Desempenho de coluna de destila??o na obten??o de ?leos lubrificantes b?sicos

• Autor(a) principal: Machado, Rodrigo da Silva
• Data de Publicação: 2009
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações da UFRRJ
• Texto Completo: https://tede.ufrrj.br/jspui/handle/tede/349
• topic: destila??o; ?leo lubrificante b?sico; simula??o; distillation; lube oil; simulation; Engenharia Qu?mica

The present work is part of a Research Project that aims to study base lube oil separation from heavy feedstocks, by hydrocracking route, using a high vacuum distillation column, operating continuously, filled with Sulzer DX gauze metallic packing. The feedstocks used were donated by CENPES/PETROBRAS and were named as: a) light feed, composed of spindle and light neutral oil; b) medium feed, a mixture of spindle, light neutral oil and medium neutral oil and c) heavy feed, mix of light neutral, medium neutral and bright stock. In order to set up the best operational conditions, simulations using PROII commercial software was used, which also is able to evaluate the influence of the main operational parameters on unit revamp. After analyzing simulation results, as there was good agreement between original product curves and the ones from the simulation reports, the properties of the pseudocomponents generated on those reports could be utilized on column performance calculations, by using HETP (Height Equivalent to Theoretical Plate) concept. Also, due to that small deviation between those curves, the simulated HETP was considered the same as experimental HETP when compared to the models investigated. Concerning column performance study, two approaches have been adopted: a more practical one, using empirical model by Carrillo and coworkers (2000) and comparing it to vandom's curve (Sulzer); and a more theoretical one, comparing mass transfer models such as Olujic et al (2004) and its modified version and Rocha, Bravo and Fair's model (1985), adapted by Orlando Jr. (2007). It has been concluded that Olujic and coworker's original model (2004), that is, using Olujic's correlation (1968), which produces very low values for effective areas, making HETP values unrealistically high and, consequently, designing columns in a conservative manner. The use of the modified version of Olujic and coworker's correlation (2004), though, corrects the effective areas evaluation, failing however to reproduce the behavior of base lube oil mixtures, concerning HETP calculation, showing average deviations ranging from 40 - 80%. Deviations of the same order of magnitude were found when using Bravo, Rocha and Fair's correlation (1985), modified by Orlando Jr. (2007).