Intelligent geoscience information mining and knowledge discovery using big data analytics: A case study of the Shangfanggou Mo (Fe) mine in Henan Province

doi:10.13745/j.esf.sf.2022.2.85

Abstract

Abstract:

Industry 4.0 of the 21^st century has given birth to smart mines. The multidisciplinary datasets of smart mines-such as geology, exploration, mining, geometallurgy, environment and survey/map datasets-constitute big data of mines, and they play an important role for the rapid advancements of geoscience in areas of geoscience digitization and application of information/Al technology in geoscience. Taking the Shangfanggou Mo (Fe) mine, a 5G⁺ smart mine, in Henan Province as an example, using big data of mines, this paper carried out geoscience information mining to highlight emerging engineering research with integrated multidisciplinary approach. Innovative results and geological knowledge discoveries from this study are summarized as follows: (1) According to theories on porphyry-associated skarns and mineralogical approach to minera resources prospecting, using borehole datasets and large-scale open-pit mapping and microscopic identification analysis, a 3D temporo-spatial model of the identified key minerals and predicted minerals in the study area was established, and a NE trending ore-bearing fault section and a penetration-type ore-bearing section were discovered. (2) Using UAV remote sensing and ground hyperspectral short-wave/long-wave infrared techniques, more than 20 types of key altered minerals in the study area were delineated, and a 3D multi-parameter mineral model was constructed. (3) Using geochemical techniques such as XRF and in-situ microscopy, a rock dataset with matching hyperspectral interpretation was established, and a dual-matrix mapping software for useful/harmful elements of rocks/ores in the study area was developed. In addition, mathematical modeling combining traditional geostatistics (gauss simulation, kriging interpolation) with machine learning (deep learning) was realized, and the composition of ore blends used between March-April 2021 was identified and the cause of the resulting low recovery rate was clarified. (4) Based on process mineralogy practice in the study area, multi-stage, multi-type mineral processing datasets (>1800 data on quarterly/monthly/daily processing of rock powder, mud powder, concentrate, tailings, etc.) were used to develop rock/mineral powder testing techniques and analysis methods, and the types of refractory ores and harmful minerals in the Shangfanggou Mo mine were identified. The multivariate, multi-type datasets of mines have the “5V” (volume, variety, velocity, veracity, value) characteristics of big data. The accurate management control of dynamic correlation measurement/analysis and rapid/efficient evaluation of big data of mines is conducive to intelligent mining decision-making and improvement of economic benefit (recovery rate). Among them, high-precision multi-parameter 3D modeling can be applied not only to deep mining of geological, structural, alteration and mineralization information models of rocks/ores as well as reserve/resources verification, but also to facilitating 4D control on real-time mining of fourth generation industrial 5G⁺ mines, such as 3D visualization of geological and mineral resources prediction/evaluation/storage expansion, virtual simulation of “year-quarter-month-day” dynamic ore blending and mining, and real-time digital twin for mine beneficiation. The research results provide a reference for in-depth geoscience research on mineral exploration and mineral resources assessment in smart mines.

Key words: mine big data, hyperspectral, XRF, in-situ geochemical analysis, geoscience information mining, knowledge discovery, Shangfanggou Mo polymetallic deposit

CLC Number:

P628
P618.65

WANG Luofeng, WANG Gongwen, XU Wenhui, XU Senmin, HE Yaqing, WANG Chunyi, YANG Tao, ZHOU Xiaojiang, HUANG Leilei, ZUO Ling, MOU Nini, CAO Yi, LIU Zhifei, CHANG Yulin. Intelligent geoscience information mining and knowledge discovery using big data analytics: A case study of the Shangfanggou Mo (Fe) mine in Henan Province[J]. Earth Science Frontiers, 2023, 30(4): 317-334.

Figures/Tables 15

Fig.1 Simplified geological maps showing the location of the study area (a, adapted from [11]), location of the Shangfanggou Mo (Fe) mine (b, adapted from [11]), and tectono-geological setting of the Shangfanggou deposit (c, modified after [11-12])

Fig.2 UAV multi-phase remote sensing images (years 2020, 2021, 2022) and radar point cloud maps of Shangfanggou mine

Fig.3 3D surface images of minding areas (upper panel) ), UAV captured image of the study area showing distribution of sampling points (RTK positioning with cm resolution), and ore specimen (lower panel)

Fig.4 (A) 3D mineral content modeling of ore standard samples for lithofacies classification based on 108 borehole databases, (B) sampling locations of rock/ore standard samples and layer samples in the study area, and (C) short-wave infrared stope location map.

Fig.5 Relative percentages of TSA minerals in rock powder samples by short-wave (a) and long-wave (b) infrared spectroscopy

Fig.6 3D geological modeling of the study area using GOCAD software

Fig.7 3D control of geological bodies at the 5G+ control room, Shangfanggou mine, Luoyang Luanchuan Molybdenum Mining Co., Ltd., showing oxidized Mo ore (main ore body) (yellow), 2021 mined (blasted) ore block (blue), wall-rock (gray), and 3D multivariate information model of sampling points

Fig.8 3D Kriging interpolation model of the Shangfanggou Mo deposit

Table 1 Mineral composition (by TIR, XRF) dataset (part) of rock/ore powder samples from Shangfanggou deposit

编号	各主成分矿物TIR(热红外)光谱分析结果						w(Fe)/%	w(Mo)/%	w(W)/10^-6
编号	第一主成分	含量	第二主成分	含量	第三主成分	含量	w(Fe)/%	w(Mo)/%	w(W)/10^-6
02-11-1	石英	0.631	白云石	0.369			4.019 8	0.075 2	75
02-11-2	石英	0.658	蛇纹石	0.342			3.737 5	0.101 6	67
02-11-3	石英	0.622	蛇纹石	0.378			4.412 4	0.120 4	106
02-12-1	绿脱石	0.397	石英	0.319	微斜长石	0.284	4.579 4	0.141 9	73
02-12-2	绿脱石	0.422	石英	0.300	微斜长石	0.278	4.171 5	0.124 2	91
02-12-3	绿脱石	0.462	石英	0.291	微斜长石	0.247	5.220 6	0.139 8	117
02-13-1	绿脱石	0.462	石英	0.275	微斜长石	0.262	5.013 5	0.119 3	89
02-13-2	绿脱石	0.409	石英	0.308	微斜长石	0.283	4.734 5	0.130 7	82
02-14-1	拉长石	0.361	石英	0.339	滑石	0.3	4.237 8	0.094 7	76
02-14-3	绿脱石	0.382	石英	0.339	微斜长石	0.279	4.039 0	0.112 3	88
02-14-2	绿脱石	0.406	石英	0.308	微斜长石	0.287	6.733 2	0.133 0	113
02-15-1	绿脱石	0.373	石英	0.332	微斜长石	0.294	6.733 2	0.133 0	113
02-15-2	石英	0.573	蛇纹石	0.427			6.733 2	0.133 0	113
03-31-1	绿脱石	0.569	蛇纹石	0.261	石英	0.17	9.198 9	0.129 6	99
03-31-2	滑石	0.459	角闪石	0.294	钠钙长石	0.248	10.623 5	0.124 4	140
03-31-3	滑石	0.355	角闪石	0.349	拉长石	0.297	10.623 5	0.124 4	140
04-01-1	滑石	0.484	角闪石	0.355	石英	0.161	10.724 8	0.127 6	137
04-01-2	滑石	0.424	角闪石	0.423	石英	0.153	10.557 7	0.125 8	117
04-01-3	滑石	0.519	角闪石	0.306	石英	0.175	10.129 6	0.134 0	124
04-02-1	滑石	0.463	角闪石	0.332	钠钙长石	0.204	11.473 4	0.130 1	121

Fig.9 Heap map comparing elemental composition of powder samples from Shangfanggou mine between January-February 2021

Fig.10 Heap map comparing elemental composition of low-recovery powder samples from Shangfanggou mine between March-April 2021

Fig.11 Geological information mining and decision making process based on ore processing result for high-talc, medium-oxidation, high-Fe, high-Mo ore from the Shangfanggou deposit

Fig.12 Knowledge discovery-based 3D geological models of the Fe-rich section (left) and phlogopite-bearing skarn (right) in the mining area showing their associative relationship

Fig.13 3D Kriging interpolation model and fault model of the Shangfanggou Fe deposit

Fig.14 Hazard-level distribution model of altered minerals based on mineral alteration status and information on mining, mineral processing and harmful minerals

References 48

[1]	YANG L. Industry 4.0: a survey on technologies, applications and open research issues[J]. Journal of Industrial Information Integration, 2017, 6: 1-10. DOI URL
[2]	赵鹏大. 大数据时代数字找矿与定量评价[J]. 地质通报, 2015, 34(7): 1255-1259.
[3]	赵鹏大, 陈永清. 数字地质与数字矿产勘查[J]. 地学前缘, 2021, 28(3): 1-5. DOI
[4]	赵鹏大. 地质大数据特点及其合理开发利用[J]. 地学前缘, 2019, 26(4): 1-5. DOI
[5]	郭华东. 地球大数据科学工程[J]. 中国科学院院刊, 2018, 33(8): 818-824.
[6]	周永章, 陈烁, 张旗, 等. 大数据与数学地球科学研究进展[J]. 岩石学报, 2018, 34(2): 255-263.
[7]	王功文, 张寿庭, 燕长海, 等. 栾川矿集区矿产勘查大数据挖掘与三维/四维建模的资源-环境联合预测与定量评价[J]. 地学前缘, 2021, 28(3): 139-155. DOI
[8]	王功文, 张智强, 李瑞喜, 等. 华北重点矿集区大数据三维/四维建模与深层次集成的资源预测评价[J]. 中国科学: 地球科学, 2021, 51(9): 1594-1610.
[9]	CODE J. Australasian code for reporting of exploration result, mineral resources and ore reserves. The joint ore reserves committee of The Australasian Institute of Mining and Metallurgy, Australian Institute of geoscientists and Minerals Council of Australia[S]. (2012-2-20) [2021-2-30]. http://www.jorc.org.2012.
[10]	周晓将, 毕惠珍. 上房沟钼铁矿区高滑石型难选矿石配矿管理方法采矿技术[J]. 矿业研究与开发, 2021, 21(1): 170-172.
[11]	LIU Z, ZUO L, XU S, et al. 3D Multi-parameter geological modeling and knowledge findings for Mo oxide orebodies in the Shangfanggou porphyry-skarn Mo(-Fe) deposit, Henan Province, China[J]. Minerals, 2022, 12(6): 769. DOI URL
[12]	简新玲, 张书彦, 庞运超, 等. 河南省栾川县上房沟矿区钼铁矿资源储量核实报告[R]. 郑州: 河南省地质矿产勘查开发局地球物理勘查队, 2012.
[13]	孙大勇, 肖舜元, 祁忠旭, 等. 洛钼上房沟高滑石型难选钼矿选矿试验研究[J]. 矿业研究与开发, 2019, 39(3): 43-47.
[14]	王玲, 赵战锋. 工艺矿物学在地质冶金学中的应用及问题[J]. 矿产综合利用, 2020, 2: 37-43.
[15]	郝跃. 智能矿山大数据关键技术与发展研究[J]. 世界有色金属, 2022, 592(4): 15-17.
[16]	黄蕾蕾. 内蒙古乌努格吐山矿山高精度三维地质建模与评价[D]. 北京: 中国地质大学(北京), 2020.
[17]	王功文, 张寿庭, 陈建平, 等. 大型-超大型矿床定量评价技术手册[S]. 北京: 地质出版社, 2019.
[18]	王功文, 张寿庭, 燕长海, 等. 基于地质与重磁数据集成的栾川钼多金属矿区三维地质建模[J]. 地球科学: 中国地质大学学报, 2011, 36(2): 360-366.
[19]	杨艳, 张静, 杨永飞, 等. 栾川上房沟钼矿床流体包裹体特征及其地质意义[J]. 岩石学报, 2009, 25(10): 2563-2574.
[20]	瓮纪昌, 高胜淮, 石聪, 等. 栾川上房沟特大型钼矿床蚀变分带规律研究[J]. 中国钼业, 2008, 139(3): 16-24.
[21]	韩江伟, 云辉, 胡红雷, 等. 河南栾川矿集区深部钨钼矿体特征及资源预测[J]. 金属矿山, 2020, 533(11): 141-151.
[22]	何亚清, 杜红波, 彭府华. 三道庄矿露天和地下岩体工程灾害监测预警应用[J]. 有色地质, 2017, 69(4): 81-85.
[23]	BOOYSENREN E, LORENZ S, THIELE S T, et al. Accurate hyperspectral imaging of mineralised outcrops: an example from lithium-bearing pegmatites at Uis, Namibia[J]. Remote Sensing of Environment, 2022(269): 112790.
[24]	ALI S H, GIURCO D, ARNDT N, et al. Mineral supply for sustainable development requires resource governance[J]. Nature, 2017, 543(7645): 367-372. DOI
[25]	JAKOB S, ZIMMERMANN R, GLOAGUEN R. The need for accurate geometric and radiometric corrections of drone-borne hyperspectral data for mineral exploration: MEPHySTo: a toolbox for pre-processing drone-borne hyperspectral data[J]. Remote Sensing, 2017, 9(1): 88. DOI URL
[26]	JAMES M R, ROBSON S, D’OLEIRE-OLTMANNS S, et al. Optimising UAV topographic surveys processed with structure-from-motion: ground control quality, quantity and bundle adjustment[J]. Geomorphology, 2017, 280: 51-66. DOI URL
[27]	KRUPNIK D, KHAN S. Close-range, ground-based hyperspectral imaging for mining applications at various scales: review and case studies[J]. Earth Science Review, 2019, 198: 102952. DOI URL
[28]	GALLWEY J, ROBIATI C, COGGAN J, et al. A Sentinel-2 based multispectral convolutional neural network for detecting artisanal small-scale mining in Ghana: applying deep learning to shallow mining[J]. Remote Sensing of Environment, 2020, 248: 111970. DOI URL
[29]	YANG Y, CHEN Y J, ZHANG J, et al. Ore geology, fluid inclusions and four-stage hydrothermal mineralization of the Shangfanggou giant Mo-Fe deposit in Eastern Qinling, Central China[J]. Ore Geology Review, 2013, 55: 146-161. DOI URL
[30]	常瑜琳, 王功文, 黄蕾蕾, 等. 用于矿山岩粉矿物光谱测试分析的压片设备: CN218330841U[P]. 2023-2-17.
[31]	HOULDING S W. 3D geoscience modeling-computer techniques for geological characterization[M]. Berlin: Springer-Verlag, 1994.
[32]	MALLET J L. Discrete modeling for natural objects[J]. Mathematical Geology, 1997, 29(2): 199-219. DOI URL
[33]	MARJORIBANKS R. Geological methods in mineral exploration and mining[M]. Second Edition. New York: Springer, 2010.
[34]	HUANG L, WANG G, CARRANZA E J M, et al. Multi-scalenumerical simulation and 3D modeling for deep mineral exploration in the Jiaojia gold district, China[J]. Natural Resources Research, 2020, 29: 415-438. DOI
[35]	JORDAN M I, MITCHELL T M. Machine learning: trends, perspectives, and prospects[J]. Science, 2015, 349(6245): 255-260. DOI PMID
[36]	KERR R A. The coming copper peak[J]. Science, 2014, 343(6172): 722-724. DOI PMID
[37]	KREUZER O P, YOUSEFI M, NYKÁNEN V. Introduction to the special issue on spatial modelling and analysis of ore-forming processes in mineral exploration targeting[J]. Ore Geology Reviews, 2020, 103391.
[38]	MOGOLLON C U, MAHER K. White mica geochemistry: discriminating between barren and mineralized porphyry systems[J]. Economic Geology, 2020, 115(2): 325-354. DOI URL
[39]	WANG G, ZHANG Z, LI R. et al. Resource prediction and assessment based on 3D/4D big data modeling and deep integration in key ore districts of North China[J]. Science China: Earth Sciences, 2021, 64(9): 1590-1606. DOI
[49]	KOKALY R F, CLARK R N, SWAYZE G A, et al. USGS spectral library Version 7[R]. Denver: Geological Survey Data Series, 2017(1305): 61.
[40]	KOKKALA A M. An engineering geological database for managing, planning and protecting intelligent cities: the case of Thessaloniki city in Northern Greece[J]. Engineering Geology, 2022(301): 106617.
[41]	NEWMAN C, POULSON S R, MCCREA K W. Contaminant generation and transport from mine pit lake to perennial stream system: multidisciplinary investigations at the Big Ledge Mine, Nevada, USA[J]. Geochemistry, 2020(80): 125552.
[42]	BOX G E P. Science and statistics[J]. Journal of the American Statistical Association, 1976, 71: 791-799. DOI: 10.1080/01621459.1976.10480949. DOI URL
[43]	DUTTA S, BANDOPADHYAY S, GANGULI R. Machinelearning algorithms and their application to ore reserve estimation of sparse and imprecise data[J]. JILSA, 2010, 2: 86-96. DOI: 10.4236/jilsa.2010.22012. DOI URL
[44]	NEZAMOLHOSSEINI S A, MOJTAHEDZADEH S H, GHOLAMNEJAD J. Theapplication of artificial neural networks to ore reserve estimation at Choghart Iron Ore Deposit[J]. Analytical and Numerical Methods in Mining Engineering, Special Issue, 2017, 6: 73-83.
[45]	THIELE S T, LORENZ S, KIRSCH M, et al. Multi-scale, multi-sensor data integration for automated 3-D geological mapping[J]. Ore Geology Reviews, 2021(136): 104252.
[46]	YIN B, ZUO R, XIONG Y, et al. Knowledge discovery of geochemical patterns from a data-driven perspective[J]. Journal of Geochemical Exploration, 2021(231): 106872.
[47]	CHOI Y, NGUYEN H, BUI X, et al. Optimization of haulage-truck system performance for ore production in open-pit mines using big data and machine learning-based methods[J]. Resources Policy, 2022(75): 102522.
[48]	BĖRUB C L, OLIVOB G R, CHOUTEAU M, et al. Predicting rock type and detecting hydrothermal alteration using machine learning and petrophysical properties of the Canadian Malartic ore and host rocks, Pontiac Subprovince, Québec, Canada[J]. Ore Geology Reviews, 2018, 96: 130-145. DOI URL