Use of data mining in the corrosion classification of pipelines in catalytic reforming units (CRU)
- 1 April 2020
- journal article
- research article
- Published by Sami Publishing Company in Eurasian Chemical Communications
- Vol. 2 (4), 571-581
- https://doi.org/10.33945/sami/ecc.2019.6.7
Abstract
Nowadays, computational tools for analyzing and collecting data in the operation of petroleum units are essential. One of the methods is the classification or regression is to step in the overall process of knowledge extraction. In this article, one of specific type of decision called the conditional contract arrangement, is used to extract the relevant knowledge in Catalytic Reforming Units (CRU) for 4 factors: Density, PH, total iron ions in vessels (S.FE) and H2S. All of these factors are related to corrosion in CRU units and this paper aims to optimize some conditions to eliminate it. In this regard, using ammonium water with a specific range and pH can be helpful. According to the obtained results the best range of density (in Feed) is less than 0.515 kg/m3, PH (water in vessels) is more than 6.7, S.FE is less than 1.5 ppm and H2S in recycle gas is less than 700 ppm. The outcomes also show how this approach can be used to gain insight into some refineries and how to deliver results in a comprehensible and user-friendly way.Keywords
This publication has 30 references indexed in Scilit:
- Optimizing the catalyst circulation ratio in a reformer with a moving bed via a combination of real and computational experimentsCatalysis in Industry, 2012
- Decision tree-based technology credit scoring for start-up firms: Korean caseExpert Systems with Applications, 2012
- Decision tree approach for classification and dimensionality reduction of electronic nose dataSensors and Actuators B: Chemical, 2011
- A one-pot multicomponent reaction for synthesis of 1-amidoalkyl-2-naphthols catalyzed by PEG-based dicationic acidic ionic liquids under solvent-free conditionsMonatshefte für Chemie / Chemical Monthly, 2011
- Prediction intervals to account for uncertainties in neural network predictions: Methodology and application in bus travel time predictionEngineering Applications of Artificial Intelligence, 2011
- A survey of cross-validation procedures for model selectionStatistics Surveys, 2010
- Decision tree and first-principles model-based approach for reactor runaway analysis and forecastingEngineering Applications of Artificial Intelligence, 2009
- Development of a Kinetic Model for Catalytic Reforming of Naphtha and Parameter Estimation Using Industrial Plant DataEnergy & Fuels, 2009
- A G2(MP2) theoretical study of substituent effects on H3BNHnCl3−n (n= 3-0) donor-acceptor complexesOpen Chemistry, 2008
- Potassium dodecatungstocobaltate trihydrate (K5CoW12O40·3H2O): A mild and efficient reusable catalyst for the synthesis of aryl-14H-dibenzo[a.j]xanthenes under conventional heating and microwave irradiationCatalysis Communications, 2007