Premium
Modeling of the Removal of Arsenic Species from Simulated Groundwater Containing As, Fe, and Mn: A Neural Network Based Approach
Author(s) -
Mondal Prasenjit,
Mohanty Bikash,
Balomajumder Chandrajit,
Saraswati Samir
Publication year - 2012
Publication title -
clean – soil, air, water
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.444
H-Index - 66
eISSN - 1863-0669
pISSN - 1863-0650
DOI - 10.1002/clen.201000536
Subject(s) - arsenic , adsorption , factorial experiment , artificial neural network , chemistry , groundwater , detection limit , mathematics , chromatography , computer science , engineering , statistics , organic chemistry , geotechnical engineering , machine learning
The present paper deals with the modeling of the removal of total arsenic As(T), trivalent arsenic As(III), and pentavalent arsenic As(V) from synthetic solutions containing total arsenic (0.167–2.0 mg/L), Fe (0.9–2.7 mg/L), and Mn (0.2–0.6 mg/L) in a batch reactor using Fe impregnated granular activated charcoal (GAC‐Fe). Mass ratio of As(III) and As(V) in the solution was 1:1. Multi‐layer neural network (MLNN) has been used and full factorial design technique has been applied for the selection of input data set. The developed models are able to predict the adsorption of arsenic species with an error limit of −0.3 to +1.7%. Combination of MLNN with design of experiment has been able to generalize the MLNN with less number of experimental points.