Three-Dimensional Geological Modeling of Coal Seams Using Weighted Kriging Method and Multi-Source Data

Defu Che,Qingren Jia

doi:10.1109/access.2019.2936811

Defu Che, Qingren Jia

Open Access

https://doi.org/10.1109/access.2019.2936811

Copy DOI

Abstract

Constructing high quality three-dimensional (3D) geological models of coal seams from multi-source data, such as boreholes and geological maps, is a crucial task in seam distribution analysis and production planning of coal mine. However, widely used interpolation methods in geology, such as inverse distance weighting (IDW) and ordinary kriging (OK), ignore the variety of uncertainties in multi-source data. In this paper, sample points extracted from the archival data of a coal mine were classified into two categories, the hard data and the soft data, according to their sources. Field boundaries and fault observations were also obtained from data processing. A 3D geological modeling workflow was developed to integrate these data, in which (1) the weighted kriging (WK) method was used to interpolate both hard data and soft data, and (2) a fault modeling method was proposed to build faults’ geometry as well as their impact on coal seam surface model. The WK method improves the accuracy of interpolation compared with the OK and IDW methods, and the proposed fault modeling method can integrate expert interpretations into the final 3D geological model of the coal seams.

Highlights

Coal still keeps hundreds of years of identified reserve and serves as the largest source of fuel for this generation [1]
To integrate expert knowledge embedded in these data, this paper developed a new fault modeling method, in which a fault surface calculated from the fault horizontal cutoff lines was used to intersect the interpolated continuous coal seam surface to get a zero-displacement line
3) PROPAGATING THE DISPLACEMENT TO THE COAL SEAM SURFACE In a certain domain where the displacement caused by fault occurs, a circular displacement region passing through two pinch-out points and having a diameter equal to the distance between two points was used here (Fig. 7c)

Summary

INTRODUCTION

Coal still keeps hundreds of years of identified reserve and serves as the largest source of fuel for this generation [1]. Most interpolation methods above, including kriging approaches, consider multi-source data as ground truth data, whereas they are much more heterogeneous in both nature and quality [16] When integrating these data into geological models, care is needed to control their uncertainty [17]. Three categories of data from these sources were acquired, including sample points (namely, hard data and soft data), boundaries, and faults horizontal cutoff lines. They serves as the input data of the modeling procedure. Each fault was constructed in a three-phase procedure, namely, calculating the intersection line of the fault and coal seam surfaces, computing the movement of horizontal cutoff lines, and propagating the movement to the coal seam surface model

WEIGHTED KRIGING

MODELING OF FAULTS

CASE STUDY

Findings

CONCLUSION