A method is demonstrated to highlight hydrocarbon migration related to faulting in seismic data. The method uses multiple seismic attributes and neural networks to highlight the vertically aligned low-energy chaotic seismic data described as gas chimneys, gas clouds, or seepage pipes. The result is a gas chimney probability volume. Methods are also described for highlighting faults in seismic data using directional attributes. When the gas chimney probability data are overlain on the fault data, obvious vertical gas chimneys can be distinguished. However subtler fault-related hydrocarbon migration can also be seen. This hydrocarbon migration is often associated with fault intersections or splinter faults related to shear along the fault. Overlaying the chimney information on fault planes can often indicate which parts of the fault have been migration pathways and which parts of the fault have not. More than 125 chimney analyses have been performed on drilled structural closures with effective reservoir. Approximately one quarter of the wells are dry holes. Thus, the dry holes failed because of either ineffective charge or seal. These structures were classified, based on the character of the chimneys below and above the reservoir, into eight classes. We will demonstrate how this information can be used to risk vertical seal and charge prior to drilling.

Gas chimneys are a means by which deep pressures can be transmitted into the shallow subsurface. This partly explains why producing reservoirs are often near the top of abnormal pressure. We will demonstrate the correlation of rock property data (fracture pressure and pore pressure), gas chimneys, and hydrocarbon column height. Case studies are shown from the Gulf of Mexico onshore, shelf, and slope.