Seismic Multi Attributes Pattern Recognition - Case Study from Malaysian Hydrocarbon Reservoirs

In complex reservoirs, individual seismic attributes are usually insufficient to delineate detailed stratigraphic features of a specific target. More information can be derived from post-stack seismic data by combining different attributes to reveal detailed seismic facies belonging to different stratigraphic features. Seismic Multi Attribute patteRn recogniTion or “SMART” is an advanced technique which delivers relatively high value for exploration and development applications. Research into seismic pattern recognition provides an effective method to map stratigraphic features. Generating seismic facies using a neural network based pattern recognition technology can extract additional features. This paper describes ways to take advantage of various seismic attributes and automatically derive more information from post-stack seismic data. Pattern recognition is based on a sample-to-sample approach. It combines different attributes to reveal details of the underlying geologic features and help in the interpretation of complex stratigraphic settings and rapid facies change in reservoirs.