ITEE Innovation Expo 2008 : Project Details

Electronic Nose Technology and Applications

Transformers are a major component in electricity distribution networks. They are expensive and require regular monitoring. Currently, the most widely-used method of monitoring is Dissolved Gas Analysis (DGA).

DGA analyses the gases that are formed by the transformer oil through gas chromatography. However, the costs of performing this procedure are high, as trained personnel and a laboratory are required. Thus, an alternative more economic method of gas analysis is sought after, which is why this study explores the idea of building an electronic nose (e-nose) using cheap components.

The design of the e-nose is based around the eZ430-RF2500 module as it is cheap and enables wireless transmission of data. The e-nose utilises a headspace odour handling and delivery system, which traps the odour, followed by an electrochemical sensor array that detects the odour and a pattern recognition system that identifies the odour. The pattern recognition system was developed by training it with a given set of transformer oil samples, each possessing its own distinct odour. Using the same set, each sample was placed inside the headspace delivery system and tested.

The electronic nose was able to distinguish each of the samples based on their distinct smells.

This demonstrates that an economical method of testing transformer oil can be implemented but more research is needed in order to enable these tests to classify the state of the transformer oil.