Comprehensive modeling of flow assurance : scales, hydrates, and asphaltenes

In the oil and gas industry, “flow assurance” incorporates the efforts to prevent inadvertent disruptions to the hydrocarbon flow from the wells to processing facilities. It is a reference to the study of both organic and inorganic deposits that may hinder production. Such deposits are mainly formed by scale, hydrate, asphaltene, and wax. This dissertation enhances the modeling of flow-assurance issues within a single platform—UTWELL, a wellbore simulator developed at The University of Texas at Austin. Scales result from mineral precipitation due to changes in pressure and temperature along the water flow, inherent to any gas/oil production. This research focuses on how water evaporation and CO₂ affect scaling tendencies in an oil well. The results demonstrate that evaporation is only relevant for very a small amount of water in the system, characteristic of the early stages of production. Additionally, the model shows that gas lift can increase mineral precipitation depending on the CO₂ content from the injected gas. Hydrates are ice-like solids formed under high-pressure and low-temperature conditions that are commonly found in an oilfield. Hydrate formation can be inhibited either by dissolved ions (electrolytes) in the produced water or by deliberate injection of chemicals. This research develops a hydrate-check model to verify formation conditions along the flow. The integration with a geochemical package (PHREEQC) provides the tools to consider electrolyte inhibition, and a newly included equation of state (CPA) assesses the inhibition effect from added glycols (and alcohols). The model predicts that when gas-water ratio (GWR) exceeds 10⁵ scf/STB, water condensation reduces electrolyte inhibition significantly. On asphaltenes, this research also discusses two prediction methods: a consolidated model from Li and Firoozabadi (2010), using a simplified version of the cubic-plus-association equation of state (CPA EoS); and a newly proposed version of a solid model, based on the Peng-Robinson EoS. It is shown that, if provided with adequate onset data, the solid model can match results from the CPA model quite successfully, while requiring only half the computational time. However, the solid model cannot adjust to composition changes in the same manner as CPA. Therefore, its adoption seems more suitable for wellbore simulation than in the reservoir, where fluid mixing is widespread.