In-Reservoir Mixing Dynamics Over Geologic Time of Separate Gas and Oil Charges in Well-Connected Reservoirs

Abstract

Abstract Many reservoirs experience separate gas and oil charges that can lead to a variety of different outcomes of fluid type and distribution. There has been fundamental uncertainty even as to which charge fluid can arrive first, let alone what fluid dynamic processes can result over geologic time. For high-pressure basins such as the Gulf of Mexico, this mixture can lead to increased solution gas, large GOR gradients and sometimes cause formation of viscous oil and tar at the oil-water contact, impacting aquifer support. In some reservoirs, the present-day outcome of oil and gas mixing over geologic time is clearly established by detailed chemical evaluation of reservoir fluids from many reservoir locations. Our objective is to understand the dynamics of the gas and oil mixing processes. Chemical measurements show that the extent of mixing includes thermodynamic equilibration in young reservoirs by 1) FHZ equation of state (EoS) asphaltene gradients and cubic EoS modeling of solution gas for reservoir fluids, 2) analysis of liquid-phase geochemical biomarkers, and 3) methane carbon isotope analysis. Specifically, in the common charge of primary biogenic gas and oil into reservoirs, methane isotope analysis is unequivocal. We employ reservoir simulation of a point gas charge into oil with various geometries and charge rates to establish parametric conditions which lead to excellent mixing vs those conditions that lead to large, disequilibrium gradients. The roles of compositional diffusion vs. momentum diffusion induced by forced convection are explored both in simulation and overall fluid mechanics analysis, which helps both to validate the results and extend the range of applicable parameters. Modeling results and simple fluid mechanics estimates also establish that there is no possibility that these reservoirs could have a gas charge followed by an oil charge; in the selected reservoirs, oil must have arrived first, followed by a biogenic gas charge. Seismic images of gas chimneys offer guidance regarding how the latter process can take place. Second, modeling results clearly establish a surprisingly wide range of charge conditions that can lead to excellent mixing and equilibration even for a point gas charge. Modeling results also show that for a very fast charge, results are consistent with those expected for CO2 injection and sequestration. The evaluation of geodynamic processes of separate biogenic gas and oil charges into reservoirs has rarely been accomplished. Even the result that biogenic gas charge must occur after oil charge challenges widely-held conventional thinking. In addition, the rapid and thorough mixing (less than 2 million years) of gas and oil charges is unexpected yet readily reproduced by reservoir simulation. The ability to connect CO2 sequestration to a wide range of reservoir studies is a novel way to constrain CCS modeling.