Study On Full-3D Surface Related Multiple Prediction For Marine Seismic Data

Oil and gas is an important resources, and play a decisive role for the country’s economic security。In recent years, the demand for energy has increased year by year with the rapid growth of China’s economy. China’s oil resources are not abundant, onshore oil/gas fields have been more in-depth exploration, but the exploration of offshore oil resources are relatively low. With the depletion of oil resources of onshore, the oil resources will mainly depended on the sea. In recent years, China has established the strategy of oil and gas exploration for deepwater, and the research on marine seismic data processing technology is the prerequisite for promoting marine seismic exploration.Marine seismic data generally have a higher resolution, but the multiple is a major problem in marine seismic data processing. Surface related multiples are the main noise of marine seismic data, and now the 2D wave-theory-based surface related multiple elimination (SRME) is the main multiple eliminate technique. But there are some difficulties that process 3D field data by 2D SRME. First, there are great errors when process 3D field data by 2D SRME because the 2D SRME does not consider the multiple contribution of crossline direction. Secondly, the prerequisites of 3D SRME can’t be meeting because of the streamer feathering and the large cable interval for current 3D marine acquisition. At present various for 3D SRME algorithm both in the theoretical study stage, and the effectiveness and computational efficiency for the actual seismic data still can’t achieve the required actual production.For this contradiction between the 3D field data and the SRME algorithms, the thesis to research surface related multiple prediction technology that suitable for 3D marine seismic data relies on the " National Science and Technology major projects for large oil/gas fields and coal seam gas development (2011ZX05019-003)" and the "CNPC major projects for core geophysical software integration and upgrade (2012E-31-03)".The thesis systematically studies the current situation of multiple suppression techniques of domestic and international, and summarizes the criteria of multiple classifications, the classes of multiples, the characteristics of multiple in seismic data, the methods to identify multiple. The thesis studies the typical multiple suppression methods for marine seismic data process and applied those methods to field seismic data. The thesis analyzes and summarizes the advantages, disadvantages and the respective scope of current multiple suppression methods for marine seismic data.Based on the above analysis, we choose to study data-driven surface multiple elimination techniques to solve the multiples of 3D marine seismic data.The thesis has made the following four innovative achievements:1. The 3D surface related multiple has been suppressed by data regularization joint with sparse inversion for the first time.For the contradiction between the 3D field marine seismic data and the SRME algorithms, the thesis using data regularization and de-regularization to makes the data distribution satisfies the prerequisites of 3D SRME, and using sparse inversion to eliminate the adverse effects of sparse sampling at crossline for multiple prediction.The thesis has established the surface related multiple suppression procedure that based on data regularization, sparse inversion and data de-regularization, and the multiples of 3D field marine data have been suppressed by using this procedure.2. The new fast 3D surface related multiple prediction method based on aperture optimization has been presented for the first time.For the problem that low efficiency of the sparse inversion-based 3D SRME, the thesis researched the aperture optimization method, proposed and implemented the fast 3D surface related multiple prediction method. This method can greatly reduce the computer storage and computer cost but have the similar multiple prediction result of 3D SRMP with sparse inversion.3. Form two different sets of 3D surface related multiple prediction technology solutions.Through comprehensive study, we formed an accurate 3D surface related multiple prediction technology solution and a fast 3D surface related multiple prediction technology solution, so users can select either the accurate solution or the fast solution according to the processing requirements of field data process.4. Form a full 3D surface related multiple suppression procedure for deepwater cable seismic data process.By applied those techniques to various 3D field marine seismic data process, we formed a full 3D surface related multiple suppression procedure for deepwater cable seismic data process:first use the surface related multiple elimination techniques to attenuate the surface related multiples of near and middle offset, and then use beam-forming technique to attenuate multiples of middle and far offset, final use the diffracted multiple attenuation technique to attenuate the residual multiples. The result of 3D field marine seismic data process proved that those techniques are effective and practical.