Event-Based Forward Modeling – Visualizing and Predicting Turbidite Channel Architectures

Event-based forward modeling is a rules-based approach to construct a self-consistent 3-dimensional computer model that incorporates both erosion and deposition of reservoir-scale objects in a time series. By selecting appropriate input parameters, complex 3D turbidite channel architectures can be constructed, interrogated quantitatively and compared to subsurface reservoirs. Multiple realizations of event-based forward models can serve to constrain, validate, or condition geocellular reservoir models prior to simulation.

Rules may be conceptual or empirical, based on observations of numerous outcrops, high resolution 3D seismic volumes, near modern oceanographic data, and wells. An example rule is the influence of under-filled channel relief on the location of the subsequent channel. Channel elements with a low channel-fill percentage tend to exert a strong control on the location and morphology of the next channel element. The result is an “organized” channel stacking pattern in which the separate elements can be characterized by the same range of parameters. In contrast, channel elements with a high channel-fill percentage tend to have little influence on the location and morphology of the next channel. The result is a “disorganized” stacking pattern in which separate channel element morphologies and locations are controlled only by other rules. The implications of these two end member stacking patterns are significant. If other inputs are held constant, the volume of preserved channel sand is much higher in a complex with disorganized channel stacking compared to a complex with organized channel stacking.

The event-based platform serves as a laboratory for (1) testing the robustness of incorporated rules, (2) constructing an analog for specific hydrocarbon reservoirs that objectively constrain input parameters for reservoir models and (3) investigating the relative importance of specific rules and parameters on reservoir volume and performance.