Um método automático de preparação de misturas gasosas para determinação cromatográfica de metano, etano e propano em GNV

• Autor(a) principal: Barbosa, Mayara Ferreira
• Data de Publicação: 2015
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações da UFPB
• Texto Completo: https://repositorio.ufpb.br/jspui/handle/tede/8214
• topic: Sistema automático; Automatic system; Preparação de misturas padrão; Cromatografia a gás; Gás Natural Veicular - GNV; Preparation of standard mixtures; Gas chromatography; Compressed Natural Gas -CNG; CIENCIAS EXATAS E DA TERRA::QUIMICA

The compressed natural gas (CNG) is a fossil derivative mostly comprised of light hydrocarbons (methane, ethane, and propane), which has great energy importance for modern society. Considering the growing demand for CNG, it is also necessary to an effective control of the composition of the majority gases, which are mainly responsible for the energy efficiency of this fuel. The composition of these gases in CNG follows a certain pattern of conformity (control group) established by the ANP (National Agency of Petroleum, Natural Gas and Biofuels), to ensure the energy efficiency of CNG sold in Brazil. This work proposes the development of a methodology for automatic preparation of gas mixtures of methane standards, ethane and propane, in order to chromatographic determination of these gases levels in CNG, employing elaborate calibration models and validated by analysis of variance (ANOVA) and test recovery. The automatic system is of the dynamic type and is fully controlled by a program in order to prepare, as programmed by the user, binary mixtures of nitrogen / methane / nitrogen and nitrogen ethane / propane at levels (at fraction mol / mol) ranging in order to obtain good calibration models that consider the concentrations of these gases in the CNG. The evaluation methodology using ANOVA was quite satisfactory, the models built for the three gases methane, ethane and propane showed a high correlation (R2> 0.99) and significant in the regression, and showed no lack of fit and systematic variation at the residual plot. The performance parameters obtained from built models had lower detection and quantification limits 10-2 to 10-1, the analytical frequency was 4 samples per hour. A recovery study , involving three gas mixtures of certified composition and Natural Gas sample Vehicle (NGV) was also performed and the percentage values mean were 99.7 ± 3.1; 100.7 ± 4.4 and 98.0 ± 5.8 for methane, ethane and propane, respectively. The conclusion is based on the recovery values, as well as ANOVA, the proposed method was validated, with satisfactory precision and accuracy.