An Advanced Reservoir Simulator for Tracer Transport in Multicomponent Multiphase Compositional Flow and Applications to the Cranfield CO2 Sequestration Site

Abstract:

Reservoir simulators are widely used to constrain uncertainty in the petrophysical properties of subsurface formations by matching the history of injection and production data. However, such measurements may be insufficient to uniquely characterize a reservoir’s properties. Monitoring of natural (isotopic) and introduced tracers is a developing technology to further interrogate the subsurface for applications such as enhanced oil recovery from conventional and unconventional resources, and CO₂ sequestration. Oak Ridge National Laboratory has been piloting this tracer technology during and following CO₂ injection at the Cranfield, Mississippi, CO₂ sequestration test site. Two campaigns of multiple perfluorocarbon tracers were injected together with CO₂ and monitored at two wells at 68 m and 112 m from the injection site. The tracer data suggest that multiple CO₂ flow paths developed towards the monitoring wells, indicative of either channeling through high permeability pathways or of fingering. The results demonstrate that tracers provide an important complement to transient pressure data. Numerical modeling is essential to further explain and interpret the observations.

To aid the development of tracer technology, we enhanced a compositional multiphase reservoir simulator to account for tracer transport. Our research simulator uses higher-order finite element (FE) methods that can capture the small-scale onset of fingering on the coarse grids required for field-scale modeling, and allows for unstructured grids and anisotropic heterogeneous permeability fields. Mass transfer between fluid phases and phase behavior are modeled with rigorous equation-of-state based phase-split calculations. We present our tracer simulator and preliminary results related to the Cranfield experiments. Applications to noble gas tracers in unconventional resources are presented by Darrah et al.