基于渗透系数序贯高斯模拟的水库渗漏量不确定性分析

Uncertainty analysis of reservoir leakage based on sequential Gaussian simulation of hydraulic conductivity

水库库底地层渗透系数是影响水库渗漏量的主要因素之一,研究渗透系数的空间变异特征可为水库渗漏量计算结果的可靠性分析提供科学依据。以天津市北塘水库库底地层渗透系数为研究对象,先利用克里金插值对该参数系列进行了插补,再对插补后的渗透系数系列进行了500次序贯高斯模拟,最后分析了其空间变异性对库区渗漏量计算值不确定性的影响。结果表明：库底地层垂向渗透系数空间结构可用球状模型描述,在变程1210 m范围内具有高度空间相关性。以2005年7月16日渗漏量计算值总体为例,最小值与最大值分别为1 389.49 m3/d、1 897.30 m3/d,且落在区间1 600~1 850 m3/d的频率高达77.6%。

Hydraulic conductivity of the reservoir bottom stratum is one of the main factors affecting reservoir leakage. Studying the spatial variability of hydraulic conductivity can provide a scientific basis for the reliability analysis of the calculation results of reservoir leakage. Using the bottom stratum of the Beitang Reservoir as a case study,the parameter series were extended through Kriging interpolation. Then, the interpolated hydraulic conductivity was estimated by sequential Gaussian simulation which was carried out 500 times. Finally,the effect of the spatial variability of hydraulic conductivity on the uncertainty of calculated leakage was analyzed. The results show that the spatial structure of vertical hydraulic conductivity can be described by a spherical model,and it has a high spatial correlation within 1 210 m. As for the leakage calculated on July 16,2005,the smallest value was1 389. 49 m3/ d,the largest value was 1 897. 30 m3/ d,and the frequency of the calculated value falling into the range of 1 600 to 1 850 m3/ d reached 77. 6%.