by Keyword: Thermal conductivity

Udina, S., Carmona, M., Carles, G., Santander, J., Fonseca, L., Marco, S., (2008). A micromachined thermoelectric sensor for natural gas analysis: Thermal model and experimental results Sensors and Actuators B: Chemical 134, (2), 551-558

Natural gas may show significant changes in its chemical composition depending on its origin. Typically, natural gas analysis is carried out using process gas chromatography. However, other methods based on the evaluation of physical properties have recently been reported. Thermal conductivity sensors are currently used in the analysis of binary mixtures of dissimilar gases. In contrast, natural gas is a complex mixture of mainly hydrocarbons, plus other residual gases as carbon dioxide and nitrogen. In this work, the response of a micromachined sensor integrating a heater and a thermopile is studied, regarding its potential use for natural gas analysis. A finite element thermal model of the device is described, and thermal operation simulations as well as a preliminary sensitivity analysis are reported. Experimental data has been collected and compared with simulated data, showing very good agreement. Results show that small variations in the gas mixture composition can be clearly detected. The sensor appears as a good candidate to be included in low-cost natural gas property analysis and quality control systems.

JTD Keywords: Natural gas, Thermopile, MEMS, Thermal conductivity, Modeling, FEM simulation

Udina, S., Pardo, A., Marco, S., Santander, J., Fonseca, L., (2008). Thermoelectric MEMS sensors for natural gas analysis Electronic Proceedings of the Seventh IEEE Sensors Conference 2008 Sensors, 2008 IEEE (ed. Frech, P., Siciliano, P.), IEEE (Lecce, Italy) , 1364-1367

T Multivariate data analysis techniques have been used for the first time in thermoelectric MEMS sensors in order to determine the composition of natural gas mixtures. Experimental measurements with different thermoelectric devices have been performed, the gathered data have been used to calibrate the sensor responses to four main components of natural gas: CH4, C2H6, N2 and CO2. Presence of the three first components was predicted with good accuracy while CO2 prediction was poor. Presented results indicate that thremoelectric sensors operated at different heater temperatures open the possibility of low-cost natural gas analysis.

JTD Keywords: Natural gas, Multivariate calibration, Thermal conductivity, Thermal sensor