Asphaltenes are nanocolloidal structurespresent in various petroleum deposits. High asphaltene content in heavy oils causes the formation of clusters, the largest asphaltene nanocolloidal particles, driving a large gravity gradient of asphaltenes. Viscosity is highly correlated with asphaltene content; therefore, heavy oils are often associated with large viscosity gradients. Large oil-to-water viscosity ratios result in the fingering of injection water into oil columns and reduce sweep efficiency during production. Cluster accumulation at the base of a heavy oil column often leads to tar mat formation at the oil-water contact (OWC), sealing the aquifer from the oil and precluding pressure support and aquifer sweep. Thus, modeling the formation of viscous oils and tar mats is critical for production planning.
Asphaltenes are established to have a nanocolloidal structure in a large, anticlinal Middle East reservoir that is well connected as proven by equilibrated asphaltenes, production data, and pressure measurements, and contains black oil in the crest and 100-km rim of heavy oil underlain by ~10m tar mat. Asphaltenes in the black oil exist as nanoaggregates causing a small gradient under the influence of gravity, while asphaltene clusters in the heavy oil exhibit large gradients in density and viscosity. Both asphaltene gradients are matched by Flory-Huggins-Zuo equation of state (FHZ EoS). Over 60m of depth, asphaltene concentrations and viscosity values increase by factors of 10 and 1000, respectively. We use reservoir flow simulation of charge to show the formation of the viscous oil and tar mat. Our simulations honor the well-established nanocolloidal structure of asphaltenes and show how to generate gradients under the influence of gravity. We also investigate the impact of simulation parameters (e.g., reservoir geometry, dip angle) and charge conditions (e.g., charge duration) on asphaltene distributions.
We model the following reservoir fluid geodynamic processes leading to current fluid measurements. First, the reservoir is charged with heavy oil. Second, gravity gradients are established from vertical migration of heavy asphaltenes to lower positions (base of the crest or base of the formation interval). This creates a density inversion; the heavier oil at the base of the interval upstructure is more dense than the asphaltene-depleted oil at the top of the interval downstructure. Third, this density inversion causes Boycott convection which transports newly-created heavier oil to the base of the reservoir and lighter oil to the top of the reservoir. Subsequent to Boycott convection, asphaltene equilibration at the base of the column yields very high asphaltene concentrations at the oil-water contact. If the asphaltene concentration exceeds the solvency capacity of the oil for asphaltenes, then asphaltene phase-separation and deposition proceed creating a tar mat. The modeling mechanism is consistent with geochemical measurements, saturation pressure, and gas-oil ratio (~100 SCF/STB). Core plugs from six tar mat wells exhibit significant variation in asphaltene content from 30% to 65%. Tar mat wells establish that 1) tar mat is a two-phase system of very high asphaltene and oil phases and 2) the mechanism proposed agrees with tar mat well properties.
Asphaltene nanocolloidal gravity gradients are established vertically while Boycott convection transports dense oil to the base of the reservoir leading to formation of viscous oil and tar mat. The simulations also show that upwelling at the oil-water contact associated with countercurrent flow of Boycott convection precludes asphaltene colloidal settling at the OWC which is required to obtain very high asphaltene concentrations and tar mat formation. Thus, tar mats form when charge is complete but generally not prior to the completion of charge. Here we show for the first time that reservoir flow simulation of charge can honor the basic physics of mass transport in the reservoir and predict tar mat formation.
