Application of the SAFT-γ Mie Equation of State for Reservoir-Fluid Modeling in the Petroleum Industry

In the equation of state toolbox used in upstream oil and gas modeling, a truly predictive equation of state is still absent. SAFT-γ Mie is a predictive equation of state that has been validated for a wide range of applications. Here we aim to demonstrate that the SAFT-γ Mie group-contribution approach can also be used in petroleum-fluid modeling. This necessitates addressing the specific traits of oil and gas mixtures: their multicomponent nature and the presence of undefined fractions in these mixtures. A model validation is first conducted on multicomponent mixtures that closely resemble reservoir fluids, both in composition and component number. For these well-defined mixtures, we demonstrate that the predictive use of SAFT-γ Mie is able to achieve parity in performance with the engineering equations of state widely used in industry. Thereafter, we introduce the notion of employing modern structure-elucidation techniques in tandem with SAFT-γ Mie to model undefined petroleum fractions. Computer-aided-molecular-design is proposed to describe pseudocomponents for crude-oil modeling. This is applied to predict PVT test data for a model crude oil that has properties representative of typical reservoir fluids. Again, the SAFT-γ Mie equation of state is seen to perform respectably when compared to an engineering equation of state, PPR78. For clarity on the robustness of the proposed ‘hypothetical-structure + SAFT-γ Mie’ framework, we predict asphaltene-instability onset behavior using synthetic asphaltene structures generated with a quantitative molecular representation. In service of this, a new aromatic-carbon group is parametrized using data from small fused-aromatic-ring systems. When applied to our model asphaltene in simplified “oils”, the SAFT-γ Mie approach can be used to predict all prototypical asphaltene instability behavior. Altogether, our work provides evidence that the SAFT-γ Mie equation of state can contend with the important traits of petroleum fluids, but does so with the added benefit of predictive modeling.