A Power Law Committee Scaling Equation for Quantitative Estimation of Asphaltene Precipitation

Abstract

Precipitation and deposition of asphaltene as a challenging issue have drawn much attention in oil industry owing to severe problems that created during petroleum production and processing. Due to its adverse impact on petroleum production, proposed potent model which capable to predict amount of asphaltene precipitation with high accuracy is necessary. Recently different researcher proposed novel model so-called scaling equation for predicting the amount of asphaltene precipitation. Although derived equation is valuable it possess with flaw as limit accuracy of prediction. The current study proposes a novel technique to predict the accurate value of asphaltene precipitation amount by integration of different scaling equation using the concept of power law committee machine (PLCM). Elements of PLCM model are Rassamdana scaling (RE) model, Hu scaling (HU) model, and Ashoori scaling (AS) model. PLCM model has a parallel architecture that combined the outputs of the aforementioned models in order to reaping the benefits of individual models and increases the accuracy of final asphaltene precipitation amount prediction. Optimal contribution of individual scaling equations in final output is computed by virtue of genetic algorithm (GA) tool. Finally, determined result from PLCM model is compared with individual scaling approaches. It is observed, implementation of PLCM can lead to more accurate prediction compared to different scaling models which conducted alone for predicting amount of asphaltene precipitation.