ABSTRACT

Today's exploration and, particularly, exploitation methods, with a major reliance placed on mapping with 3-D seismic data, generate a great deal of potential information about prospective reservoirs. Effective prospect evaluation requires consideration of the sealing characteristics of faults, and techniques have been developed to improve fault surface analysis. "Allan" fault surface profiles permit assessment of sand juxtaposition across the fault, and can be prepared by manual mapping methods if adequate structural maps are available from 3-D seismic interpretation. Commercial software is available to perform similar analysis directly from the 3-D data.

Vermilion Block 331 field, operated by Marathon, was selected for a pilot study. The field consists of a low relief anticline, downthrown to a regional growth fault. Numerous small faults, with limited vertical separation, cross the crest of the anticline and compartmentalize reservoir sands of Trimosina A (Pleistocene), Angulogerina B (Pleistocene), and Lenticulina (Miocene) age. Faulted reservoirs with multiple, stacked sands are particularly prone to loss of hydrocarbons by leakage across fault surfaces, so that this field was considered ideal for testing the effectiveness of fault surface analysis. Both lateral and top seal risk were evaluated by means of fault surface profiles along five of the crestal faults, to determine the limits of trapping potential and paths for vertical migration. A detailed review of actual hydrocarbon distribution was then compared with the predictions made from fault surface analysis. Seventy percent of a total of 83 predicted hydrocarbon/water contacts were found to be correct within ten meters (30 feet).

The role of faults in permitting up-fault migration along the fracture surface or in providing shale smear barriers to cross-fault migration from sand to sand may confound interpretations based only on fault surface profile geometries. For this reason, shale smear factors were also determined and used in assessing trapping potentials. A critical value of shale smear factor appropriate for this field was found empirically to be between 1.85 and 2.0. Capillary limited cross-fault migration was blocked in all cases where the value was lower than critical, while spill point limited traps occur where values are above critical. This analysis explained all the remaining discrepancies between predicted and actual hydrocarbon/water contacts mentioned in the preceding paragraph.