Sign in Subscribe
Natural gas

Selective Monitoring of Natural Gas Odorants using Machine Learning

Strategic Insights

— 1 min read

Researchers at the University of British Columbia have developed a portable device integrated with an array of five different sensors to detect a mixture of tert-butyl mercaptan and methyl ethyl sulphide in natural gas (NG) for concentration ranges of 1 ppm to 10 ppm. The device utilizes a machine learning model to predict the presence and concentration of NG odorants from sensor data, achieving a classification accuracy of 98.75% between NG odorants and hydrogen sulphide. The sensor system also exhibited high sensitivity and selectivity indicators, with a Mean Squared Error (MSE) of 0.50 and R error of 95.16%.

Key Takeaways:

  • A natural gas odorization system requires continuous monitoring to satisfy odorization guidelines, minimize over-odorization, and prevent hazardous gas leaks.
  • The developed portable device and machine learning model have promising applications for the selective monitoring of NG odorants.
  • The device achieved high sensitivity and selectivity indicators of 0.3667 (1/ppm) and 0.125, respectively.
  • The machine learning model predicted the presence and concentration of NG odorants from sensor data with a classification accuracy of 98.75%.
  • The MSE and R error of the sensor system were 0.50 and 95.16%, respectively.
  • The device was developed to detect a mixture of tert-butyl mercaptan and methyl ethyl sulphide in NG for concentration ranges of 1 ppm to 10 ppm.
  • The research was peer-reviewed and published in the Journal of Hazardous Materials.

Statistics:

  • Concentration range of NG odorants detected: 1 ppm to 10 ppm.
  • Classification accuracy of the machine learning model: 98.75%.
  • MSE of the sensor system: 0.50.
  • R error of the sensor system: 95.16%.
  • Number of sensors in the array: 5.
  • Numbers of authors on the research paper: 3 (including Nishat Tasnim, Mahan Ghazi, and Mina Hoorfar).

Sources:

  • NewsRx LLC. (2022, September 9). University of British Columbia Reports Findings in Machine Learning (Selective monitoring of natural gas sulphur-based odorant mixture of t-butyl mercaptan and methyl ethyl sulphide using an array of microfluidic gas sensors). Energy Weekly News, p. 971.
  • Selective monitoring of natural gas sulphur-based odorant mixture of t-butyl mercaptan and methyl ethyl sulphide using an array of microfluidic gas sensors. Journal of Hazardous Materials, 2022; 438:129548.

Read more