Research Of Seismic Characterization Method On Fluvial Facies Reservoir Of Laohekou Oilfield

This issue derives from the research project of Shengli Oilfield, with the name Research of reservoir weak seismic reflection effective information extraction technology .Along with the old oilfield in North China entered into later stable production period, the exploration goal has gradually changed from large-scale structural reservoirs which has simple structure and abundant reserves ,to small-scale subtle reservoirs which has complex structural, small target and changes quickly horizontally and vertically such as fluvial sand body , buried hill and shoal-bar sand body. The reservoir of Laohekou oilfield is a kind of fluvial sand body, of which the exploration is a difficult issue in the whole world because the thickness of fluvial sand body is generally smaller than the limit of seismic vertical resolution. So geophysical exploration technology is unable to be used to describe the sand body microcosmic. In this paper, some geophysical exploration technology will be utilized to describe the fluvial sand body, especially its distribution in order to lay the foundation for the following research.Now the technology used for predicting the fluvial sand body reservoir utilizing post -stack seismic data are mainly like these: joint time-frequency analysis (JTFA) and spectral decomposition, frequency band expanding technology, seismic attributes, impedance inversion and so on. In this paper, three methods are carried out to describe the distribution of fluvial sand body and multiple attributes technology with RGB technology is used first. Seismic attributes technology is very effective in reservoir prediction. However, signal attribute is general unavailable, so it is important to consider all the attributes. First, abundant attributes can be extracted and optimized, then these attributes can be joined up and displayed by RGB technology in order to describe the fluvial sand body more clearly, which can be used as a available information for the interpretation following. The most important technology in the paper is spectral decomposition. This technology contents tuning body and equal frequency body, and the equal frequency body technology is utilized in the paper. In this paper, the seismic data should be computed by a joint time-frequency analysis method (JTFA) in order to get a T-F spectrum, then all the time series with the same frequency will be extracted from all the spectrums and then compose a new data(frequency-divided data), So it is very important to choose a kind of JTFA technology to get a high-precision T-F spectrum. In this paper, matching pursuit method, combining with Wigner-Viller Distribution (MP-WVD) technology is used, which is one of the most precise JTFA technology now. Moreover, some improvements are used in the MP-WVD method in order to improve operation efficiency while maintain its excellent time-frequency resolution. Last, it is used for describe the distribution of the fluvial sand body.In the last paragraph, two methods which are all based on the multi-wavelet imagine are carried out. The multi-wavelet imagine is that seismic data is composed of numerous wavelets, of which the amplitude, frequency and phase are all different because the energy of seismic wavelets will become weak, the dominant frequency will be lower and the bandwidth will also become narrow when the wavelets spread in the rock underground. So traditional signal wavelet imagine (for example convolution model) is inadequate to meet the exploration. In this paragraph, the unusual response of fluvial sand body can be highlighted, and the distribution can also be displayed obviously especially some unconspicuous fluvials on the slice by these two technologies.