Identifying the coal seams and their sedimentary environments in the Nui Beo mine, Quang Ninh province using K-means and regression methods

Abstract

The Nui Beo mine is located in the southern part of the Quang Ninh province. It belongs to the Hon Gai-Cam Pha coal zone, where many coal resources have been estimated in Vietnam. Based on synthesizing and processing data of the coal seam parameters and coal quality by using K-means and regression methods, results allow dividing the Nui Beo mine into three areas, the area 1 bearing coal thickness greater than or equal to 18.5m, the area 2 containing coal thickness less than 18.5m, and coal seam dipping angle less than 32.0 degrees; and the area 3 obtaining coal thickness less than 18.5m, and coal seam dipping angle greater than or equal to 32.0 degrees, combining with five interlayers as the scared boundary. Identifying results using K-means and regression methods for 2 or 3 coal seam parameters is very low; only coal seams of V4, V12, and V14 are higher (over 50%). After dividing the mining area, the identifying rate is higher, especially the coal seams of V5, V7, V11, V12, and V14 reach 30.38% on average compared to 16.83% for the whole mining area. Identifying coal seams based on six analyzed parameters shows 16.64% on average; the nine analyzed parameters show 25.93% on average. Generally, each environment has increased significantly after dividing the identifying rates of coal seams according to seam parameters and coal quality. Using the seam characteristics as thickness, dip angle, the number of interlayers, and coal quality, they help divide the Nui Beo mine into 07 relatively homogenous blocks, namely A1, A2, A3, B, C, D, and E, respectively. In which block A1 is formed in a swamp environment; blocks A2 and A3 are characterized by a mudflat environment; block B is suggested a flow environment; block C is characterized by mudflat; block D features a swampy and block E in the periphery. The results confirm that applied mathematical methods in geological science are effective, especially in identifying coal seams and sedimentary environments.