Metamorphic Buried Hill Fractural Reservoir Prediction Study In Damintun Depression

The fracture reservoirs in oil and gas production accounted for more than half of the whole production. Domestic discovery of reservoir is also increasing in recent years. With the increasing degree of exploration, fracture-type and other non-structural reservoir prospecting has become the main direction of major oil fields. Damintun sag area, located to the North of Liaohe basin, has the whole 3D seismic data. Judging from drilling in the study area, it is of high research and exploration value.Based on the theory of seismic attributes, this paper proposed an application of the fracture prediction in Damintun buried hill block. With the analysis of the basic data of logging and geology, the petrophysic property gives an understanding of the relationship between the reservoir properties. The response characteristics based on the seismic forward modeling technology provide the identification of the fractural zones in the view of seismic method. The analysis of the seismic attributes and the selection of the sensitive attributes leads to the comparison of the geology data and drilling data. At last, the preferred fractural development zones can be predicted based on the application of pre-stack coherent detection of fracture, ant-tracking technology, multi-scale edge detection and multiple attribute integrated analysis techniques, such as neural network technology.The logging characteristic analysis of the fractural imaging should be conducted before the application of the seismic attribution. The accuracy of the fractural identification can be promoted using the logging imaging technology. At the same time, through the analysis of the character curves, different logging response (such as velocity, density, resistivity, gamma, neutron) can be summarized and help to categorize the sensitive logging data, based on which the response relationship can be established. Meanwhile, for different density and different development degree, the forward modeling can be set up. The result of the forward modeling provides the post-stack seismic response characters of different lithology, velocity and fracturing degree.The paper gives an analysis of the seismic attribution application, especially for some commen attributions. Further study on the post-stack coherent technology and ant-tracking method is conducted. (1) The coherent technology can provide a macro information of the fracture development based on the analysis of the longitudinal and horizontal waveform similarity. (2) The ant-tracking can give a clear distribution of the fractures. (3) The wavelet multi-scale edge detection technique with the threshold Fl=0.012376 and F2=0.137816 shows the fracturing zone whose value is higher than F2 and the sub-development zone. (4) The comprehensive analysis method based on the correlation and attributions characters gives 7 sensitive attributions. The optimization of the attributions helps the prediction of the fractural zone based on neural network technology.At last, based on a series study of the prediction method of the metamorphic facture, some conclusions are presented:(1) In fractural development zones, the double lateral resistivity curves have large positive differences, especially in acoustic curves. (2) The forward modeling for the fractures gives 23 geology models with different fractural parameters in 5 categories. The forward modeling results are also analyzed in this paper. It shows that the fractural density and fractural equivalent velocity are the main influence factors. (3) In the analysis of the post-stack seismic attributions, the abstracted attributions including amplitude, complex seismic trace statistics, sequence statistics and spectral. The optimized 7 sensitive attributions are set as the basic data for the neural network prediction method. The result shows the main fractural development zones. Based on the fractural density map, the interested area can be divided into three blocks. The validation well H20, H30-28-28, S3O3, S269 and H31-16-18 present a good error evaluation with mean error lower than 8%. (4) The conclusion of the series of studies shows that the results of different conventional methods are basically similar. The multi-scale seismic attributions methods performance better with lower error and high accuracy.