Abstract: Integration of Static, Dynamic and Pseudo-dynamic Data to Detect Reservoir Compartmentalization
SMALLEY, P. CRAIG
Reservoir compartmentalization is an issue that affects field economics at all stages from appraisal through development to production. This presentation shows the benefit of integrating all available data to determine compartmentalization. Such data can be grouped into 3 types:
Static data (e.g., sedimentology, high-resolution stratigraphy, heterogeneity modeling, production analogues, fault seal analysis, connectivity analysis) tell about reservoir shape, structure and heterogeneity - but not fluid flow. These data are valuable because they start to become available early in field life.
Dynamic data (e.g., tracer studies, production data, well test analysis, production logs) are the most definitive identifiers of compartmentalization. Unfortunately such data only become available relatively late in field life, too late to influence development decisions.
Pseudo-dynamic data (e.g., fluid contacts, natural pressures, petroleum PVT properties and geochemistry, Sr residual salt analysis) tell about fluid movement in the reservoir on a geological time scale. With care this can be extrapolated to predict reservoir behavior on a production time scale. These data are of extreme value, as they can provide dynamic information early in field life, before production start-up.
Integration of these data types requires a cross-disciplinary approach. Informal integration may consist simply of obtaining and interpreting the available data types, and forming a conceptual reservoir model that best fits the combined data. A more rigorous approach is to use a set of integration tools: a set of standard and non-standard simulation tools (e.g., reservoir simulation, tracer response simulation, reservoir filling and mixing simulation, dynamic connectivity analysis), that allow the different groups of data (static, pseudo-dynamic and dynamic) to be compared together, often enabling compartmentalization models to be tested or "history-matched" against other dynamic or pseudo-dynamic data.
A series of cameo case studies demonstrates the different types of data and how they can be integrated. Future compartmentalization studies are likely to utilize a 3-D reservoir computer model that can be used as a repository for all of the relevant data types, and can communicate freely with the different integration tool simulators.
AAPG Search and Discovery Article #90942©1997 AAPG International Conference and Exhibition, Vienna, Austria