Analysis Of Well Logging Fractures In Carbonate Rocks

In carbonate reservoirs,the existence of fracture cavities will change the response characteristics of conventional logging and electrical imaging logging.Because of the complex lithology of carbonate rocks,often the conventional logging is difficult to identify accurately,so the fracture can be identified by using conventional logging and imaging logging data in accordance with the characteristics of the carbonate rocks in the A area.The main work of this paper is to deal with the data of A area.The main work is to preprocess the conventional log data,including the depth correction of the log curve,the smooth filtering of the log curve,the correction of the environment impact of the log curve and so on.This paper introduces the calculation of facture parameters of conventional measurement,analyzes the characteristics of conventional logging curves,and identifies conventional logging fractures with coring data.The conventional logging feature parameters and extracted and multiple logging information is analyzed,and the qualitative identification of cracks is made through the method of difference ratio and curve reconstruction.The imaging log is generated,and the imaging log is processed preliminarily.Compared with the coring data,the location of the crack is marked.Imaging logging image processing,the pre processing of imaging logging images includes EMAX voltage correction,failure electrode detection and compensation,acceleration correction and depth correction,and image balance processing.Based on the conventional logging data and imaging logging data of A area,the digital image processing is realized in the java platform.It includes image enhancement,image grayscale,imaging two value and so on.When using human-computer interaction to identify cracks,it is necessary to manually select points on the cracks,use the three B spline function to interpolate,and use the least square fitting method of sinusoidal curve to obtain the parameters of the sinusoidal curve,so as to complete the human computer interaction identification crack and obtain the crack parameters according to the parameters of the sinusoidal curve.It is very important to identify the artificial selection points of cracks by man-machine interaction,because each person selects different points,and there are differences in the identified cracks.We can use Hough transform to automatically pick up cracks to reduce this artificial difference,but using Hough transform to automatically identify cracks has a large amount of data.The imaging logging combined with the actual data,using Hough to identify the results of fracture,using the voting mechanism to obtain the position of the sinusoidal baseline,and then obtain the Hough transformation of two parameters,which is good for the identification of single crack.As Hough changes identify multiple cracks in the search for baseline,there will be a false peak.At this time,multiple peaks can be selected and multiple cracks can be identified under the condition of error allowable.Finally,we combine conventional logging and imaging logging to identify fractures.The conventional logging data is used to analyze the crack in the large scale section,and the imaging logging data are used to detect the small scale fracture section,and the two phases are combined to improve the accuracy of the fracture recognition.