Research Of Fault Detection Based On Coherency Algorithm

With the development and progress of our society, human demand for energy isincreasing continuously. Oil is major power energy and chemical materials, which plays anextremely important role in today’s economic society. As a strategic material, it is widely usedin industry, agriculture, military and all fields of our daily life. Therefore, the oil and gasexploration is critical to energy development. Fault detection is an important progress in theexploration of oil and gas, which is significant to the migration and accumulation of oil andgas.Traditionally, faults are picked in3D seismic volumes as discontinuities in seismicamplitude on vertical sections and time slices, this approach detects fault by visuallyrecognizing reflector’s discontinuity. However, the results of conventional fault interpretationgreatly depend on the geological prior knowledge and experience of the interpreter. Besides, ifthe fault is complex, the progress of fault detection becomes complex and the efficience is low.The strong subjectivity of the interpreter and the complexity of geologic targets may affect thefault interpretation’s confidence and accuracy. Therefore, it is imperative to improve theefficiency and confidence of fault interpretation.Considering the drawbacks of the traditional fault interpretation, the paper proposes afault detection method based on the coherency algorithm. The main works of the thesis arefocus on:1.Study the seismic data format SEG-Y. SEG-Y format seismic data is one of the standardtape data formats made by SEG (Society of Exploration Geophysicists), which generallyorganizes seismic data in traces. The SEG-Y format is extremely slow when the interpreterdisplays time/depth slice form large seismic data. Therefore, the paper translates the SEG-Ydata into3D seismic volume, which improves the efficience of achieving time/depth slicelargely.2.In order to enhance the discontinuity of the seismic data, the paper uses the coherencyalgorithm to generate the coherence cube after studying the seismic attributes. The coherencecube is used as the fault attribute cube. The thesis mainly analyzes the basic principle of thefirst generation algorithm(C1),the second generation algorithm(C2) and the third generationalgorithm(C3). The improved C3algorithm based on Rayleigh acceleration is applied to thepaper, which improves the efficience largely. Besides, we can achieve good experimentalresults by comparing time/depth slice of the C1algorithm with the improved C3algorithm.3.Introduce the basic principle and working mechanism of ant colony algorithm by solving TSP. Considering the successful application of ant colony algorithm in multiple fieldsand its characteristics, the paper combines ant colony algorithm with fault identification anduses ant colony algorithm to extract the fault lines from a series of coherence slices, whichcan be used to reconstruct the fault surfaces later.