An intelligent model for predicting wax deposition thickness during turbulent flow of oil

Abstract

To mitigate wax deposition, having an accurate model is of vital importance. In this study, an artificial neural network (ANN), with 19 and 8 neurons at its hidden layers, was developed to predict wax deposition thickness (WDT) during single-phase turbulent flow of oil. The proposed ANN takes wax content, Reynolds number, oil/pipeline temperature, and deposition time as input arguments. Predicted WDT by ANN was in close agreement with experimental data, with AARD% and RMS of 4.5369 and 0.011, respectively. Prediction of ANN was compared with that of adaptive neuro-fuzzy inference system (ANFIS). Results demonstrate superiority of ANN over ANFIS.

KEYWORDS:

• artificial neural network
• fuzzy inference system
• intelligent model
• turbulent flow
• wax deposition