Chemical Characteristics Of Groundwater And Discriminate Models Of Water Brusting For Xiandewang Coalmine

The karst mine in North China is the typical areas where the water disaster is one of the most serious problems.The geological and hydrogeological conditions of the mine is complex,and the type of water disaster stands completely.Especially,with the increasing depth of mine,Ordovician limestone water pressure to coal floor increases gradually,which has become a major threat to the coal mine safety production.In fact,beside some visible sign of water inrush,it also contains no visible sign of all type of water inrush such as water hydrogeochemical implicit information.If this information can be made full use of timely,geochemical characteristics of hydrological can be analysis and study,which has important practical significance to ascertain the mine water bursting source accurately as well as the possibility of water disaster.This article discusses the geochemical analysis of hydrological characteristics of four major aquifers as Yeqing,Daqing,Benxi and Ordovician limestone of Xiandewang mine in Hanxing areas.The results showed that runoff conditions of each aquifer is good and they have the similar genetic type so that different hydrogeochemical characteristics of the aquifer has not only a certain independence but also has a certain similarity.Water-rock interaction is the main cause of X mineral water hydrological characteristics.In addition,the geological environment also plays a role in controlling to its hydrological composition.Using Piper,Cluster analysis and Boxplot,the standard water sets of Daqing,Benxi and Ordovician limestone aquifer are built.Determined by analyzing standard water sets,three models of water bursting source discrimination is established such as Characteristic ions based on comparative analysis of water quality characteristics,Nonlinear gray correlation method discrimination and Fisher based on statistical principles,and the three models have been combined to improve the accuracy rate.Finally,comprehensive discriminant model has been examined.The results show that the use of comprehensive discriminant model increased discrimination accuracy rate from 80%in single discriminant model to 93.33%,that greatly improved the reliability of discrimination.