Odour discrimination from wastewater treatment plants with a portable electronic nose

A home-made and home-developed electronic nose based on a non-specific array of tin oxide sensors was built in order to discriminate among odors from wastewater treatment plants. The prototype uses headspace sampling system: The headspace of samples of water stored in 15 ml vials is carried out to the sensors cell using an integrated pump controlled by the PC. It also uses and integrated multisensor in silicon composed by six tin oxide gas sensors with different geometries and operating at different temperatures between 400 and 500 degrees C. The measurement system is based on an USB data acquisition card and an electronic card designed for this application and controlled by pc with a program executed on it. The sensor responses from the six sensor array were stored on the PC and preprocessed for further analysis. Later, linear methods like PCA and non linear ones like Artificial Neural Networks were used to chech the discrimination capability of the system.. Different samples of water polluted by several typical contaminants have been used to evaluated the performance of the sensory system. Other set of samples coming from different stages of a wastewater treatment plants has been also measured. Results show a complete discrimination in PCA plot and a success rate near to 100% is obtained with Artificial Neural Networks in the Classification of Samples.