Premium
Determination of Input Parameters for Crude‐Oil Fouling Using an Integrated Non‐Parametric Method
Author(s) -
Forstenhäusler Marc,
Mirsadraee Alireza,
Malayeri M. Reza
Publication year - 2017
Publication title -
chemical engineering and technology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.403
H-Index - 81
eISSN - 1521-4125
pISSN - 0930-7516
DOI - 10.1002/ceat.201600392
Subject(s) - fouling , crude oil , asphaltene , parametric statistics , instability , adsorption , materials science , petroleum engineering , chemistry , pulp and paper industry , chromatography , mathematics , chemical engineering , mechanics , engineering , membrane , biochemistry , statistics , organic chemistry , physics
An integrated, non‐parametric model has been proposed based on neural networks and partial derivatives (PaD) to determine the influence of the operating parameters during crude‐oil fouling. The input parameters included surface and bulk temperatures, fluid velocity, colloidal instability index, and surface substrate, whereas the output was the initial fouling rate. The PaD method was used to slightly change an input while the others were kept constant. The input that changed the output most had the highest impact on the objective output, i.e., the initial fouling rate. This method identified the surface temperature as the most dominant parameter. This indicates that crude‐oil fouling dominantly resulted from chemical reactions due to asphaltene instability.