Coaltech makes a Breakthrough in Underground Coal Mining Ventilation: A Validation Study2016-11-09 Karl Du Plessis
In the previous blog article, ESTEQ Assists Coaltech in Creating Safer Underground Coal Mining Environment, we discussed how Coaltech utilises FloEFD to perform underground coal mining ventilation studies. This article is a continuation of that work and presents some revelations and detailed findings based on literally hundreds of different scenarios that have been simulated thanks to the ease of use of the FloEFD technology. Comparisons with actual measurements as part of a verification process are also shown which provided Coaltech with confidence in the use of FloEFD as the CFD tool of choice to aid smarter ventilation system designs in the future. Please refer to the previous article for an introduction to the topic. In that article we mentioned that a recent series of methane ignitions (thankfully not explosions) has led to the research being conducted.
The reason for the sudden increase in methane ignitions can on one hand be ascribed to increased production demands and on the other hand, poor ventilation design due to little to no insights as to how the complete system is going to interact. Increased coal demand results in larger machines capable of mining more coal at higher rates, resulting in more coal dust that has to be removed and resulting in larger ventilation, dust suppression and scrubber systems. It is with the increased ventilation requirements that the problem arises. Merely increasing the flow rates of the various systems can produce unpredictable behaviour and have a negative impact on the dilution of methane from the cut face as the system can become imbalanced. Furthermore, careful consideration has to go into matching the different auxiliary ventilation systems, i.e. matching a suitable scrubber arrangement with either jet fan, scoop brattice or ventilation ducting arrangement for the supply of fresh air. Consider the figures below showing the effect of an imbalanced (top) system versus a better balanced (bottom) system. Figure 1 shows velocity streamlines. It is quite clear that in the case of the imbalanced system the scrubber flow cuts off the fresh air supply from the duct resulting in a large amount of recirculation of the contaminated air and the build-up of methane, as is evident in the iso-surfaces showing the methane concentrations in Figure 2. In general the methane concentrations are lower for the balanced system as the surfaces representing the different levels of concentration encapsulates smaller volumes. Note that in these particular simulations the rotation of the cutting drum was not considered, hence the methane being trapped between the drum and the cut-face. Of course the rotation of the drum will aid in washing out the methane between the face and the drum, provided sufficient and effective ventilation is ensured. This can also be simulated in FloEFD through the sliding mesh rotation function (refer to the series of articles under “CFD Simulation of complex rotor equipment using FloEFD“ at the bottom of this article, pertaining to the different approaches for modelling rotating bodies in FloEFD). Figure 3 shows a similar scenario (with a different heading and CM configuration) where the rotation of the cutting drum was simulated. The images are mere static screenshots of the animations from the transient analyses conducted, but one can immediately identify the build-up of methane for the imbalanced system. Poor penetration of the fresh air supply results in high levels of recirculation of contaminated air from the scrubber to yield the build-up of methane at the cutting face, regardless of the rotation of the cutting drum. The bottom image shows much better dilution of methane. Just to clarify, the slightly higher local methane concentration in the bottom image is attributed to the physical time at which the simulations were stopped. While the imbalanced case showed increasing methane concentrations up to this point the balanced case had reached steady conditions by the same time. Now, it is all good and well that one can simulate complex situations like these, but does it have any merit? Can simulations like these really be utilised to reliably predict what would happen under certain circumstances? Which brings us to the next point, that of verifying the results.
Figure 1. Velocity streamlines of imbalanced (top) vs. balanced (bottom) systems.
Figure 2. Iso-surfaces of methane concentration for imbalanced (top) vs. balanced (bottom) systems.
Figure 3. Iso-surfaces and velocity vectors for imbalanced (top) vs. balanced (bottom) systems – Cutting drum rotation considered.
Breakthrough in mining ventilation:
In order to verify the CFD results with any measurements it is necessary to establish a quantifiable and measurable parameter. Considering the flow trajectory images from above it is easy to imagine it to be quite difficult to take velocity measurements in order to determine the flow profile, or even doing some form of smoke tracing as one very quickly ends up with no more than a cloud of smoke. In this respect one has to rely on CFD in order to investigate the interactions of various ventilation system configurations. Coaltech have through an extensive number of simulations, made possible by the ease of use of FloEFD, derived a novel method of determining the actual amount of recirculation. Which surely sounds a lot easier than it really is, since it is not immediately apparent exactly how much of the fresh air actually penetrates the cutting region. By performing CFD simulations and making use of the extensive post-processing features available, the amount of recirculation can be calculated and the methane build-up can be estimated subsequently. Consider the comparison below of the methane build-up as predicted by the recirculation equation and the actual methane concentration extracted directly from the CFD Results.
Figure 4. Methane build-up for various recirculation ratios.
Verification of results:
It is only through verification that high fidelity simulations can be conducted. Thus, in order to do the verification, a physical underground test was conducted, with controlled levels of methane being released and the cutting head is moved up and down simulating typical operation. Comparing the predicted methane levels with actual underground measurements is a more reliable and quantifiable way of verifying the CFD results, as opposed to taking point velocity measurements in order to derive a flow profile. The scenario considered was with the CM in the right hand sump position, similar to Figures 1 to 3 above.
Consider figure 5 below showing a time-history graph extract from the CM telemetry during the underground test. The black line represents the cutting head position as it moves from top to bottom. The blue and red lines are the methane levels detected from the sensors on the left and right side of the cutting head respectively. Two observations are made; one is that the left sensor (blue) constantly measures higher levels than the right side (red), indicating the right to left motion induced by the auxiliary ventilation systems, the other is that the left sensor level peaks whenever the cutting head is in the bottom position (circled), indicating the methane getting trapped under the head as the head moves down. From the graph it is clear that the process is inherently transient and conducting such a simulation would pose some challenges. However, it turns out that one can do quasi-transient simulations with reasonable accuracy in terms of the methane concentrations measured at the sensors by simulating the transient rotation of the cutting drum but with the head fixed in either the top or bottom cutting positions.
Figure 5. Underground test CM telemetry. Right hand sump scenario.
Consider Fig 6 below which shows the graph of methane build-up over time from the CFD simulation. The graph shows the methane levels as averaged across the sensors located all around the CM. Also displayed are the methane levels as calculated from datasheets also averaged across all the sensors. One can see that the CFD predictions compare reasonably well with a slight over prediction when the head is in the bottom position and conversely a slight under prediction when the head is in the top position. This can be attributed to the difference between the quasi-transient modelling approach and the physically transient nature of the cutting sequence. In the simulations the head is kept in a fixed position for a longer period of time allowing for more methane build-up when the head is down and more methane dilution when the head is up. When the head is down the methane gets trapped because the scrubber cannot effectively draw the contaminated air in as the inlet is obscured causing the build-up of methane below the cutting head which results in the higher averaged methane levels. When the head is up the scrubber inlet is no longer obscured and the contaminated air gets drawn in more effectively, thus aiding in the dilution of methane. Thus, when comparing the combined effect from the head up and down positions, one can see a really good correlation between the simulated and measured results.
Figure 6. Verification. Average methane levels on sensors comparison.
These results prove that FloEFD is very capable of producing accurate and reliable results even with such complex systems as underground mining ventilation. The various innovative and advanced technologies within FloEFD have enabled coaltech to consider the entire system of primary and auxiliary ventilation while maintaining a high level of complexity in terms of the geometry of the CMs, scrubbers, shuttle cars and jet fans etc. Being able to provide reliable results with confidence has started a paradigm shift in underground mining ventilation which will lead to more integrated and involved designs where the CM and the ventilation systems will be designed as a complete integrated system from the ground up. Furthermore, smarter designs in terms of early detection of methane have now been made possible by utilising CFD to identify the most suitable placement of methane sensors, which will not be possible without modelling the entire system.
Links relating to the series on complex rotor equipment: