Nashrieh Shimi va Mohandesi Shimi Iran

Nashrieh Shimi va Mohandesi Shimi Iran

Using Online Machine Learning Methods for Prediction of the Loss of Alumina and the Caustic Soda and Increasing the Carbonate Contamination in the Bayer Process

Document Type : Research Article

Authors
Reza Monsefi
Mostafa Mahmoudian
Nima Salehi Moghadam
Chemical Engineering Department, University of Ferdowsi, Mashhad, I.R. IRAN
Abstract
In this study, machine learning tools were used to investigate the effect of various factors on the waste of caustic soda (NaOH) and alumina (Al2O3) and increasing carbonate (Na2CO3) pollution in the first stage of the Bayer process. The parameters studied in this paper are the concentration of Na2Oc, Na2Ou, Na2Ot and Al2O3 in the output solution to the wet grinding unit. The investigated factors included mass flow rate and chemical analysis of various compounds in bauxite and lime consumed, flow rate and chemical analysis of sodium alumina solution in a daily basis for 3016 days. In this study, Fourier Online Gradient Descent (FOGD) and Nysgd were used to model the process. The results show that the FOGD method has more accuracy and speed of learning than the NysGD method. These results indicate that each of these methods can model the process with high precision.
Keywords
Bayer Process
alumina
caustic soda
sodium carbonate
machine learning
FOGD and NysGD

Subjects

[1] Technoexport Praha. Alumina Plant Jajarm Process and Operating Manual. (1997).
[2] Kaußen F.M., Friedrich B., Methods for Alkaline Recovery of Aluminum from Bauxite Residue. Journal of Sustainable Metallurgy., 2: 353–364 (2016)
[3] Yousry L. Sidrak., Dynamic Simulation and Control of the Bayer Process, Ind. Eng. Chem. Res., 40: 1146-1156 (2001).
[4] Liang W.T., Couperthwaite S.J., Kaur G., Yan C., Johnstone D.W., Millar G.J., Effect of Strong Acids on Red Mud Structural and Fluoride Adsorption Properties, Journal of Colloid and Interface Science, 423: 158–165 (2014).
[5] Peter S., “Reactions of Lime under High Temperature Bayer Digestion Conditions”, 10th International Alumina Quality Workshop; 273-381,  19-23 Aprol, Western Australia (2015).
[6] Gerald R., Dunstan D.G., Charmaine R.G. William., Method for the Causticisation of Bayer Process Solutions, EP06704821A (2006).
[7] Power G., Gräfe M., Klauber C., Bauxite Residue Issues: I. Current Management, Disposal and Storage Practices, Hydrometallurgy, 108: 33–45 (2011).
[8] Radomirovic T., Smith P., Southam D., Tashi S., Jones F., Crystallization of Sodaliteparticles under Bayer-Type Conditions, Hydrometallurgy, 137: 84–91 (2013).
[9] Xu B., Smith P., Wingate C., De Silva L., The Effect of Calcium and Temperature on the Transformation of Sodalite to Cancrinite in Bayer Digestion, Hydrometallurgy., 105: 75-81 (2010).
[10] Rosenberg S.P., Tichbon W., Wilson D.J., Heath C.A., Process for the Removal of Oxalate and/or Sulphate from Bayer Liquors, United States Patent 72444040 (2007).
[11] Ostap S., Bartok D., Removal of Sodium Carbonate and Sodium Sulfate from Bayer Solutions. US3508884A (1967).
[12] Peter S., The Processing of High Silica Bauxites—Review of Existing and Potential Processes. Hydrometallurgy., 98: 162–176 (2009).
[13] Piga L., Pochetti F., Stoppa L., Recovering Metals from Red Mud Generated during Alumina Production, JOM., 45: 54-59 (1993).
[14] McGill W.B., Paul E.A., Fractionation of Soil and 15 N Nitrogen to Separate the Organic and Clay Interactions of Immobilized, Canadian Journal of Soil Science, 56: 203-212 (1976).
[15] Shin H.J., Lee S.O., Kim S.J., Tran T., Kim M.J., Study on the Effect of Humate and its Removal on the Precipitation of Aluminum Trihydroxide from the Bayer Process, Minerals Engineering., 17: 387–391 (2004).
[16] Brady J. P., “An examination of the Applicability of Hydrotalcite for Removing Oxalate Anions from Bayer Process Solutions”, Masters by Research Thesis, Queensland University of Technology, (2011).
[17] Armstrong L., Hunter J., McCormick K., Warren. H., Bound Soda Incorporation during Hydrate Precipitation-Effects of Caustic, Temperature, and Organics, Light Metals, 37–40 (1993).
[18] Grocott S.C., Rosenberg S.P., “Possible Mechanisms for Soda Incorporation in Smelter Grade Alumina”, 1st International Alumina Quality Workshop, 271–287 (1998).
[19] Hind A.R., Bhargava S.K., Grocott S.C., The Surface Chemistry of Bayer Process Solids, Journal of Colloids and Surfaces, 146: 359-374 (1999).
[20] Atkins P., Grocott S.C., Impact of Organic Impurities on the Product of Refined Alumina. Proceedings of Science, Technology and Utilisation of Humic Acids. CSIRO Division of Coal and Energy Technology, Australia, 85–94 (1988).
[21] Zhang Y.F., Cao S.T., Zhang Y., Zheng S.L., Method for Producing Alumina from Bauxite. China patent, Appl. No.: 200810227930.5 (2008).
[22] Stopa G.G., Brandão A.A.C., Prasad S., Improving the Bayer Process Productivity – An Industrial Case Study, Minerals Engineering., 22: 1130–1136 (2009).
[23] Solymar K., Orban M., Zoldi J., Baksa G., Methods for Reducing NaOH Losses in the Hungarian Alumina Plants, Travaux ICSOBA., 13: 377–390 (1983).
[24] Baksa G., Vallo F., Sitkei F., Zoldi J., Solymar K., “Complex Causticization: An Effective Means for the Reduction of NaOH Losses in an Alumina. Light Metals”, New Orleans, LA USA, 75–80 (1986).
[25] Picaro T., Red Mud Processing, WO9729992-A, (2000).
[26] Harato T., “The Development of a New Bayer Process that Reduces the Desilication Loss of Soda by 50% Compared to the Conventional Process”, Fourth International Alumina Quality Workshop, Darwin, N. T: 311–320 (1996).
[27] Armstrong L., Richter K., Taylor D., Mitchell V., Fane T., Glastras M., “A New Membrane Process to Purify Bayer Liquors”, 6th International Alumina Quality Workshop (2002).
[28] Zachary T.W., Nikolaos, V.S., The ALAMO Approach to Machine Learning, Computers and Chemical Engineering, 106: 785-795 (2017).
[29] Beck A.C.D., Data Science: Accelerating Innovation and Discovery in Chemical Engineering, AIChE Journal., 62: 1402-1416 (2016).
[30] Shokry h., Vicente P., Escudero G., Pérez-Moya M., Graells M., Espuña A., Data-Driven Soft-Sensors for Online Monitoring of Batch Processes with Different Initial Conditions, Computers and Chemical Engineering, 118: 159-179 (2018).
[31] Han H., Zhu S., Qiao J., Guo M., Data-Driven Intelligent Monitoring System for Key Variables in Wastewater Treatment Process, Chinese Journal of Chemical Engineering, 1-9 (2018).
[32] Onel M., Kieslich C.A., Guzman Y.A., Floudas C. A., Pistikopoulos E.N., Big Data Approach to Batch Process Monitoring: Simultaneous Fault Detection and Diagnosis Using Nonlinear Support Vector Machine-based Feature Selection, Computers and Chemical Engineering, 1-47 (2018).
[33] Garcia G.M., Munaro C.J., Isolation of Plant-Wide Faults using Causality Detection Methods, IFAC-Papers OnLine, 49: 13–18 (2016).
[34] Tsai Y., Lu Q., Rippon L., Lim S., Tulsyan A., Gopaluni B., Pattern and Knowledge Extraction using Process Data Analytics: Atutorial, IFAC-PapersOnLine, 51: 13–18 (2018).
[35] Cheng Y., Chen K., Sun H., Zhang Y., Tao F., Data and Knowledge Mining with Big Data Towards Smart Production, Journal of Industrial Information Integration, 9: 1–13 (2018).
[36] Clerget C.H., Grimaldi J.P., Ch`ebre M., Petit N., An Iterative Algorithm for Dynamic Optimization of Systems with Input-Dependent Hydraulic Delays, IFAC-PapersOnLine, 51(18): 7–12 (2018).
[37] Houben C., Lapkin A.A., Automatic Discovery and Optimization of Chemical Processes, Current Opinion in Chemical Engineering, 9: 1–7 (2015).
[38] Schweidtmann A.M., Clayton A.D., Holmes N., Bradford E., Bourne R.A., Lapkin A.A., Machine learning Meets Continuous Flow Chemistry: Automated Optimization Towards the Pareto Front of Multiple Objectives, Chemical Engineering Journal., 352: 277–282 (2018).
[39] Ye L., Cao Y., Skogestad S., Global Self-Optimizing Control for Uncertain Constrained Process Systems, IFAC-Papers OnLine., 50: 7672–7677 (2017).
[40] Vazan P., Janikova D., Tanuska P., Kebisek M., Cervenanska Z., Using Data Mining Methods for Manufacturing Process Control, IFAC-Papers OnLine., 50(1): 6178–6183 (2017)
[41] Zhou Zhenpeng., Li X., Zare R.N., Optimizing Chemical Reactions with Deep Reinforcement Learning, ACS Cent. Sci., 3: 1337−1344 (2018).
[42] Tan P., Xia J., Zhang C., Fang Q., Chen G., Modeling and Optimization of NOX Emission in a Coal-Fired Power Plant using Advanced Machine Learning Methods, Energy Procedia, 61: 377 – 380 (2014).
[43] Wei X., “Modeling and Optimization of Wastewater Treatment Process with a Data-Driven Approach”, Ph.D. Theses. University of Iowa. (2013).
[44] Cregan V., Lee W.T., Clune L., A Soft Sensor for the Bayer Process, Journal of Mathematics In Industry, 7-17 (2017).
[45] Chelgani S.C., Jorjani E., Artificial Neural Network Prediction of Al2O3 Leaching Recovery in the Bayer Process—Jajarm Alumina Plant (Iran), Hydrometallurgy., 97(1-2): 105-110 (2009).
[46] Mahmoudian M., Ghaemi A., Hashemabadi H., Shahhosseini S., Comparing the Capability of Various Models for Predicting the Bayer Process Parameters, Journal of Advanced Materials and Processing, 6: 71-86 (2018).
[47] Lu J., Large Scale Online Kernel Learning, The Journal of Machine Learning Research, 17(1): 1613-1655 (2016).
[48] Alpaydin E.., “Introduction to Machine Learning. 2nd ed.: Cambridge, Mass”, MIT Press, (2010).
[49] Wang Z., Crammer k., Vucetic S., Breaking the Curse of Kernelization: Budgeted Stochastic Gradient Descent, Journal of Machin Learning Research, 13(1): 3103-3131 (2012).
[50] LU J., Hoi S.C., Large Scale Online Kernel Learning, Journal of Machin Learning Research, 17(1): 1-43 (2016).
[51] Kivinen J., Online Learning with Kernels, IEEE Transactions on Signal Processing, 52: 2165-2176 (2004).