_BR
dc.format.mimetypeapplication/pdfpt_BR
dc.languagePortuguêspt_BR
dc.relationDisponível em formato digitalpt_BR
dc.subjectTesespt_BR
dc.subjectFísico-químicapt_BR
dc.subjectBioreatorespt_BR
dc.subjectModelos matemáticospt_BR
dc.subjectGlicerinapt_BR
dc.subjectAdsorçãopt_BR
dc.subjectEngenharia Químicapt_BR
dc.titleModelo de inferência para a estimação da umidade do leito de um biorreator de fermentação no estado sólidopt_BR
dc.typeDissertaçãopt_BR


Arquivos deste item

Thumbnail

Este item aparece na(s) seguinte(s) coleção(s)

Mostrar registro simples