The Impact of Broadband Seismic Acquisition and Implications for Interpretation and Reservoir Model Building

Abstract

Abstract This paper addresses the impact of acquiring a new 3D Broadband seismic survey over an amplitude-supported, discovered gas field containing legacy 3D conventional towed streamer seismic data. The new seismic data were acquired in shallow water depths using Western Geco's dual level streamer technique and processed through PreSDM. Five gas discovery/appraisal wells with reservoirs ranging from approximately 1.2 to 2.5 s TWT (3,500 ft TVDss to 9,500 ft TVDss) existed prior to broadband acquisition and two additional wells were drilled after acquisition was completed. These seven wells serve as control points that provide a valuable link between the seismic and reservoir properties. Acquisition of new seismic data was primarily motivated by imaging challenges, particularly in the deeper (>1.6 s TWT) section, to which legacy surveys and (re)processing attempts have failed to completely find a solution. These imaging problems stem from: 1) Abundant shallow gas pockets, which contribute to both amplitude and frequency decay in the underlying image; 2) Fault shadow noise resulting from velocity variations which are difficult to capture in the velocity model; and 3) The presence of sub-regional coal layers having high-impedance contrasts that further attenuate the signal and contribute to generation of multiples. A complimentary paper discusses the novel processing techniques applied to overcome these problems. Here, we discuss the comparative benefit of the broadband acquisition versus the legacy conventional acquisition, namely higher signal to noise ratio throughout the entire record. In this context, we demonstrate through interpretation of broadband data that overall, complex geologic layers are better resolved versus conventionally acquired seismic data. We also compare conventional and broadband wavelets and discuss the implications for layer detection and resolution. This is particularly important for imaging and interpreting thin (<20 ft) reservoirs. We compare horizon-based attribute maps with legacy interpretation and address the implications for reservoir model building. We also apply various degrees of de-multiple algorithms and assess the resultant AVO effects. The broadband data has provided significant uplift in terms of reservoir detection and has thus bolstered confidence in our interpretation of thin, gas charged reservoirs.