Apache is exploring a large block offshore Australia. Part of the area is covered with good quality 3D seismic and a number of prospects have been identified from this data. Additional potential prospects were inferred from 2D seismic in the relatively under-explored deep water area of the block. With the current high cost of 3D seismic and drilling programs, it was considered very important to high-grade the prospects and to correctly assess the prospectivity of the deep water area.Initially it was thought that traditional target-oriented CSEM surveys over each prospect would provide the required information. However, after further study, it was decided to cover the entire area with electromagnetic scanning. This approach provides a coarse 3D view of the entire area providing information not just about prospects identified from seismic, but potentially also revealing new hydrocarbon leads.A highly rugose seafloor in combination with a high resistive overburden of varying thickness appeared as a major concern for the scanning survey. A wide, deep submarine canyon provided both operational challenges and data processing issues. The rugosity of the seafloor and varying overburden thickness constitute significant local as well as regional variations to the background resistivity distribution, making it difficult to extract potentially hydrocarbon related anomalies from the scanning data.A novel approach to dealing with this problem was adopted, which takes advantage of complementary information from existing well logs and the available seismic data. A number of 1D inversions constrained by resistivity logs were performed at various locations across the survey. The results of the 1D inversions were then used to build a reference resistivity model that conforms to the bathymetry and the seismically derived overburden thickness. Detailed 3D simulation of the scanning survey for this resistivity model generated synthetic reference responses, which adequately account for most of the bathymetry and overburden related variations in the scanning data. Using these synthetic reference responses to normalize the scanning data, a number of interesting anomalies became apparent, one of which coincides with a known oil reservoir. The same anomalies had been masked by regional trends in previous results obtained by conventional single-receiver referencing.The results obtained significantly increased our confidence in the interpretation of the scanning data and highlight the increased value obtained from an integrated analysis with complementary geophysical data.

No abstract