Files
Document type
ArticleVersion
Accepted versionPublication date
All rights reserved
Please use this identifier to cite or link to this item: https://hdl.handle.net/2445/187960
A micromachined thermoelectric sensor for natural gas analysis: Multivariate calibration results
Journal Title
Director/Tutor
Additional contribution
Journal ISSN
Volume Title
Abstract
The potential use of a micromachined thermopile based sensor device for analyzing natural gas is explored. The sensor consists of a thermally isolated hotplate, which is heated by the application of a sequence of programmed voltages to an integrated heater. Once the hotplate reaches a stationary temperature, the thermopile provides a signal proportional to the hotplate temperature. These signals are processed in order to determine different natural gas properties. Sensor response is mainly dependent on the thermal conductivity of the surrounding gas at different temperatures. Seven predicted properties (normal density, superior heating value, Wobbe index and the concentrations of methane, ethane, carbon dioxide and nitrogen) are calibrated against sensor signals by using multivariate regression, in particular partial least squares. Experimental data have been used for calibration and validation. Results show property prediction capability with reasonable accuracy except for prediction of carbon dioxide concentration. A detailed uncertainty analysis is provided to better understand the metrological limits of the system. These results imply for the first time the possibility of designing unprecedented low-cost natural gas analyzers. The concept may be extended to other constrained gas mixtures (e.g. of a known number of components) to enable low-cost multicomponent gas analyzers.
Related resource
Subject (English)
Citation
Citation
UDINA OLIVA, Sergi, et al. A micromachined thermoelectric sensor for natural gas analysis: Multivariate calibration results. Sensors and Actuators B-Chemical. 2012. Vol. 166-167, num. 338-348. ISSN 0925-4005. [consulted: 6 of October of 2026]. Available at: https://hdl.handle.net/2445/187960