Early detection system for coal spontaneous combustion by laser dual-species sensor of CO and CH4.

• A dual-species sensor was constructed based on two-channel ALIA. • A Herriott cell and normalized technique were adopted to improve sensitivity. • A sensitivity for CO and CH 4 was about 0.20 ppm with a 2.5 s sampling time. • The CO emission satisfied the exponential growth tendency during CSC. • Two CTs were recognized and were able to precisely forecast CSC. As a common hazard, coal spontaneous combustion (CSC) causes the loss of coal in piles and mines. To rapidly and reliably forecast CSC, a portable and powerful detection system based on a laser spectroscopic technique is proposed, to precisely measure gaseous inhibitors; the technique uses a single, distributed-feedback diode laser emitting at 2.33 μm as a sensing light source for dual species. The detection system carries out wavelength-modulation spectroscopy to suppress detection noise by using a two-channel analogue lock-in amplifier (ALIA) and a complex programmable logic device (CPLD) controller. To improve detection sensitivity, a Herriott cell is adopted to increase the interaction length between the laser and the target species. The embedded system on a chip (SoC) of STM32F103 is the charge of the current driver and temperature control of the laser. Deployment of the SoC of STM32H743 acquires and digitally filters the original demodulated absorption signal, after which communication with a personal computer takes place. The second demodulation component, normalized by the first demodulation component, is used to improve the signal-to-noise ratio. The sensor-calibration experiment reveals the square of the correlation coefficients R2 to be 0.99718 and 0.99905 for carbon monoxide (CO) and methane (CH 4), respectively, over the concentration range of 5–200 ppm. Using Allan-Werle deviation analysis, the detection sensitivity for CO and CH 4 are 0.27 ppm and 0.20 ppm with a 2.5 s sampling time, and 0.05 ppm and 0.03 ppm with ~125 and 95 s integration times, respectively. Finally, the detection system is used for the online measurement and analysis of the growth pattern of gaseous products which are released during the CSC process, the temperature varying from 30 to 200 °C. Two critical temperatures were correctly determined to further forecast CSC status. On beginning a vigorous oxidation process, the exponential-evolution trends of the CH 4 and CO gases agree with each other. Compared with CH 4 , CO may be more suitable for online, early detection of CSC. [ABSTRACT FROM AUTHOR]