Abstract:

The gas chimney is a special associated structure formed by the hot fluid flow in the sedimentary basin. As a specific channel for oil and gas migration, the gas chimney is a direct evidence for the existence of shallow reservoirs, and it can also influence hydrocarbon accumulation. These days, many scholars have paid attention to gas chimneys. Based on the research results and thoughts about gas chimneys home and abroad, through the artificial neural network (ANN) which is based on multi-layer perceptron and provided by the brand new seismic solution, i.e.OpendTect, this paper predicts the probability of occurrence for gas chimney, improves the algorithm by using dip steering, presents and applies seismic chimney detection technology based on dip steering. The application of pattern recognition technology in seismic exploration field has been well supplemented and developed. Generally, the 3D seismic chimney detection technology can not only interpret the migration pathway and reason of shallow gas accumulation, but also forecast the development of source rocks. In addition, it is also applicable for distinguishing fault sealing. At last, this process was applied to research the migration pathway and accumulation rules of an oil-bearing block in Buir Sag of Hailar Basin, and the fracture characteristics and types of migration pathways were put forward, then the special configuration relations of oil source, pathway, reservoir and cap rock were summarized, and the corresponding accumulation mode was  established.

Key words: gas chimney, dip steering, multi-layer perceptron, artificial neural network, oil-gas migration pathway, fault sealing, reservoir mode