Characterizing Compartmentalization in Structurally Heterogeneous Reservoirs Using Fluid Mixing Time-Scales

Abstract

Detection and characterization of reservoir compartmentalization during appraisal is significantly improved by using fluid data (pressure, contacts, density and composition) and the rate at which these observed fluid variations equilibrate over geological time-scales. This essentially involves comparison of the time-scales for any observed fluid property variation(s) to homogenise with the time since the reservoir filled. A suite of published analytical expressions for fluid mixing via molecular diffusion, gravitational overturning or pressure diffusion have been used previously to quantify mixing time-scales. These have subsequently been applied to field studies to identify and quantify barriers and baffles to flow. These analytical mixing relations however have been derived for idealized reservoir geometries (e.g. 1D and 2D box models) where the fluid mixing time-scales are simply estimated over straight line distances between two observed points (e.g. wells). In reality, most reservoirs are structurally heterogeneous (e.g. with folding, anticlines, faulting) and thus mixing times may be increased due to the non-linear mixing distances within the reservoir. It is not clear whether such analytical estimates of mixing time are reliable in these cases. In this study, we investigate the time taken for fluid contacts and fluid densities in a faulted anticlinal reservoir to reach equilibrium using detailed numerical simulation, compared with existing analytical solutions for a box reservoir. We present an easy method for estimating an effective mixing distance and thus the mixing time in such cases without recourse to simulation. A simple field case study from a giant Middle Eastern oil field is presented demonstrating these principles. This confirms previous work that observed fluid contact variations in the field do not necessarily indicate the presence of barriers to flow. Using the effective mixing distance of ∼100 km, the estimated mixing time is long (∼1My) compared to the time since the reservoir filled or aquifer started flowing and thus the overturning of fluid contacts in the field has not yet reached steady state.

AAPG Datapages/Search and Discovery Article #90189 © 2014 AAPG Annual Convention and Exhibition, Houston, Texas, USA, April 6–9, 2014