The prediction and mitigation of asphaltene precipitation is an ongoing issue in upstream operations with native oils and in refinery processes of fluids from thermocracking and hydrocracking processes. Asphaltenes can precipitate with changes in pressure, temperature and composition, promoting problems with deposition and fouling during crude oil operations. Fouling problems are a key concern in the oil industry because of their associated cost due to shut downs that decrease production and increase cost for maintenance for cleaning or replacing equipment. Hence, reliable methods are needed to predict asphaltene precipitation at different conditions during upstream and downstream crude oil operations.
The modified regular solution approach (RSM) has been successfully applied to predict asphaltene precipitation from native oils or upstream processes. This approach requires mole fractions, molar volumes, and solubility parameters for the pseudo-components representing the crude oil mixture. The pseudo-components used in the RSM, based on a solubility fractionation method (modified ASTM D4124 technique), are namely saturates, aromatics, resins, and asphaltenes (SARA). SARA fractionation applies to heavy oils and bitumens which do not have significant amounts of volatile compounds distillable at atmospheric pressure. However, the reactions in refinery processes convert a significant amount of the bitumen and heavy oil into lighter compounds creating a significant amount of distillables in the product streams. These reactions also alter the chemical structure of the crude oil and, therefore, the properties and correlations used in the RSM must be modified for these oils.
The objective of this thesis is to characterize several native and reacted oils and develop a general model for asphaltene precipitation from both native and reacted oils using the standard SARA characterization method for the heavy fraction and distillation assays for the lighter fraction of the oils. The oils investigated included native, thermocracked (in situ converted and visbreaker samples), and hydrocracked samples.
Each of the distillable fractions and SARA fractions were characterized in detail to identify the changes with reaction in comparison with native oils and to estimate the properties required for the regular solution model. Properties including densities, molecular weight, and solubilities were measured for the SAR fractions. Asphaltenes tend to self-associate and form nano-aggregates and, therefore, a more detailed characterization was required to determine the asphaltene property distributions. Molecular weight and density distributions of asphaltenes in toluene were obtained from a previous study. Solubility parameter distributions were determined by fitting the RSM to asphaltene precipitation yield data measured for “pure” asphaltenes in mixtures of n-heptane and toluene (heptol). Yield data was also measured for the heavy fractions of the oils and whole oils diluted with n-heptane and used to tune the model.
The updated model is applicable to both native and reacted oils and includes a new correlation for the solubility parameter of the distillables as a function of their boiling point, and new asphaltene density and solubility parameter correlations, as a function of mass fraction and molecular weight, respectively. The required input data are: a distillation assay of the oil, SARA assay of the non-distillable residue, and density of the asphaltenes. Distillables, if present, are represented with 5 pseudo-components which are added to the SARA characterization. The density and molecular weight of the saturates, aromatics and Resins (SAR) fractions can be estimated from average values from this study. The asphaltene nano-aggregate molecular weight distribution in the crude oil cannot be measured or estimated a priori and is represented with a gamma distribution which requires the average molecular weight of asphaltenes nano-aggregates and the distribution shape factor. The two unknown molecular weight distribution parameters are found by tuning the model to fit the heavy oil (or heavy fraction) solubility data. Once the model is tuned to fit the solubility data of the heavy fraction of the oil, the model predicts the whole crude oil (distillables + SARA) stability.
The model and the property correlations developed in this project can potentially be implemented in a simulation software to predict asphaltene precipitation for different crude oil samples. A predictive model for asphaltene precipitation is an important tool for production and refinery engineers to find operating conditions that avoid asphaltene precipitation and the consequent lost production and increased costs from fouling and plugging of production and refinery equipment. Additionally, the characterization data for the pseudo-components of native and reacted oils can be used to develop models for other fluid modeling applications including property correlations and fouling models.