Over the last decade, there have been several discoveries of significant oil accumulations in deepwater reservoirs. With advances in research and the availability of high-resolution seismic data, their variability and complexity have been well documented. For hydrocarbon exploitation purposes, however, the industry lacks an integrated approach to interpret these reservoirs, using multiple domains, and efficiently consume the available data. This paper describes a series of innovative interrelated techniques, plus observations to help improve the understanding of these reservoir types at three scales:

• Basin scale: Stratigraphic forward modeling is a useful method to simulate regional geological processes. Using this technique a paleo-basin floor surface reconstructed from seismic can be used to set a series of simulations. By varying several parameters, such as water level and water/sediments coefficients, within a reasonable range of values and comparing results to observed data, plausible models are created. These, considered jointly, provide an integrated understanding of the uncertainty of the occurrence of observed features and sediment properties.
• Sequence scale: Sinuosity and stacking configurations are critical factors that control local channel-fill patterns. By using seismic sections along channels, coupled with seismic reconstruction, complex paleo-sequence geometries can be resolved. In our case a low-sinuosity channel was identified and a backstepping-fill pattern defined.
• Reservoir scale: The quantification of net-to-gross is critical in hydrocarbon exploitation. By using training grids and multipoint statistical methods a comprehensive facies model can be built. Training grids, based on fill patterns and channel geometry, are combined with probability properties to condition the overall stochastic distribution of facies. Additionally, uncertainties associated with this process are captured in an uncertainty workflow.

Integrating these three techniques leads to a better understanding of reservoir-critical factors. The above approach is proven by using a seismic dataset from Campos Basin, Brazil, which contains well-documented reservoirs originally deposited in a deepwater setting.