DEVELOPMENT OF MINE SUBSIDENCE PREDICTION MODEL FOR INDIAN COALFIELDS

Abstract

Subsidence is an unwanted phenomenon. It may cause damage to surface and subsurface structures of both kinds; manmade and natural, unless proper subsidence control measures are taken. Subsidence control measures may be taken in three stages: (i) prediction, (ii) prevention, and (iii) protection. The effectiveness of preventive and protective measures greatly depend upon the accuracy of the prediction. The empirical methods are quick, simple to use and yield satisfactory results but they are site specific i.e. their use is restricted to the areas for which they have been developed or to areas having identical geological and mining conditions. The theoretical methods, though more fundamental in approach, have main handicap in general availability of accurate rock properties and geological data which are usually difficult and expensive to obtain. In order to overcome disadvantages of these two approaches, a hybrid approach of carrying out parametric studies by numerical modeling and using its conceptual results in field database to develop a semi-empirical model has been adopted in the present work. The objectives of the present study were : 1) To study the effects of various subsidence contributing parameters by a theoretical approach and to enhance / consolidate the present knowledge of subsidence mechanism. 2) To develop a practical and comprehensive subsidence prediction model with reference to Indian coalfields in particular and of global application in general. (ii) The semi empirical model thus proposed follows the following sequential steps : prediction of maximum possible subsidence, prediction of maximum subsidence for any specified mine geometry, prediction of subsidence profile and subsequently subsidence trough, and prediction of slope and horizontal strain profiles. The effects of various subsidence contributing factors were studied by considering two elastic models (isotropic and transversely Isotropic). Two computer programs MSEAMS, [Crouch, 1976] and MULSIM/BM [Beckett and Madrid, 1988] based on Displacement Discontinuity Method [Sinha, 1979],- a subvariation of Boundary Element Method have been used in present studies. MSEAMS is a two dimensional BEM program which assumes the rockmass in the overburden as a homogeneous, linearly elastic and transversely isotropic medium. MULSIM/BM is a three dimensional Boundary Element program which also assumes the overburden rockmass as homogeneous, linearly elastic but isotropic model. For these models the values of Young's moduli were obtained from an empirical correlation suggested by sheorey [1989] and modified by the author. The values of shear moduli were obtained by back analysis from program MSEAMS such that subsidence values equal to those measured in the field were simulated. The values of G/Ey ratio thus obtained were used to classify the overburden rockmass from subsidence point of view. A Table giving such classification based on G/E ratio and condition of rockmass has been presented. The model studies showed that the results obtained through these programs were only qualitative and the results obtained through transversely isotropic model were more realistic and closer to measured value than those obtained through isotropic models. The isotropic medium (iii) normally does not deform sufficiently to simulate desired displacements. They produced shallower and flatter subsidence trough than observed in the field. Effects of each of the subsidence contributing factors on maximum subsidence S or maximum possible subsidence S has been evaluated bv u max J varying the value of that factor in the model and keeping other factors constant. The analysis of these qualitative results served as guideline for quantitative assessment of subsidence by considering the effect of each individual factor separately using the field database. The effect of the nature of overburden rockmass was simulated by the use of a Rock factor. Rf. The value of Rf depends not only on the composition of rockmass in terms of percentage of hard rock layers in it but also on its condition represented by the 'disturbances' in it either due to natural discontinuities or the degree of fragmentation caused by repeated workings or by both. Chart for obtaining the value of R for a given overburden rockmass has been developed for five different classifications of rockmass based on the degree of discontinuities present. A formula for Smax has been presented which includes the effects of seam thickness, depth of workings, goaf support, extraction ratio, other working seams and condition of overburden rockmass, time and dip of the seam. Numerical values have been assigned to these factors (except for time factor, which was ignored, since the cases of finished subsidence were considered) with the help of field data base. Parametric studies have also been carried out for obtaining subsidence profile. The effects of width-depth ratio and overburden (R ) on the parameters controlling the shape of the subsidence profile have been studied. The effects of width/depth ratio and R on the angle of draw, subsidence over ribside, and the distance of point of inflection was studied by numerical modelling. The relationship between angle of draw and width-depth ratio for different ranges of R were established from field database as well. Similarly relationship was obtained between Rf and a profile constant M, which controls the shape of the subsidence (iv) tjj profile. These relationship help in determining the values of angle of draw (on static end and imping face end as well) and proportionality constant M. Slopes or tilts of the subsidence profile were calculated from subsidence values and the horizontal displacements were obtained with the help of a proportionality constant B which relates slopes to horizontal displacements. The values of 3 were correlated to R values for given width/depth ratios. The horizontal strains were then found out from the calculated horizontal displacements. Field data from 125 mine workings was used in these studies. The field data of 65 mine workings from longwall and bord-and-pi 1lar mines were used to establish correlation between Rf and percentage of hard rock layers in the overburden and rest of 60 mine workings were used to validate the proposed semi-empirical model. The results of application of the model in coalmines in U.S.A and U. K have also given satisfactory results. i ' Subsidence, slope and strain profiles have been predicted from the proposed profile function for Indian coalfields. Comparison of these predicted profiles with those obtained from field measurements and a few other profile functions has been given for 8 panels of different coal mines. These comparisons show that the proposed profile function gave the profiles closer to those measured in the field. The model presented here is easy to understand, simple to use in the field, consists of simple but comprehensive correlations and does not require expensive data and complex analysis; and at the same time it is supported by theory of elasticity. The model has been tested for 60 coal mine-workings in India and 22 longwall mines in North Appalachian basin and compares favourably with NCB predictions in 16 longwall British coal mines. The satisfactory predictions of subsidence, slope and strains.giving results comparable to field measurements validate the prediction mpdel proposed in the present work. t •-'..• •