Geo-Hazard Detection with Chimney Cubes

Abstract We describe geo-hazard applications of "chimney cube" processing, an interpretation methodology that highlights vertical anomalies on the seismic data, associated with gas clouds and gas chimneys. Among such applications are identifying drilling hazards caused by shallow gas pockets. Ensuring platform stability problems through avoiding sub-sea mud volcanoes is another such application. The foundation of chimney processing is based on the meta-attribute concept that will be briefly highlighted. Chimneys are also used for exploration of hydrocarbon targets both in high grading prospects and better understanding the petroleum system. Except for a brief overview of exploration applications of chimney cubes, only geo-hazard applications will be emphasized. Meta Attribute Concept As it was described in Aminzadedh and Connolly (2002), meta attributes are aggregation of a number of seismic attributes where the interpreter's insight and power of a trained neural network are combined to detect a particular seismic anomaly. Figure 1, shows a number of seismic attributes as input to a neural network from known or suspected chimneys (or faults). The training of neural network using interpreter's insight renders the "meta attribute" suitable for detection of a given seismic body, in this case, chimneys. Figure 2 shows a slice of input seismic data and the corresponding chimney output. Other seismic objects we can detect using meta attribute concept are fault planes, fractures, reefs, sand channels and salt bodies. In what follows we will adopt from an earlier Leading Edge of Geophysis publication by Aminzadeh et al (2002) highlighting several geo-hazard applications of gas chimneys. Also see Hegland et al (1998) OTC paper for related topics. Geo-hazard applications discussed here are: detecting shallow gas pockets, distinguishing shallow fault geo-hazard, and predicting shallow geo-pressure. For various exploration applications of gas chimney cubes see Hegland et al (2000), Meldahl et al (2001) and Aminzadeh et al (2001). Some of these applications are: understanding of deep petroleum migration processes, distinguishing between charged and non-charged fault segments, detecting sealing versus leaking faults, and predicting oil versus gas-prone prospects and distinguishing oil-prone versus gas-prone prospects. Detecting Shallow Gas Pockets One of the original impetus' for developing the chimney cube technology was related avoiding shallow gas pockets. This originated from different sources of knowledge and experience, for example the blow out of a well and destruction of a platform in the North Sea due to a shallow gas pockets, and resulting research into the detection of such hazards. Both the presence of new seismic characterization technology using neural networks, and the unused potential in seismic acquisition and processing techniques also motivated the tool building. Gas hazards are often manifested by pockmarks or sea floor mud volcanoes. Figure 3 shows an example of the output of chimney processing that highlights near surface gas pockets in the deepwater, Gulf of Mexico. Obvious chimneys, which extend to the sea floor (Figure 3, events marked A) can be observed from conventional seismic. The chimney cube data however can highlight more subtle chimneys and those which do not extend to the sea-floor (Events marked B).