Methodology Of Fracture Identification And Software Development Based On Log Data

For the geophysical exploration of fractured reservoir,the image logging technology is the most advanced and the most direct method.However,image logging requires great expenses,which limits the application of the image logging technology.The conventional logging data are dominantly used in most oilfield companies in China,thus the effective fracture identification method with conventional log data in the absence of image log data is very important.In this paper,the fracture identification method based on conventional log data is systematically studied and innovated.In the aspect of numerical analyses,14 characteristic parameters are defined through hybrid operation of conventional log data.After calculating the weight value of each parameter,we can get the Comprehensive Index.In the aspect of curve shape,Fractal Dimension of each curve is defined,and Comprehensive Dimension is calculated based on the weight value and fractal dimension.With the combination of numerical and curve shape analyses,the index of fracture can be determined.The relationship between fracture probability and logging curves is defined using the Logistic Regression Model.The result of this analysis can be used on other wells in the work zone.Finally,developing a fracture identification software with the team.The processing results of actual data are compared with measurement results of core samples,thus the validity and the feasibility of the method is confirmed.