Abstract
The emergence of sophisticated integrated asset simulation software places radically new demands on our fluid modeling capability. Part of this challenge arises from the different modeling objectives and priorities within the broader group of users of such extended systems. For instance, a reservoir engineer will typically focus on detailed resolution of geological features for accurate flow simulation, requiring large grids while using a simple, isothermal fluid model to capture gross volumetric behavior. A production engineer on the same project team, by contrast, will be concerned with thermal losses in wells and flowlines and the attendant effect on fluid flow properties, as well as the impact of fluid blending in common production facilities. Common to both disciplines is the desire for a simple, fit-for-purpose conceptual Black Oil model that makes use of available data without mandating detailed experimental analyses at differing producing conditions. To facilitate fluid model sharing across discipline boundaries we have developed a Universal Black Oil formalism. In this approach, Black Oil correlations are tuned to existing data in a guided workflow, providing a Black Oil model capable of being used in all parts of an asset model. The problem of fluid blending in co-mingled production systems is treated by converting Black Oil models into analogous two-component compositional representations, and by exploiting weaving under the assumption of ideal mixing. The presence of multiple fluid models within an interconnected asset introduces the need for a set of heterogeneous fluid transformation models that must to be solved simultaneously along with reservoir and production system equations. The validity and limitations of the Black Oil blending approach are discussed in the context of an EOS model characterized to actual PVT data and the flexibility of the overall approach is demonstrated in an integrated simulation example.
