The Study Of Strata Sequence And Sedimentary Facies Based On Well Logging Data In Su-Bei Basin

The Su-bei basin is a continental rift-lake oil and gas-bearing basin whose geological structure is characteristic of its various faults. The exploration in the past 30 years gained superior achievements, it got massive geological data, especially the well-logging data. This paper base on the standpoint that the massive routine well-logging data include sufficient and overall geological information, studying the well-logging information which indicates the strata sequence, lithology and then sedimentary facies, then combining with the analysis to the actual target and test of programming and processing, find out the method which divides the sequence and recognizes the facies reasonably based on the well-logging data. Finally realize the method by programming and deal with the data in local region of interest.The paper surveyed and studied the manifestation of the strata sequence and the sedimentary facies on the well-logging data firstly: gaining some samples from the multi-material study to the features of sequence and facies of the chief hydrocarbon-bearing strata in the typical interest region, so accomplished the quest to the relationship between the well-logging data and the geology. What's immediately following was the surveying and comparing to the expression and acquirement of the information of the sequence and the facies based on the logging data.The next step is the selection of the suitable means by which study the sequence and facies on logging data: on the respect of dividing sequence on logging data, the general measures were compared for a best one, for examples, the qualificative measures such as dividing sequence by step suddenly change on SGR(natural gamma ray spectrometry log) and by the change of stratum attitude showing on the dip vector figure etc., and the quantitative measures such as dividing sequence by activity method, the method which analyze the ratio of sand and mud, the method of depth-frequency analysis and wavelet analysis etc., ultimately the wavelet analysis method was chosen, with which the dividing of the sequence and the strata sedimentary cycle association is obtained by study the vertical change of the lithology and physical property. on the respect of recognizing facies on logging data, the general measures were compared for a best one also, the modeling methods such as cluster analysis, the grey-correlation method, artificial neural network technique, fuzzy logic on the logging data etc., finally the artificial neural network method based on the characteristic quantity which are formatted and ranked was chosen, with which realized the auto-recognizing of the micro-facies in local area.Based on the above, with programming, actual data treating, and the uninterrupted adjustment, the program holds sequence's auto-dividing and facies'auto-recognizing functions was completed. The result of the running of the program show, the sequence's auto-dividing method is relatively reasonable on the cycle of sedimentation in mid-high frequency, and it's helpful to the correlation of the strata. The facies'auto-recognizing method has a remarkably effect on the micro-facies'auto-recognition in local area and specially strata.