Study On The Response To Mining And Risk Prediction Of Mining-induced Seismicity

As the superficial resource in mine is beening exhausted,it's a inevitable trend to mine deeply for sustainable development of company.During mining deeply,there are some secure problems.Mining-induced seismicity is the most serious secure problem in deeply mining.It's necessary for sustainable development of company and people's life to research the response law and risk for prediction of mining-induced seismicity.Based on national technological project of tenth five-years plan and practice of Dongguashan copper mine, this paper deeply studys response to mining at Dongguashan copper mine.It included space-time law,characteristics of magnitude-frequency and energy-seismic moment, stress-strain law and relations between seismicity and exploitation.At last,simulate and predict the risk of seismicity in artifical neural network.The main results are as follows as:(一)Studys the law of seismicity in time using statistics and nonlinear dynamics methods.The conclusion is as follows as:(1)Mining-induced seismicity in Dongguashan copper mine mostly occur in 01∶00-03∶00 Am,06∶00-08∶00 Am,14∶00-16∶00 Pm and 22∶00-24∶00 Pm.(2)The seismic events magnitudes are more than zero mainly occur in 05∶00-07∶00 Am,13∶00-16∶00 Pm and 22∶00-24∶00 Pm.They occur before the exploiting time.Large seismic events had obvious foreshock phenomenon.(3)Sum of seismic events that magnitudes are between -1 and 1 was increasing.(4)The seismic events which magnitude is between -2 to-1,-1 to 0,and 0 to 1 had perfect fractal characteristics.The fractal characteristics of small seismic events is obvious than those which magnitudes is larger.The former was mainly affected by exploiting condition,and the later by natural seismicity and geological structure.(二)studys the law of seismicity in time using statistics and quantitative seismology methods.The results are as follow as:(1)The seismic events mostly occur in No1-14 stopes and exploratory drifts between the line 52^thand 54^th.The depth is between -670m to 850m.(2)The concentrative areas of seismic events mainly distribute in No 2,6 to 10,14 stopes of line 52^th,No2,6,8,10 stopes and drifts between line 52^thand 54^th.The concentrative areas starte from No2 and 6 stope of line 52th,and then vary as exploiting. All concentrative areas are independed.(三)Statistically studys the magnitude-frequency and energy-seismic moment distribution.The conclusion is as follows as:(1)The both of distribution of magnitude-frequency and energy-seismic moment have two obvious distribution models, and then get their mathatical expressions in linearization.(2)During hardening process of rock,there are little seismic events witch magnitudes is small.During softening process of rock,there are more seismic events witch magnitudes is large. (四)Studys the seismic stress-strain in quantitative seismology methods,the results are as follow as:(1)In time,the seismic stress increases as the exploiting.The seismic strain are stable.(2)In space,the seismic stress mainly concentralize on exploiting areas, and varys as the exploiting areas.The seismic strain is mainly related with the spatial structure.Big stress and strain mostly occur around adjacent rock mass.(五)Statistically and numerically analyzes the relationship between exploiting and seismicity.The results are as follow as:(1)There is a minus logarithmic linear relationship between cumulate seismic energy and cumulate exploited rock volume,and also there is a plus logarithmic linear relationship between cumulate seismic event counts and cumulate exploited rock volume.(2)Large seismic events mostly occur around stope after big blasting events.The seismic events beyond zero magnitude occur before the blasting event in time.Seismic events change from strongness to weakness and strongness again.(3)There is a plus logarithmic linear relationship between cumulate seismic event counts and cumulate blasting capacity.(六)Statistically analyzes the factors of inducing seismicity.And then imulate the risk of seismicity induced by various factors.The results show that there is a strong relationship between seismic risk and blasting capacity,rockburst orientation,the natural seismicity and past energy release,and also a weak relationship between seismic risk and spatial volume, depth of the exploiting and the ratio to intensy of terrane.BP neural network is an exact and short-cut way to forecast seismic risk affected by many complicated factors.