The advantage of high-resolution sequence stratigraphy, which takes base-levels as reference, is that it can be applied to continental depositional basins controlled by multiple factors and can effectively improve the...The advantage of high-resolution sequence stratigraphy, which takes base-levels as reference, is that it can be applied to continental depositional basins controlled by multiple factors and can effectively improve the accuracy and resolution of sequential stratigraphic analysis. Moreover, the principles of base-level cycles are also suitable for analyzing sequential stratigraphy in continental coal-bearing basins because of their accuracy in forecasting distribution of coal measures. By taking the Dongsheng coalfield in the Ordos basin as an example, the extensive application of base-level cycles in exploration and exploitation of coal is analyzed. The result shows that the Yan’an formation in the Dongsheng area is a long-term base-level cycle which is bordered by nonconformities and made up of five mid-term cycles and 13 short-term cycles. The long-term cycle and the mid-term cycles are obvious in comparison with a transverse profile. The episodic coal accumulation in the Mesozoic Ordos basin means that the deposition of primary matter (peat bogs) of coalification is discontinuous, periodical and cyclical in the evolution of the basin. The episodic accumulation of coal measures in the Yan’an stage is controlled by ascending-descending changes of a long-term cycle and middle-term cycles. Coal measures formed during the early and late periods of the long-term cycle are characterized by multiple layers, big cumulative thickness and poor continuity. Coal measures formed in the mid-term of the long cycle are dominated by good continuity, fewer layers and a small additive thickness, which is favorable for the accumulation of thick and continuous coal measures in the transition stage of mid term base-level cycles.展开更多
The authors discussed the method of wavelet neural network （WNN） for correlation of base-level cycle. A new vectored method of well log data was proposed. Through the training with the known data set, the WNN can re...The authors discussed the method of wavelet neural network （WNN） for correlation of base-level cycle. A new vectored method of well log data was proposed. Through the training with the known data set, the WNN can remenber the cycle pattern characteristic of the well log curves. By the trained WNN to identify the cycle pattern in the vectored log data, the ocrrdation process among the well cycles was completed. The application indicates that it is highly efficient and reliable in base-level cycle correlation.展开更多
基金Project2003CB214603 supported by Development Plan of the State Key Fundamental Research, China
文摘The advantage of high-resolution sequence stratigraphy, which takes base-levels as reference, is that it can be applied to continental depositional basins controlled by multiple factors and can effectively improve the accuracy and resolution of sequential stratigraphic analysis. Moreover, the principles of base-level cycles are also suitable for analyzing sequential stratigraphy in continental coal-bearing basins because of their accuracy in forecasting distribution of coal measures. By taking the Dongsheng coalfield in the Ordos basin as an example, the extensive application of base-level cycles in exploration and exploitation of coal is analyzed. The result shows that the Yan’an formation in the Dongsheng area is a long-term base-level cycle which is bordered by nonconformities and made up of five mid-term cycles and 13 short-term cycles. The long-term cycle and the mid-term cycles are obvious in comparison with a transverse profile. The episodic coal accumulation in the Mesozoic Ordos basin means that the deposition of primary matter (peat bogs) of coalification is discontinuous, periodical and cyclical in the evolution of the basin. The episodic accumulation of coal measures in the Yan’an stage is controlled by ascending-descending changes of a long-term cycle and middle-term cycles. Coal measures formed during the early and late periods of the long-term cycle are characterized by multiple layers, big cumulative thickness and poor continuity. Coal measures formed in the mid-term of the long cycle are dominated by good continuity, fewer layers and a small additive thickness, which is favorable for the accumulation of thick and continuous coal measures in the transition stage of mid term base-level cycles.
基金Supported by Project of Dagang Branch of Petroleum Group Company Ltd,CNPC No TJDG-JZHT-2005-JSDW-0000-00339
文摘The authors discussed the method of wavelet neural network （WNN） for correlation of base-level cycle. A new vectored method of well log data was proposed. Through the training with the known data set, the WNN can remenber the cycle pattern characteristic of the well log curves. By the trained WNN to identify the cycle pattern in the vectored log data, the ocrrdation process among the well cycles was completed. The application indicates that it is highly efficient and reliable in base-level cycle correlation.
