网络裂隙介质的刻画是水文地质研究领域最具挑战性的难题之一。介绍布尔模拟和多点地质统计两种生成裂隙网络的随机模拟方法基础上,分别利用布尔模拟和多点地质统计方法对某一典型裂隙介质网络进行随机模拟,并通过与真实裂隙场的迹长、走向等几何参数的对比评价其模拟精度。结果表明,两种随机方法都能实现对裂隙结构特征的近似刻画,其模拟误差均在10%以内。相对来说,布尔模拟对裂隙迹长的模拟精度更高,多点地质统计模拟对裂隙走向的模拟效果更好。同时,布尔模拟较多点地质统计模拟对裂隙的已知数据要求更高。裂隙网络的随机模拟为准确刻画裂隙介质中的地下水流运动及污染物运移提供了现实可能性。
Characterization of fractured rocks is one of the most challenging problems in hydrogeology. In this paper,two stochastic methods,Boolean method and multiple-point statistical simulation,are introduced to establish fracture network models of a specific field. We compare the geometrical parameters with real data,such as length and orientation,to assess its characterizing accuracy. The results turn out satisfactorily with both of the methods,and the simulation errors are both less than 10%. Relatively,the Boolean method is more accurate in characterizing fracture length,and multiple point statistical method is more accurate in characterizing orientation. The Boolean method has higher requirements on conditioning data. Stochastic simulation provides a possibility for accurately simulating flow and mass transport in fracture networks.