Stochastic Simulation Of Karst Groundwater Dynamics In Coal Mine And Its Application

Karst groundwater dynamics in coal mines is an integrative reflection of geologic settings and water resource forming condition. In northern China, a lot of coal mines are facing the threat of Ordovician confined water. So it is of great significance to grasp and understand the Karst groundwater dynamics.Based on an analysis of the stochastic nature of geometric parameters in karst medium field, the dissertation focuses on the simulation of mining-induced impact on the natural Karst groundwater dynamics in coal mines. The mining activity further enhances the complexity of Karst groundwater dynamics and makes the prediction of groundwater dynamics more difficulty.The dissertation deals at first with the main characteristics of karst groundwater dynamics under natural and nature-mining coupling conditions, respectively. It can be seen that water yield in coal mines has a close relation with precipitation, hydro- geology condition and mining activity etc.. Other factors involved include water level, salinity,thermal dynamics, and openness of geology structures.Lyapunov exponent calculated by the author and from references have proved the chaos character of mine groundwater dynamics. And the chaos mechanism has been explained in different ways, like in system theory, hydrogeology, engineering geology and man-made mining, and"three zone"theory, etc., Because of mine activity, mine groundwater system structure was changed, and water from multi-sources could flow into mine. So the system was turned to a complex nonlinear system, which makes mine karst groundwater system not only determinative,stochastic, but also chaotic.To predict the water yield of coal mine from January to June in 2008, the author uses a series of stochastic simulation methods, including ARIMA time series(by EViews software), chaotic time series, and RBF nerve network chaotic time series(Matlab program) etc., Among these methods, chaotic time series method is preferable,the least relative error by chaotic time series is 0.9%, and others is less than 12%. A certain-random model was built to describe Ordovician groundwater dynamics in ZCD coal mine, under consideration of the stochastic border and stochastic input-output. Simulation results also show the complexity of mine groundwater system.According to the chaos nature of groundwater system, a chaos degree model was built in the dissertation to forecast water hazard on the basis of groundwater dynamics. In the model, Lyapunov exponent represents the predictability,and the fractal dimension D, the complexity degree, and R, the differentiation degree of groundwater dynamics. When lyapunov exponent is greater than zero, then the model could be used to predict the groundwater dynamics by chaotic time series method, and according to the different R values, four relevant hydrogeological conditions could be recognized.Finally, the models was applied in Gujiao coal mine and some conclusions were brought out, for example, average 11 months are needed groundwater recharge in the coal mine. And groundwater dynamics change was closely related to the structure. The R value of 0.977 in ZCD mine show that the water yield is normal and there is no water hazard threat for the coal production.And 34 figures, 20 tables and 249 references are included in this dissertation.