We use cookies to enhance your experience. By continuing to browse this site you agree to our use of cookies. More info.
China has 36 coal mines with a production capacity of more than 10 Mt/a, with a total capacity of 612 Mt/a. Three of these mines, Hongliulin, Zhangjiamao, and Ningtiaota coal mines, are located in the Shennan mining region.
Image Credit: Parilov/Shutterstock.com
The Ningtiaota Coal Mine is a massive and major coal mine with a production capacity of 12 Mt/a. This mine has piqued the interest of many scholars because of its unique characteristics and representativeness. Their findings served as a model for other coal mines’ water resource management and water inrush control.
Researchers used Bayes discriminant analysis and Fisher discriminant analysis (FDA) to create a mine-water source-identification model, and they evaluated the different models’ relevant circumstances, sample size, accuracy, and identification capabilities.
Mine-water inflow source discrimination has a positive impact on water resource preservation, development, and usage, coal mine safety, and ecological and environmental protection. It can also give theoretical assistance for developing water-conserving coal-mining designs and preventing water inflows.
Water inrush in mines, in particular, causes more catastrophic and frequent disasters, posing serious hazards to human life and property. The weathering zone of the bedrock was the source of the water inrush. The principal source of water inrush was the weathered bedrock aquifer, which reached the working face through the caving cracks.
The conventional hydrochemistry technique, the isotope method, the tracer test method, the trace element method, and the water temperature and water level method are all examples of hydrogeochemical characteristic analysis methods. To immediately determine the source of water inrush, laser-induced fluorescence technologies, GIS, and temperature analysis have been implemented.
For the first time, a self-organizing feature map (SOM) was used to distinguish mine-water inrush sources in this investigation. From a quantitative and qualitative standpoint, the FDA, the Piper diagram, the Gibbs diagram, and the water temperature were also employed to examine and validate the water input source.
The origins of the unidentified water samples were discovered. When water gushing or indicators of water gushing develop in a mine, real-time detection should be used to swiftly determine the kind of water source.
The study of regional hydrogeochemical impacts and the replenishment and drainage interaction between groundwater and surface water would benefit greatly from this research. This study provides a solid foundation for mine-water hazard prediction and water source forecasting and prevention. It provides theoretical support as well as a practical foundation for policy formation and the mitigation of mine-water hazards.
Ningtiaota Coal Mine is located in Shenmu County, Yulin City, Shaanxi Province, and is in the southern section of the Shennan mining area (38°57′24′′–39°07′57′′ N, 110°09′29′′–110°16′23′′ E) (Figure 1). The geological structure of this region is simple; the whole area is a monoclinic layer oriented west, with no faults or magmatic rocks visible. Figure 1 depicts a stratigraphic map of the area.
Figure 1. Map of the study area and the stratigraphic map ((a) Geographical location of the coal mine. (b) Hydrogeologic profile. (c) stratigraphic map). Image Credit: Zhao, et al., 2022
Quaternary loose-pore phreatic aquifers (Q), weathered-fissure water of layered clastic rocks from the Middle Jurassic Anding Formation (J2a) and Zhiluo Formation (J2z), fissure-confined water from the Middle Jurassic Yan’an Formation (J2y), and phreatic aquifers of burnt-rock fissure holes are the most common water-filled aquifers (Figure 2).
Figure 2. Hydrogeological profile. Image Credit: Zhao, et al., 2022
The concepts of dimensionality reduction and projection are used in Fisher discriminant analysis. Its goal is to project high-dimensional training-sample data in an appropriate direction onto a low-dimensional environment.
Figure 3 shows a neural matrix depiction of the individual indices in the SOM. TDS, Cl–, K+ + Na+, Ca2+, Mg2+, and SO42- all showed the same color gradient, indicating that the distributions of these ions were highly comparable and that the correlations between them were quite strong.
Figure 3. Neural matrix visualization of the SOM for individual indexes—pH, TDS, HCO3−, SO42−, Cl−, K+ + Na+, Ca2+, Mg2+, and NO3−—in all samples. Image Credit: Zhao, et al., 2022
A SOM neural network clustering pattern classification map with five clusters was created by clustering based on the shortest DBI index (Figure 4).
Figure 4. SOM neural network clustering pattern classification map. Image Credit: Zhao, et al., 2022
To examine the hydrogeochemical features of the samples, AqQA was used to create a Piper trilinear diagram (Figure 5).
Figure 5. Piper trilinear diagram. Image Credit: Zhao, et al., 2022
Gibbs diagrams were also used to separate natural water into three primary categories based on its ionic components: evaporation dominance, rock-weathering dominance, and precipitation dominance (Figure 6).
Figure 6. Gibbs diagrams. Image Credit: Zhao, et al., 2022
Both discrimination functions were used to compute the geographical distribution of water samples. Figure 7 shows the water sample types as Quaternary pore water (P), Middle Jurassic Yan’an Formation fissure water (J2y), surface water (S), and Middle Jurassic Zhiluo Formation fissure water (Zhiluo) (J2z).
Figure 7. Spatial distribution of water sample discriminant function. Image Credit: Zhao, et al., 2022
Table 1 shows the distances between gushing water and the centroids of the four water sample types.
Table 1. Distance of gushing water to the centroid. Source: Zhao, et al., 2022
The self-organizing feature map was developed as a simple and rapid way of identifying water sources in this study.
SOM determined that the sources of G1, G2, G3, G4, G5, and G6 were all surface water based on the above investigation. The hydrogeochemical approaches yielded the same results. Because the distance between aquifers, external interference, and the number of samples all had a substantial impact on water temperature, this can only be used as an auxiliary identification approach.
The gushing water sample G4 was recognized as Quaternary pore water by using FDA. G4 and G2 were quite close in terms of hydrogeological features, and the two-sample stations were in the same aquifer. In the result, surface water was determined to be the source of all the mine gushing water samples.
To define the hydrogeochemical features of mine gushing water in Ningtiaota Coal Mine, 11 indicators, comprising pH, TDS, T, and the principal ions, were determined in 40 water samples by field studies and experimental testing.
A self-organizing feature map was used in this study to determine the source of mine gushing water. The hydrogeochemical features of water samples G1–G6 were compared to those of the surface-water cluster, and they were recognized as surface water. As supplementary measurements, FDA, water temperature, and classic hydrogeochemical discrimination, such as Piper and Gibbs diagrams, were used to confirm the SOM analysis.
The SOM method offers significant advantages in terms of data visualization, dimensionality reduction, and grouping, and it is particularly well suited to source analysis with large training samples. It may be used to not only measure a source but also to assess the hydrogeochemical features of a sample at the same time.
As a consequence of the reduction of human influence in sample sets, the findings of source classification of mine gushing water have high efficiency and precision. In conclusion, this study establishes a scientific foundation for mine-water hazard prediction as well as water source anticipation and control. It offers theoretical support as well as a practical foundation for policy formation and the prevention of mine-water hazards.
Zhao, D., Zeng, Y., Wu, Q., Du, X., Gao, S., Mei, A., Zhao, H., Zhang, Z. and Zhang, X. (2022) Source Discrimination of Mine Gushing Water Using Self-Organizing Feature Maps: A Case Study in Ningtiaota Coal Mine, Shaanxi, China. Sustainability, 14(11), p.6551. Available Online: https://www.mdpi.com/2071-1050/14/11/6551/htm.
Written by
Laura Thomson graduated from Manchester Metropolitan University with an English and Sociology degree. During her studies, Laura worked as a Proofreader and went on to do this full time until moving on to work as a Website Editor for a leading analytics and media company. In her spare time, Laura enjoys reading a range of books and writing historical fiction. She also loves to see new places in the world and spends many weekends looking after dogs.
Please use one of the following formats to cite this article in your essay, paper or report:
APA
Thomson, Laura. (2022, June 06). Research Highlights Scientific Framework for Forecasting Mine-Water Risk. AZoMining. Retrieved on June 06, 2022 from https://www.azomining.com/Article.aspx?ArticleID=1684.
MLA
Thomson, Laura. "Research Highlights Scientific Framework for Forecasting Mine-Water Risk". AZoMining. 06 June 2022. <https://www.azomining.com/Article.aspx?ArticleID=1684>.
Chicago
Thomson, Laura. "Research Highlights Scientific Framework for Forecasting Mine-Water Risk". AZoMining. https://www.azomining.com/Article.aspx?ArticleID=1684. (accessed June 06, 2022).
Harvard
Thomson, Laura. 2022. Research Highlights Scientific Framework for Forecasting Mine-Water Risk. AZoMining, viewed 06 June 2022, https://www.azomining.com/Article.aspx?ArticleID=1684.
Do you have a review, update or anything you would like to add to this article?
Cancel reply to comment
This product profile outlines the Ramsey Series 30 Belt Scale System.
ZEISS Xradia 510 Versa x-ray microscope advances industry and science to address emerging research challenges.
Discover the PerGeos Digital Rock Software from Thermo Fisher Scientific.
Prof. James Tour
AZoMining speaks with James Tour from Rice University about his team's exciting method that has the very real potential to recover valuable REE from three types of waste; electronic waste, coal fly ash and bauxite residue.
Craig Liddicoat
Craig Liddicoat from Flinders University speaks to AZoMining about his team's work on effective rehabilitation in post-mining ecosystems and whether it can set up a predictable trajectory of recovery. We also ask why ecosystem recovery is important in mining operations and the current limitations af
Andrew Somers
In this interview, Andrew Somers from SciAps Australia talks to AZoMining about the new Geochem Pro application for the Z range of handheld elemental analyzers, and the improvements that have been made to their technology over the last year.
AZoMining.com – An AZoNetwork Site
Owned and operated by AZoNetwork, © 2000-2022
Research Highlights Scientific Framework for Forecasting Mine-Water Risk – AZoMining
