Experimental and numerical investigation of methane flame propagation characteristics in a 30.5-m explosion reactor with obstacles

Confined explosions are a major risk in underground coal mine operations. Depending on confinement, ignition location and obstacles in the flame path, methane explosions can transition to detonations. This research focuses on explosion experiments in a reactor with diameter of 71 cm and length of 30.5 m, along with developing the corresponding computational fluid dynamics (CFD) model that captures the observed flame behavior. The results show that the 2D and 3D CFD models are able predict the experimental flame front propagation velocity, but the 3D model shows more explicit conformity than the 2D model. Introduction Explosions in underground coal mine operations can be devastating, which can result in multiple casualties and destruction of mine infrastructure. Conducting experimental methane explosion research in a full-scale industrial setting is not practical due to cost and lack of available facilities worldwide. In recent years, CFD modeling has been used extensively to study the flame behavior and potential impact of an explosion in different industrial settings. CFD models that are capable of accurately predicting complex flame propagation behavior during an explosion scenario are highly sensitive to the chosen physical models (such as turbulence model), boundary conditions and initialization parameters. Thus, it is vital to acquire sufficient experimental data for validation to provide a reliable and robust model that is capable of simulating a methane gas explosion in an underground coal mine. The objective of this research is to investigate the impact of rock rubble obstacles on methane flame propagation velocities inside an explosion reactor that is 30.5 m long and 71 cm in diameter, and to validate and improve the predictive capabilities of a corresponding CFD methane explosion model. Fig. 1 Explosion reactor setup for large-scale experiments. Experimental and CFD modeling setup In the large-scale experiments shown in Fig. 1, a methane (CH4)-air gas mixture with a concentration of 7.5 volume percent of CH4 is filled into a semiopen reactor with length of 30.5 m. The ignition of the explosive atmosphere happens 38 cm from the closed end of the reactor. Ion sen- sors protrude 10±1 cm radially into the reactor to record the flame front as it propagates toward the open end. The reactor consists of four sections, each 7.62 m in length, providing flexibility of testing different total reactor lengths and allowing varying CH4-air mixtures to be confined to a specific zone. For the purpose of this research, three sections, approximately 23 m in total length, act as a reactive zone that includes the explosive mixture of 7.5 volume percent CH4. The fourth section, located at the open end, is designated as a nonreactive zone and is filled with air. The local atmospheric pressure is 82 kPa, and temperatures range between 283 and 295 K (depending on the day of the test) prior to ignition. After the filling process, the reactor pressure is 86 ± 3.4 kPa. A plastic sheet barrier is installed between reactors 1 and 2 to separate the reactive zone from the nonreactive zone. This barrier is designed to break at approximately 156 kPa. For every test, the reactor’s reactive zones are filled with approximately 27 m3 of premixed CH4-air mixtures, three times the volume of the reactive zone, from the closed end, while the ambient air initially inside the reactor prior to filling is allowed to bleed through valves through ports that are 0.47 m and 0.94 m from the gas barrier at the end of reactor 2. The corresponding CFD models in 2D and 3D follow the same geometry and model initialization to replicate the experiments. The CFD models are ran in two stages, with the first stage including only the 23-m reactive zone with the gas barrier intact. The gas barrier is represented as a wall boundary condition. The second stage triggers with the burst of the gas barrier and simulates the full 30.5-m reactor length. As the barrier ruptures, the boundary condition switches from “wall” to “interface.” A patch function is used to pass the initialization parameters to the nonreactive zone. The open end is treated as a “pressure outlet” boundary condition, and the simulation resumes. Fig. 2 Comparison of flame front propagation velocity over distance between experiments and CFD models. Results and discussion Figure 2 shows the comparison of flame front propagation velocity over distance from ignition location between experiments and CFD models. Both 2D and 3D CFD models show similar flame propagation trends that compare well with the experimental results. The flame propagated slowly with the gas barrier intact, and flame velocity significantly increases once the barrier is opened. The 2D CFD model shows a good prediction initially but deviates past 18 m, predicting much smaller flame velocities where the experimental results show a continuous increase. The large error bars at a distance of 15 m from ignition indicate a large fluctuation of velocity and happen right after the gas barrier bursts. This can be related to a strong turbulent flame and pressure oscillations caused by the bursting of the barrier. Overall, the CFD models capture the effect of the gas barrier opening as observed in the experiments, with both models showing a significant increase in flame speed once the gas barrier bursts. Conclusions and future work This research presents a method for developing a flame propagation CFD model in a large-scale reactor and shows how coupling experiments with model development and validation can lead to improved modeling predictive capabilities. Key findings from this study include: The 2D and 3D CFD models show good prediction of the flame behavior inside the empty reactor. Both models are able to capture the confinement effect of the gas barrier and the resulting flame acceleration after the barrier bursts. As expected, the 3D model shows improved agreement with the experimental results compared to the 2D model. However, considering the computational time required to run the simulation, 21 days for the 3D model to simulate 960 ms compared to five days for the 2D model, the 2D model provides an adequate alternative to the 3D model when relatively quick turnaround times are desired to capture general trends in flame propagation velocities. Both models used 4 × 36 cores and a combination of 192-GB and 384-GB RAMs. The 2D model is sufficient in capturing the general trend of the flame propagation and predicts the flame propagation speed reasonably well. However, the 2D model underpredicted the resulting pressure. Pressure comparisons at 820 ms, as the gas barrier ruptures, show that the pressure of the 3D CFD model (approximately 113 kPa) is in good agreement with the experiment (approximately 103 kPa), compared to the 2D CFD model (approximately 56 kPa). Depending on the purpose of the study, such as explosion barrier design, a 3D model is necessary to more accurately simulate flame propagation speed and overpressures observed with the experiments. Selected reference Investigating the impact of rock rubble on methane-air explosions: High-speed turbulent deflagrations and transition to detonations. Short-term innovative and exploratory research projects – follow up research Colorado School of Mines (Grant AFSTI14FO-69) https://www.alpha-foundation.org/outputs-and-impact/hsi-outputs/#ffs-tabbed-12|ffs-tabbed-31

Experimental and numerical investigation of methane flame propagation characteristics in a 30.5-m explosion reactor with obstacles

Authored by:
Gregory Bogin, Jürgen Brune, Erik Charrier, Aditya Juganda, Patrick Maier

To view this article you must be a member of SME/UCA.

Get your SME Foundation VIP Access Conference Pass to MINEXCHANGE 2027

“Check the box” and help support the SME Foundation

The Art and Science of the Flowsheet

Get your SME Foundation VIP Access Conference Pass to MINEXCHANGE 2027

“Check the box” and help support the SME Foundation

The Art and Science of the Flowsheet

To view this article you must be a member of SME/UCA.

Are fatigue and sleepiness the same? A brief introduction to the differences and similarities, and their implications for work safety

Fatal accident analysis and hazards identification in Turkish coal-extracting industry using analytic hierarchy process

Related Articles

Rare earth elements in the global economy: Usage, recovery and the quest for supply security — A review

A systematic evaluation of localization signals in the advancing openpit mine Aitik

From foundation to future: Revisiting AI integration in mining engineering education through current perspectives of students, educators and industry

Iron ore tailings reprocessing by electrostatic separation plus two-stage froth flotation

Get your SME Foundation VIP Access Conference Pass to MINEXCHANGE 2027

“Check the box” and help support the SME Foundation

The Art and Science of the Flowsheet

Get your SME Foundation VIP Access Conference Pass to MINEXCHANGE 2027

“Check the box” and help support the SME Foundation

The Art and Science of the Flowsheet

Use Single Sign On