Characterization of biodiesel using electronic nose and artificial neural networks

• Grant number: 06/04952-2
• Support Opportunities: Regular Research Grants
• Start date: November 01, 2006
• End date: June 30, 2009
• Field of knowledge: Physical Sciences and Mathematics - Chemistry - Analytical Chemistry
• Principal Investigator: Domingos Sávio Giordani
• Grantee: Domingos Sávio Giordani
• Host Institution: Escola de Engenharia de Lorena (EEL). Universidade de São Paulo (USP). Lorena , SP, Brazil

Abstract

The production of renewable fuels is among the priorities in of the Brazilian Government. In 2002, the Science and Technology Ministry started the Brazilian Program of Technological Development of Biodiesel (PROBIODIESEL). Energetic crisis such as the recent Bolivian Gas Nationalization shows that the Brazilian Energetic Matrix must be still optimized. Researches about the production of biodiesel has been done recently by some universities and the use of a 2% mixture with mineral diesel (B2) has already been indicated by the government, however, in order to assure the success of the program, some quality standards must be respected, such parameters must be not expensive and easily controlled. So, analysis methods must be available with low costs. Some biodiesel quality parameters, such as the cetane number and stability in oxidizing process, require expensive analysis and some equipment that can not be afforded by small producing units. This project intends to develop analytical methodologies the uses a simpler equipment, the electronic nose associated to artificial neural networks. The array of semiconductor sensors, forming the heart of the electronic nose, responds with a signal pattern characteristic for each biodiesel type or blend. The signal pattern will be correlated to the quality parameters and mixture proportions. In such a way, low cost, although powerful analytical centrals could be installed in small and medium producing units. (AU)