Modeling of Asphaltene Systems with Association Models

Asphaltenes are normally present in the reservoir oil and for industry they are analogous to“cholesterol” since their precipitation stops the entire production and causes the loss of millions of dollars. Asphaltenes consist of “ill defined” components of high molecular weight (around 500-4000gm/mol) and are considered most polar in the oil compared to the other components. This polar nature of asphaltenes is imparted by heteroatoms (O, S, N, vanadium, nickel) present in their structure. Asphaltenes molecules associate with each other and precipitate at certain temperature, pressure and composition. Asphaltenes can easily precipitate as pressure is reduced but also if the oil is diluted bylight hydrocarbons eg. A gas such as methane, CO₂ or nitrogen. Ever since the introduction of enhanced oil recovery (EOR) method with gas, the asphaltene precipitation problem has become evenworse. In addition, at refinery inlet, more than two crudes (degassed oil) are generally mixed to upgrade or downgrade the feedstock. This blending of different crudes may also cause asphaltene precipitation. However, prediction of these conditions, where asphaltenes precipitate out, is quite uncertain and detailed thermodynamic model and appropriate oil characterization is required.
There have been several attempts to model asphaltene precipitation using various equations of state (EoS) and empirical models. In the past few years, the association models based on the Cubic Plus Association (CPA) and Statistical Associating Fluid Theory (SAFT) EoS are found to be promising models for the asphaltene precipitation study. However, there are different opinions in the literature and it is still unclear whether we can successfully model asphaltene precipitation. Therefore, a systematic study is important to model asphaltene precipitation and show what we can achieve with these models.
In this PhD project, a modeling approach using the CPA and PC-SAFT EoS is developed to model asphaltene precipitation from reservoir oil and degassed crude considering asphaltenes as associating component/s. Several reservoir oils are studied in order to show that the developed approach can be used to predict gas injection and reservoir depressurization effect after calculating the model parameters from the required experimental data. The developed approach with Soave-Redlich-Kwong (SRK) with classical mixing rule and SRK with Huron Vidal mixing rule is also studied considering asphaltenes as non-associating component/s. The well known approach from the literature based on the PC-SAFT EoS, where asphaltenes are considered non-associating component, is also been studied and found that it needs temperature dependency like the developed approach in this work to correlate asphaltenes onset conditions at different temperatures. In addition, asphaltene yield from crude oil during the addition of n-paraffin is also studied and concluded that the models cannot predict it but can only correlate the data. The CPA and PC-SAFT models are also studied to predict asphaltene onset conditions when more than two crudes are mixed. A MATLAB tool is also developed in order to use these models to study asphaltene phase envelope with or without gas injection.