ABSTRACT: In predicting the integrity of under-mined structures, surface horizontal displacements and curvatures are frequently of greater importance than the more homogeneous vertical displacements. Accurate evaluation of mining induced subsidence and horizontal displacement further requires that the predictive model adequately represents the subsurface geology, mining geometry and extraction sequencing. The following documents an extension of the Subsidence Prediction and System Identification method (SPASID) to the evaluation of horizontal strains. Comparisons are made between predictions from SPASID and two other empirical methods against measured data from a case study. The system identification method is illustrated to predict horizontal displacements and strains most consistently with the in situ data.
1. INTRODUCTION
Since most structures are prone to damage by strains, ground horizontal movement may have far greater impact on the stability of structures in comparison with the surface vertical movement (subsidence). The effect of subsidence may be crucial only where the ground water level approaches the surface from which surface structures would be flooded as a result of the extraction of a medium to thick coal seam. Additionally, under subcritical mining situations, surface structures are subjected to high curvatures resulting from the subsidence. Surface structures are affected either by the dynamic horizontal strains when mining is underway, or by the static horizontal strains as the extraction of coal ceases. In either case, once the structures are fractured under excessive strains they lose their stability and integrity permanently. It is therefore important that surface structures are either capable of withstanding the effect of high horizontal strains throughout the period of mining or that these effects may be predicted to allow mitigation in populated areas. The objective of this paper is to illustrate the validity of the extended SPASID program through a case study. Comparisons of surface subsidence and deformation between SPASID predictions and measured data, together with two other methods of prediction, show that the surface horizontal displacement and strain can be predicted correctly provided that accurate surface subsidence profiles have been obtained. The case study also reveals that the SPASID program has the potential to match field curves of ground movement and deformation to a satisfactory degree in comparison with other methods.
2. PREDICTION OF SURFACE SUBSIDENCE: A FIRST STEP IN THE ESTIMATION OF SUBSIDENCE DAMAGE
Because each of the other four components of ground movement and deformation (tilt, curvature, horizontal displacement and horizontal strain) can be derived from the vertical ground subsidence either analytically or empirically, accurate prediction of subsidence plays a key role in the estimation of subsidence damage. There are a variety of methods available for the prediction of subsidence. Relatively few of them, however, possess the capability of accommodating the variations of geology and mining conditions. It is apparent that no current methods are capable of accurately predicting surface subsidence without a priori knowledge of previous mining subsidence. In practice, the first step in subsidence prediction is to determine the subsidence parameters based on in-situ measured data, then calculate the subsidence (post- calculation, because the real subsidence profile is known).
