Research On The Upscaling With Random Fractal For Typical Geological Parameters

In the hydrology and oil industry,the scaling effect of typical geological parameters such as porosity and hydraulic conductivity is a common problem to study.It is used to sampling and analyzing on geological parameters,from micro scales to macro scales,to get their distributions.So it is necessary to study the upscaling method.In recent years,there are many upscaling algorithms,mainly classified as simple average method and Laplace equations of two categories.At first,this paper studied the random distributions of typical geological parameters such as porosity and hydraulic conductivity,by truncating the high frequency and low frequency of the data,and formed Truncated Power Law Variogram(TPV)model.Secondly,to get the simulated data and to plot the new data's graph,the idea of sequential Gaussian simulation was adopted for numerical simulation.Thirdly,by using of the Least Square method,fractal dimension of the geological parameters were obtained by estimating Hurst index.Finally,the large scale grid data will be obtained by using of the K-means clustering method.The simulation and upscaling programs were designed in C language.Numerical simulation results show that the method combined truncation model with K-Means clustering could describe the characteristics of the typical geological parameters better.The upscaling results show that upscaling algorithm with random fractal dimension could be widely used in getting the large scale data based on the small scale of data.Therefore,the work of this paper provides a new calculation method in hydrogeology and other industrial operations.