Reservoir Modeling of a Structurally Complex Deepwater Field: A Gulf of Mexico Case History

Abstract

Abstract Reservoir modeling of structurally complex fields is particularly challenging when dealing with reverse or Y-shaped faults. In such cases, the modeling approaches usually tend to incorrectly reproduce or to simplify the actual geological situation. The scope of this contribution is to describe a case history where reservoir modeling activities are effectively performed for a field having high structural complexity, including Y faults. The novel structural modeling approach has been applied for a deepwater oilfield located in the Gulf of Mexico. The Miocene-Pliocene reservoir consists of a series of stacked turbidite sandstone layers, which were deposited in a small basin against a salt ridge. A main set of WNW-ESE trending normal faults, as well as the allochtonous salt presence in the southwestern part, are responsible for the high compartmentalization of the field. Because of the complex fault framework, especially the Y faults geometry, a particular type of stair-step grid was adopted for the static modeling. In presence of complex geological structures, such as reverse faults and/or Y faults, the use of the common pillar gridding modeling approach would lead to difficulties to build an effective geological model and, consequently, reliable dynamic reservoir simulations. In the current case, the adopted approach allowed building a static reservoir model accounting for its high structural complexity. The resulting grid (structural grid) is a corner point grid having specific characteristics that differentiate it from a conventional pillar grid. The main difference is the existence of apparent duplicated i, j, k indices in the structural grid. The apparent layer index (Geological Layer) property correctly reproduces the geological situation from a reservoir static standpoint. However, the flow simulators use the real grid layer index (Grid Layer), which may not correspond to the Geological Layer property, especially for complex fault patterns. Then, if such situation is well-known and handled with a proper strategy, the dynamic model will run effectively. The positive results achieved with the described approach, despite the complicated nature of the studied field, must be highlighted. In fact, in the previous 3D grid models the field history match was never achieved, while the current structural reconstruction allowed a successful reproduction of the field history and, consequently, a more robust definition of the future activities to optimize the field production.