Fouling resistance prediction using artificial neural network nonlinear auto-regressive with exogenous input model based on operating conditions and fluid properties correlations

Totok R. Biyanto*

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

9 Citations (Scopus)

Abstract

Fouling in a heat exchanger in Crude Preheat Train (CPT) refinery is an unsolved problem that reduces the plant efficiency, increases fuel consumption and CO2 emission. The fouling resistance behavior is very complex. It is difficult to develop a model using first principle equation to predict the fouling resistance due to different operating conditions and different crude blends. In this paper, Artificial Neural Networks (ANN) MultiLayer Perceptron (MLP) with input structure using Nonlinear Auto-Regressive with eXogenous (NARX) is utilized to build the fouling resistance model in shell and tube heat exchanger (STHX). The input data of the model are flow rates and temperatures of the streams of the heat exchanger, physical properties of product and crude blend data. This model serves as a predicting tool to optimize operating conditions and preventive maintenance of STHX. The results show that the model can capture the complexity of fouling characteristics in heat exchanger due to thermodynamic conditions and variations in crude oil properties (blends). It was found that the Root Mean Square Error (RMSE) are suitable to capture the nonlinearity and complexity of the STHX fouling resistance during phases of training and validation.

Original languageEnglish
Title of host publicationProceedings of the 3rd AUN/SEED-NET Regional Conference on Energy Engineering and the 7th International Conference on Thermofluids, RCEnE/THERMOFLUID 2015
EditorsTetsushi Biwa, Samsul Kamal, Hideaki Ohgaki, Jun Tanimoto, Chaiwat Nuthong, Adhika Widyaparaga, Keiichi N. Ishihara, Harwin Saptoadi, Jayan Sentanuhady, Deendarlianto, Zainal Alimuddin b. Zainal Alauddin, Iman Reksowardojo, Indarto, Yasuyuki Takata, Indro Pranoto
PublisherAmerican Institute of Physics Inc.
ISBN (Electronic)9780735413917
DOIs
Publication statusPublished - 3 Jun 2016
Event3rd AUN/SEED-NET Regional Conference on Energy Engineering and the 7th International Conference on Thermofluids, RCEnE/THERMOFLUID 2015 - Yogyakarta, Indonesia
Duration: 19 Nov 201520 Nov 2015

Publication series

NameAIP Conference Proceedings
Volume1737
ISSN (Print)0094-243X
ISSN (Electronic)1551-7616

Conference

Conference3rd AUN/SEED-NET Regional Conference on Energy Engineering and the 7th International Conference on Thermofluids, RCEnE/THERMOFLUID 2015
Country/TerritoryIndonesia
CityYogyakarta
Period19/11/1520/11/15

Keywords

  • Artificial Neural Network
  • Fouling Resistance
  • Modeling
  • Shell and Tube Heat Exchanger

Fingerprint

Dive into the research topics of 'Fouling resistance prediction using artificial neural network nonlinear auto-regressive with exogenous input model based on operating conditions and fluid properties correlations'. Together they form a unique fingerprint.

Cite this