The modelling of water treatment processes is challenging because of its complexity, nonlinearity, and numerous contributory variables, but it is of particular importance since water of low quality causes health-related and economic problems which have a considerable impact on people’s daily lives. Linear and nonlinear modelling methods are used here to model residual aluminium and turbidity in treated water, using both laboratory and process data as input variables. The approach includes variable selection to find the most important factors affecting the quality parameters. Correlations of 0.7–0.9 between the modelled and real values for the target parameters were ultimately achieved. This data analysis procedure seems to provide an efficient means of modelling the water treatment process and defining its most essential variables. 1. Introduction Water quality is becoming an ever more important issue, as water of low quality causes many significant problems. In particular, there is a wide range of microbial and chemical constituents of drinking water that can cause either acute or chronic detrimental health effects, and the detection of these constituents in treated water is often time-consuming, complex, and expensive [1]. On the other hand, water of bad quality can also be harmful from an economic perspective, as resources have to be directed towards improving the water supply system every time a problem occurs. For these reasons, there is growing pressure to improve water treatment and water quality management in order to ensure safe drinking water at reasonable costs. Systematic assessments of raw water, treatment processes, and operational monitoring issues are needed to meet these challenges. There are many parameters which can be used to measure the quality of water, of which turbidity is a common one, the purpose being to measure impurities in the water. In a physical sense, turbidity is a reduction in the clarity of water due to the presence of suspended or colloidal particles, and it is commonly used as an indicator of the general condition of drinking water [1]. Furthermore, turbidity has been used for many decades as an indicator of the efficiency of drinking water coagulation and filtration processes, so that it is an important operational parameter for this reason, too. High turbidity values refer to poor disinfection and possibly to fouling problems in the distribution network, so that turbidity should be minimized [2]. However, turbidity is a quite sensible and faulty measurement, and many variables and phenomena are influencing it.
References
[1]
World Health Organization, Guidelines for drinking-water quality, vol. 1, Recommendations, 3rd edition, 2006.
[2]
R. D. Letterman, Ed., Water Quality & Treatment, Handbook of Community Water Supplies, AWWA, 1999.
[3]
A. Campell, “The role of aluminium and copper on neuroinflammation and Alzheimer’s disease,” Journal of Alzheimer’s Disease, vol. 10, pp. 165–172, 2006.
[4]
J. E. Van Benschoten and J. K. Edzwald, “Chemical aspects of coagulation using aluminum salts - I. Hydrolytic reactions of alum and polyaluminum chloride,” Water Research, vol. 24, no. 12, pp. 1519–1526, 1990.
[5]
J. E. Van Benschoten and J. K. Edzwald, “Chemical aspects of coagulation using aluminum salts - II. Coagulation of fulvic acid using alum and polyaluminum chloride,” Water Research, vol. 24, no. 12, pp. 1527–1535, 1990.
[6]
C. Huang and H. Shiu, “Interactions between alum and organics in coagulation,” Colloids and Surfaces A, vol. 113, no. 1-2, pp. 155–163, 1996.
[7]
P. Juntunen, M. Liukkonen, M. Lehtola, and Y. Hiltunen, “Cluster analysis of a water treatment process by self-organizing maps,” in Proceedings of the 8th IWA Symposium on Systems Analysis and Integrated Assessment, E. Ayesa and I. Rodríquez-Roda, Eds., pp. 553–558, WATERMATEX, 2011.
[8]
P. Juntunen, M. Liukkonen, M. Lehtola, and Y. Hiltunen, “Dynamic modelling approach for detecting turbidity in drinking water,” in Proceedings of the 52nd International Conference of Scandinavian Simulation Society, E. Dahlquist, Ed., 2011.
[9]
C. W. Baxter, Q. Zhang, S. J. Stanley, R. Shariff, R. R. T. Tupas, and H. L. Stark, “Drinking water quality and treatment: the use of artificial neural networks,” Canadian Journal of Civil Engineering, vol. 28, supplement 1, pp. 26–35, 2001.
[10]
D. N. Thomas, S. J. Judd, and N. Fawcett, “Flocculation modelling: a review,” Water Research, vol. 33, no. 7, pp. 1579–1592, 1999.
[11]
H. R. Maier, N. Morgan, and C. W. K. Chow, “Use of artificial neural networks for predicting optimal alum doses and treated water quality parameters,” Environmental Modelling and Software, vol. 19, no. 5, pp. 485–494, 2004.
[12]
S. Haykin, Neural Networks and Learning Machines, Pearson Education, Upper Saddle River, NJ, USA, 3rd edition, 2009.
[13]
P. Kadlec, B. Gabrys, and S. Strandt, “Data-driven Soft Sensors in the process industry,” Computers and Chemical Engineering, vol. 33, no. 4, pp. 795–814, 2009.
[14]
M. R. G. Meireles, P. E. M. Almeida, and M. G. Sim？es, “A comprehensive review for industrial applicability of artificial neural networks,” IEEE Transactions on Industrial Electronics, vol. 50, no. 3, pp. 585–601, 2003.
[15]
M. Heikkinen, H. Poutiainen, M. Liukkonen, T. Heikkinen, and Y. Hiltunen, “Self-organizing maps in the analysis of an industrial wastewater treatment process,” Mathematics and Computers in Simulation, vol. 82, no. 3, pp. 450–459, 2011.
[16]
A. M. Kalteh, P. Hjorth, and R. Berndtsson, “Review of the self-organizing map (SOM) approach in water resources: analysis, modelling and application,” Environmental Modelling and Software, vol. 23, no. 7, pp. 835–845, 2008.
[17]
H. R. Maier and G. C. Dandy, “Neural networks for the prediction and forecasting of water resources variables: a review of modelling issues and applications,” Environmental Modelling and Software, vol. 15, no. 1, pp. 101–124, 2000.
[18]
M. S. Gibbs, G. C. Dandy, and H. R. Maier, “Calibration and optimization of the pumping and disinfection of a real water supply system,” Journal of Water Resources Planning and Management, vol. 136, no. 4, Article ID 023003QWR, pp. 493–501, 2010.
[19]
C. W. Baxter, S. J. Stanley, and Q. Zhang, “Development of a full-scale artificial neural network model for the removal of natural organic matter by enhanced coagulation,” Journal of Water Supply: AQUA, vol. 48, no. 4, pp. 129–136, 1999.
[20]
J. L. Giraudel and S. Lek, “A comparison of self-organizing map algorithm and some conventional statistical methods for ecological community ordination,” Ecological Modelling, vol. 146, no. 1–3, pp. 329–339, 2001.
[21]
P. J. Werbos, Beyond regression: new tools for prediction and analysis in the behavioral sciences, Doctoral thesis, Harvard University, Cambridge, Mass, USA, 1974.
[22]
A. K. Jain, R. P. W. Duin, and J. Mao, “Statistical pattern recognition: a review,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 22, no. 1, pp. 4–37, 2000.
[23]
A. L. Blum and P. Langley, “Selection of relevant features and examples in machine learning,” Artificial Intelligence, vol. 97, no. 1-2, pp. 245–271, 1997.
[24]
I. Guyon and A. Elisseeff, “An introduction to variable and feature selection,” Journal of Machine Learning Research, vol. 3, pp. 1157–1182, 2003.
[25]
H. Liu and H. Motoda, Eds., Computational Methods of Feature Selection, Chapman & Hall, Boca Raton, Fla, USA, 2008.
[26]
A. W. Whitney, “Direct method of nonparametric measurement selection,” IEEE Transactions on Computers, vol. C-20, no. 9, pp. 1100–1103, 1971.
[27]
D. J. C. MacKay, “A practical bayesian framework for backpropagation networks,” Neural Computation, vol. 4, no. 3, pp. 448–472, 1992.
