Wax precipitation modelling using Perturbed Chain Statistical Associating Fluid Theory (PC-SAFT)

Abstract

Wax precipitation is one of the most challenging flow assurance problems because of its ability to create restrictions to flow inside wellbores, pipelines, and some production facilities. Inaccuracy in predictions of wax appearance temperature (WAT) and amount of precipitated wax makes it necessary to reassess existing thermodynamic models. In addition, most of the current models require accurate description of wax composition from expensive PNA analysis. Here we propose to substitute traditional cubic equations-of-state with Perturbed Chain form of the Statistical Associating Fluid Theory (PC-SAFT). The advantage of PC-SAFT is mainly the accuracy in the estimation of fugacities of heavy components in vapor and liquid mixtures. These fugacities are important because they define the equilibrium between wax, liquid and vapor phases. The novelty of this research was in approach for fluid characterization. Such models as multi-solid or solution-solid were not used. Instead, wax phase was represented as one phase and its amount was taken from inexpensive crosspolarized microscopy data. Therefore, we did not need PNA analysis. To be able to accurately predict with one wax component, reservoir was divided into sectors to determine PC-SAFT parameters for this wax component from calibration of each sector separately. Later, when a new well is drilled, its content of wax can be determined from the cross-polarized experiment and PC-SAFT parameters are same as PC-SAFT parameters of that sector (they were obtained from calibration for that sector before). This data alone is enough to predict amount of precipitated wax at any conditions with high accuracy. First, we validated PC-SAFT with experimental PVT data such as bubble point pressure, gas-to-oil ratio (GOR) and oil formation volume factor (Bo) and compared to the results of Peng Robinson EoS. This PVT data is from one of the fields in the South Caspian Basin. The first validation of wax precipitation itself, however, was performed on experimental data in the literature. Later, the model was calibrated on the oil sample data (composition and wax data from cross-polarized lab experiment) from the field in the South Caspian Basin. Finally, we verified the model with the data from the rest of the wells in this field. The results prove the accuracy of PC-SAFT method and show that costs of PNA analysis can be avoided if cross-polarized microscopy is available.