A regression approach for seismic first-break picking

The picking efficiency of seismic first breaks(FBs)has been greatly accelerated by deep learning(DL)technology.However,the picking accuracy and efficiency of DL methods still face huge challenges in low signal-to-noise ratio(SNR)situations.To address this issue,we propose a regression approach to pick FBs based on bidirectional long short-term memory(Bi LSTM)neural network by learning the implicit Eikonal equation of 3D inhomogeneous media with rugged topography in the target region.We employ a regressive model that represents the relationships among the elevation of shots,offset and the elevation of receivers with their seismic traveltime to predict the unknown FBs,from common-shot gathers with sparsely distributed traces.Different from image segmentation methods which automatically extract image features and classify FBs from seismic data,the proposed method can learn the inner relationship between field geometry and FBs.In addition,the predicted results by the regressive model are continuous values of FBs rather than the discrete ones of the binary distribution.The picking results of synthetic data shows that the proposed method has low dependence on label data,and can obtain reliable and similar predicted results using two types of label data with large differences.The picking results of9380 shots for 3D seismic data generated by vibroseis indicate that the proposed method can still accurately predict FBs in low SNR data.The subsequent stacked profiles further illustrate the reliability and effectiveness of the proposed method.The results of model data and field seismic data demonstrate that the proposed regression method is a robust first-break picker with high potential for field application.

The origin and accumulation model of crude oils from oil reservoirs Chang 9 and Chang 10 in the Yanchang Formation of the Ordos Basin

In the lower parts of oil reservoirs Chang 9 and Chang 10 of the Yanchang Formation are oil-bearing layers newly found in oil exploration in the Ordos Basin.Based on GC,GC-MS analyses of saturated hydrocarbons from crude oils and source rocks,reservoir fluid inclusions and BasinMod,the origin of crude oils,accumulation period and accumulation models are discussed in combination with other petroleum geology data in this paper.The result shows that(1) there are two different types of crude oils in oil reservoir Chang 9 in the Longdong and Jiyuan regions:crude oils of typeⅠ(Well D86,Well A44,Well A75,Well B227,Well X62 and Well Z150) are mainly de-rived from the Chang 7 source rocks(including mudstones and shales) and distributed in the Jiyuan and Longdong regions;those of typeⅡ(Well Z14 and Well Y427),are distributed in the Longdong region,which are derived from the Chang 9 source rocks.Crude oils from oil reservoir Chang 10 in the Shanbei region are mainly derived from the Chang-9 source rocks;(2) there are two phases of hydrocarbon filling in oil reservoir Chang 9 in the Jiyuan and Longdong regions and oil reservoir Chang 10 in the Shanbei region:The first phase started at the early stage of J2z.The process of hydrocarbon filling was discontinuous in the Late Jurassic,because of the tectonic-thermal event in the Ordos Basin.The second phase was the main accumulation period,and hydrocarbons began to accumulate from the late stage of J2a to the middle-late of K1,mainly at the middle-late stage of K1;(3) there exist two types of accu-mulation models in oil reservoirs Chang 9 and Chang 10 of the Yanchang Formation:source rocks of the reservoirs in oil reservoir Chang 9 in the Jiyuan region and oil reservoir Chang 10 in the Shanbei region,the mixed type of reservoirs on the lateral side of source rocks and source rocks of the reservoirs in oil reservoir Chang 9 in the Long-dong region.