Study On The Mining Subsidence Prediction Model Of Mining Area With Thick Alluvium

Surface subsidence caused by coal mining always cause for concern, the law of strata andground movement caused by coal mining under normal conditions has been understanded deeply.However, in China since the80’s20th century, many coal mining areas covered with a thickalluvium shows some unique phenomena, for example, subsidence coefficient is greater than1.0,range of surface subsidence is even larger, surface horizontal displacementis greater than the surfacevertical subsidence in the border of subsidence basin, the stability period of surface movement islong and etc. These phenomena caused lots of attention and concern.This paper taking the Huainan mining area as an example analyzes the mechanism of miningsubsidence in coal mining area with thick alluvium.During the transfer of mining subsidenceconsolidation settlement occurs in alluvial soil layer. Thick alluvium is transferred downwards inthe form of load, which causes the response of rock mass and the synergy subsidence increases thesubsidence of the bedrock surface.Using the measured data proved the probability integral method is defective in Huainan miningareas covered with thick alluvium. The convergence on the fringe of subsidence basin is fast. Inorder to modeling a predict model suitable for alluvium mining area of Huainan,firstly theprobability integration method was modified by adding thick alluvium factor, it has improved overthe original model. Then based on the Boltzmann function, proposed a new predict model.By examples, analysis the prediction and fitting accuracy of the two new models, selectionand the physical meaning of the model parameters. Results show that, in the thick alluvium miningarea of Huainan, compared with original model the mining subsidence prediction model based onBoltzmann function improved by20%-30%in the fitting precision and improved by10%-20%inthe prediction accuracy. Finally taking the the mining subsidence predict model based onBoltzmann function as the predict model and determine the selection principle of model parameters.Furthermore, the global sensitivity of parameters, the robust ability and applicability of newmodel are analysised. Results show that,the sensitivity of parameters consistent with the law ofmining subsidence, the robust ability is better than the original model. Similar material test showsthat, the model is suitable for coal seam dip angle less than50degrees, and the accuracy is slightlybetter than the probability integration model; the model is not suitable for coal seam dip anglegreater than50degrees, and the accuracy is lower than the original model. Testing in several othermines shows that, Boltzmann function model is applicable to general mining geological conditions.Finally, application in Huainan and Fengfeng mining for engineering carried out to goodeffect.