Microsoft Word - 6 ART_1 Adrian Ioana.doc 38 journal homepage: www.fia.usv.ro/fiajournal Journal of Faculty of Food Engineering, Ştefan cel Mare University of Suceava, Romania Volume XII, Issue 1 – 2013, pag. 38 - 46 EAF ENVIRONMENTAL IMPACT ANALISYS THROUGH MATHEMATICAL MODELING AND FUZZY LOGIC *Adrian IOANA1, Augustin SEMENESCU1, Cezar Florin PREDA1 1University Politehnica of Bucharest, Spl. Independentei 313, Bucharest, S6, 060042, adyioana@gmail.com * Corresponding author Received 10 January 2013, accepted 15 February 2013 Abstract. The original optimisations mathematical model of the electric arc furnace’s charge preheating process mainly takes into consider 2 thermo-technological aspects: the heat transfer between fluids and particles and the heat transfer between the fizz layer and an exchange surface. According the energetically balance at the gaseous environment level, the conductive transfer model is also analysed through the finished elements method. The results of the mathematical model are presented as the analysis and quantification of the thermo gradients obtained during the charge preheating process. These thermo gradients are determined for various temporal moments and for different capacities of the electric arc furnace. The results of the mathematical model are presented as the analysis and quantification of the thermo gradients obtained during the charge preheating process. These thermo gradients are determined for various temporal moments and for different capacities of the electric arc furnace. Keywords: Environmental Impact, Electric Arc Furnace (EAF), Modelling, Fuzzy Logic 1. Introduction The electric arc furnaces (EAFs), as powerful energy consumers, are also polluting emissions generators with an important environmental impact. The most significant polluting emissions of the EAF furnace are metallic and oxides powders driven by emergent gases [1-6]. The powders are produced during the following technological operations: raw materials loading, steel melting, refining, alloying and evacuation. Generally, the driven powders contain heavy metals (Cr, Ni, Zn, Cd, Pb, Cu etc) and some metal oxides (iron, manganese, aluminum and silicon oxides) and they can reach values of more than 15 kg/t steel [7-8]. Fuzzy Logic (FL) is a powerful problem- solving methodology with wide applications in economical control and information processing [9-10]. It provides a simple way to draw definite conclusions from vague, ambiguous or imprecise information. It resembles human decision making with its ability to work from approximate data and find precise solutions [11, 12]. The neuro-fuzzy approach is to use neural networks and fuzzy set theory to model practical systems. A pattern match or recognition system is a black box constructed using multiple layers of neurons called neural networks. Neurons have the ability of memory and self- learning by training. FL algorithms are implemented as an inference engine which can automatically infer from facts (data) [13-15]. Mathematical modelling of the electric arc furnace’s processes (EAFP) for the optimisation of the functional and Food and Environment Safety - Journal of Faculty of Food Engineering, Ştefan cel Mare University - Suceava Volume XII, Issue 1 – 2013 ADRIAN IOANA, AUGUSTIN SEMENESCU, CEZAR FLORIN PREDA, EAF ENVIRONMENTAL IMPACT ANALISYS THROUGH MATHEMATICAL MODELING AND FUZZY LOGIC, Food and Environment Safety, Volume XII, Issue 1 – 2013, pag. 38 – 46 39 technological performances of this complex unit is based on the next principles [16-19]: A. The principle of analogy – consists in observing and analysing competently the modelated reality, using both analogy with other fields of research and logical homology. According to this principle, for mathematical models making were used the following steps: • the modelated subject definition – represents the first phase of the modelation analysis. This step must satisfy both the purpose and the simultaneous system’s aims, assuring their compatibility; • the effiency criteria’s definition – is a step imposed on the correct definition of the system’s aims and allows the optimisation of the modelling solutions; • making the options – basing on accessing some realistically, original and efficient solutions; • choices evaluating – related to the established efficiency criterials; • choosing the final solution – based on the analysis between the different solutions of the modelling. B. The principle of concepts is based on the sistems’theory, including the feedback concept. C. The principle of hierarchisation consists of making hierarchical models systems, for structuring the decision and coordinating the interactive subsystems [20-22]. Fuzzy Logic has been found to be very suitable for embedded control applications [23-25]. Several manufacturers in the automotive industry are using fuzzy technology to improve quality and reduce development time. In aerospace, fuzzy enables very complex real time problems to be tackled using a simple approach. In consumer electronics, fuzzy improves time to market and helps reduce costs. In manufacturing, fuzzy is proven to be invaluable in increasing equipment efficiency and diagnosing malfunctions. The gaseous phase (burnt gases) that comes out of the EAF mainly results from the melting and refining procedures and contains carbon monoxide, carbon dioxide together with nitrogen and sulfur oxides (NOx and SOx); however, in practice it also contains very toxic other components, such as fluorides or volatile organic compounds (dioxine, chloride derivatives of benzene or phenol) resulted from burning of organic oils that are introduced as contaminants together with the raw materials. 2. Materials and methods Environmental impact of eaf From the total polluting emissions, over 90% are generated during the technological operations of melting and refining. The chemical composition of these emissions is extremely variable and directly dependent on multiple factors, as followings [25-29]: - composition of the raw materials that make up the loading; - the melting managing way; - type of refining process that is used (with gaseous oxygen or ore); - time duration of the melting and refining steps; - desired quality degree of the elaborated steel. Figure 1 presents the main scheme of the environment polluting system through the EAF. Table 1 presents the variation limits of chemical composition for the powders generated during the steel elaboration in electric arc furnaces (EAF) in the USA and Germany, from loading that consists of scrap iron, only. Food and Environment Safety - Journal of Faculty of Food Engineering, Ştefan cel Mare University - Suceava Volume XII, Issue 1 – 2013 ADRIAN IOANA, AUGUSTIN SEMENESCU, CEZAR FLORIN PREDA, EAF ENVIRONMENTAL IMPACT ANALISYS THROUGH MATHEMATICAL MODELING AND FUZZY LOGIC, Food and Environment Safety, Volume XII, Issue 1 – 2013, pag. 38 – 46 40 Figure 1. Environment polluting system through the EAF Table 1 Variation limits of chemical composition for the EAF powders Variation limits, % GERMANY No. Component SUA Plain Basic Steel Alloy Steel 1 Fetotal 16.4 – 38.6 21.6 – 43.6 35.3 2 Si 0.9 – 4.2 0.9 – 1.7 17.0 3 Al 0.5 – 6.9 0.1 – 1.5 x) 4 Ca 2.6 – 15.7 6.6 – 14.5 0.4 5 Mg 1.2 – 9.0 1.0 – 4.5 1.2 6 Mn 2.3 – 9.3 0.9 – 4.8 2.0 7 P 0 – 1.0 0.1 – 0.5 x) 8 S 0– 1.0 0.3 – 1.1 0.1 9 Zn 0 – 35.3 5.8 – 26.2 1.4 10 Cr 0 – 8.2 0 – 0.1 13.4 11 Ni 0 – 2.4 x) 0.1 12 Pb 0 – 3.7 1.3 – 5.0 0.4 Environmental modelling system of eaf The modelling system’s central element of the EAF processes conceived consists of the system’s criteria function. Knowing that the technological processes study for EAF is subordinated to high quality steel obtaining, the modelling system’s criteria function (CF) is the ratio between quality and price: max        PRICE QUALITY CF (1) The maximum of the criteria function is assured by the mathematical model of prescribing the criteria function (M.P.C.F.) The mathematical model of prescribing the criteria function concept consists of transforming the criteria function (CF) in a quality-economical matrice MQE, as in the scheme presented in figure 2. WASTE GASES ELECTRIC ARC FURNACE (EAF) LIQUID STEEL NOx CO CO2 POWDE R SLAG SCRAP IRON CARBURIZING AGENTS LIME IRON ALLOYS O2 AIR . . . NOISE Food and Environment Safety - Journal of Faculty of Food Engineering, Ştefan cel Mare University - Suceava Volume XII, Issue 1 – 2013 ADRIAN IOANA, AUGUSTIN SEMENESCU, CEZAR FLORIN PREDA, EAF ENVIRONMENTAL IMPACT ANALISYS THROUGH MATHEMATICAL MODELING AND FUZZY LOGIC, Food and Environment Safety, Volume XII, Issue 1 – 2013, pag. 38 – 46 41 Figure 2. The modelling system’s criteria function’s evaluation The levels of prescribing the criteria function could be obtained by using a composition algorithm for three vectors: • vector – technical parameters’ vector (ti); • vector – economical parameters’ vector (ej); • vector – weight vector (p1). In figure 3 there is presented the general logical scheme used for the EAF’charge preheating. Fuzzy logic modelling for ecological impact of eaf Patterns reflect the behavioral characteristics of how a person or a system acts in a certain environment. A spending pattern may represent the way a consumer spends money on different goods, such as travels, cars, or food. A defect pattern in a semiconductor equipment may indicate the way in which a part or assembly fails. By matching various patterns, a marketing specialist at a credit card company is able to better understand consumers spending habits, and therefore he can tailor his or her marketing strategies targeted to different consumer groups. Figure 3. General logical scheme By the same token, scientists study and match various patterns of machine faults in STAR READ: h, , cp, , q, t,  T , , x, y, N, M n = 1 n  N [T] = [F], [K]-1 N Y READ xjn,yin i= 1…3, j = 1…3 WRITE [T] (N+2, N+2) STOP bin = yi+1,n – yi+2,n cjn = xj+2,n – xj+1,n i= 1…3, j = 1…3 An = ½ (x2n  y3n – x3n  y2n – x1n  y3n + x3n  y1n + x1n  y2n) 213 132 )( )( nnn nnn xxI xxI   For i = 1…3 and j = ki,j = ksni, j+ki,j For j = 1,..., N + 2 DO n = n fi,j = fsni, j+fi,j T t1 t2 … tn P p1 p2 … pn E e1 e2 … em MCE (mu,w)   ti  ej  pk i = 1 … n; j = 1 … m; k = 1 … l Food and Environment Safety - Journal of Faculty of Food Engineering, Ştefan cel Mare University - Suceava Volume XII, Issue 1 – 2013 ADRIAN IOANA, AUGUSTIN SEMENESCU, CEZAR FLORIN PREDA, EAF ENVIRONMENTAL IMPACT ANALISYS THROUGH MATHEMATICAL MODELING AND FUZZY LOGIC, Food and Environment Safety, Volume XII, Issue 1 – 2013, pag. 38 – 46 42 order to be able to predict and control the performance of equipment, including advanced warning. The application of the fuzzy logic (FL) is based on 3 simple steps defined below (figure 4): Figure 4. The steps of the FL FLS – Fuzzy Logic Steps; F – Fuzzification; A – Aggregation; D - Defuzzification Neuro-fuzzy modeling is to use neural networks and fuzzy set theory to model practical systems. Neuro-fuzzy technology was developed that can automatically extract business rules from the neuro-fuzzy model. These rules can explain which and why people tend to fraud, or why a customer price increases. Mathematically, the fuzzy utility function is a more accurate measure on the consumption utility. It can describe the relationships between: spending (S), price (P), consumption composition (CC), preference and subjective measure on commodity (C) or service values (SV). These relationships are described by the FUFC matrix concept (figure 5). Fuzzy Consumption Utility Functions (FCUF) are based on Utility Theory (UT). Fuzzy Utility Function for Consumption (FUFC) is described by the FUFC matrix concept. This concept is based on the following vectors: S (spending vector); P (price vector); CC (consumption composition vector); C (commodity vector) and SV (service values vector). Figure 5. FUCF Matrix concept FUCF – Fuzzy Utility Function for Consumption; OFM – Objective Factors Matrix; SFM – Subjective Factors Matrix; S – Spending vector; P – Price vector; CC – Consumption Composition vector; C – Commodity vector; SV – Service Values vector. 3. Results and discussion The correlation between the criteria function’s (C.F.) prescribed levels and T vector’s components’ variation is presented in figure 9. The execution of the EAF’s charge preheating modelling (CPM) was made both for a 10t EAF (fig.11a, b, c) and for a 50t EAF (fig. 11d). It was considered to be a load with medium permeability ε = 0.45. The scheme of the EAF environmental impact is presented in figure 6. In figure 7 is presented the CO concentration in the evacuated gas during the melting in the EAF. FLS F A D [FUFC [S [P [CC [C [SV [OFM [SFM] Food and Environment Safety - Journal of Faculty of Food Engineering, Ştefan cel Mare University - Suceava Volume XII, Issue 1 – 2013 ADRIAN IOANA, AUGUSTIN SEMENESCU, CEZAR FLORIN PREDA, EAF ENVIRONMENTAL IMPACT ANALISYS THROUGH MATHEMATICAL MODELING AND FUZZY LOGIC, Food and Environment Safety, Volume XII, Issue 1 – 2013, pag. 38 – 46 43 E X P E R I M E N T A L R Figure 6. Scheme of the EAF environmental impact T im e 10 20 30 40 50 60 70 80 90 L I L IV 0 2000 4000 6000 8000 10000 12000 14000 16000 CO[%] Time [min] L I L II L III L IV Figure 7. CO concentration in the evacuated gas during the melting in the EAF Figure 8 presents the carbon dioxide impact of EAF. Figure 8. Carbon dioxide impact of EAF The correlation between the criteria function‘s (CF) prescribed levels and the T vector’s components’ variation ( T 1, T 2, vector’s components’ variation ( T 1, T 2, T 3 T 3) are presented in figure 9. Figure 9. The correlation between the criteria function ‘s (CF) prescribed levels and the T The cumulated correlation between the criteria function’s (C.F.) prescribed levels and T and E vectors’ variation are presented in figure 10. Figure 10. The cumulated correlation between the criteria function’s (CF) prescribed level and T and E vectors’ variation. EAF WASTE GASES POWDER SLAG NOISE EA F Food and Environment Safety - Journal of Faculty of Food Engineering, Ştefan cel Mare University - Suceava Volume XII, Issue 1 – 2013 ADRIAN IOANA, AUGUSTIN SEMENESCU, CEZAR FLORIN PREDA, EAF ENVIRONMENTAL IMPACT ANALISYS THROUGH MATHEMATICAL MODELING AND FUZZY LOGIC, Food and Environment Safety, Volume XII, Issue 1 – 2013, pag. 38 – 46 44 Figure 11 presents the main CPM’s results. a) c) d) b) Figure 11. The main CPM’s results Food and Environment Safety - Journal of Faculty of Food Engineering, Ştefan cel Mare University - Suceava Volume XII, Issue 1 – 2013 ADRIAN IOANA, AUGUSTIN SEMENESCU, CEZAR FLORIN PREDA, EAF ENVIRONMENTAL IMPACT ANALISYS THROUGH MATHEMATICAL MODELING AND FUZZY LOGIC, Food and Environment Safety, Volume XII, Issue 1 – 2013, pag. 38 – 46 45 The components of two vectors T and E which are considered to have important weight in the criteria function’s evaluation are:  t1 – the steels chemical composition;  t2 – the steels purity (in gases);  t3 – the steels purity (inclusions );  e1 – the specific consumption of basic material and materials;  e2 – the specific consumption of energy;  e3 – the elaboration process’s productivity in EAF. The best level (NO) for each component of the 2 vectors is: - for t1 – the prescribed variation limits of the elaborated steel quality composition arithmetical mean. The cumulated correlation between the criteria function’s (C.F.) prescribed levels and T and E vectors’ variation are presented in figure 10. - for e1 - the minimum content specific consumption prescribed of basic materials - for e2 - the minimum prescribed specific energy consumption. - for e3 - the maximum prescribed productivity of the elaboration process. We can notice the obtaining of:  the criteria function’s maximum level FOT,max = 43,76 for the T vector’s variation (t1 component - the prescribed variation limits of the elaborated steel quality composition arithmetical mean).  the criteria function’s maximum level FOE,max = 55,31 for the E vectors’ variation (e3 component - the maximum prescribed productivity of the elaboration process). And respective the criteria function’s maximum level FOCUM,max = 19,85 for the T and E vectors’ cumulated variation. 4. Conclusions From the total polluting emissions of EAF, over 90% are generated during the technological operations of melting and refining. The gaseous phase (burnt gases) that comes out of the EAF mainly results from the melting and refining procedures and contains carbon monoxide, carbon dioxide together with nitrogen and sulfur oxides (NOx and SOx); however, in practice it also contains very toxic other components, such as fluorides or volatile organic compounds (dioxine, chloride derivatives of benzene or phenol) resulted from burning of organic oils that are introduced as contaminants together with the raw materials. Fuzzy Logic has been found to be very suitable for embedded control applications. Every application can potentially realize some of the benefits of FL: performance, productivity, simplicity and lower cost. FL algorithms are implemented as an inference engine which can automatically infer from facts. The application of the fuzzy logic is based on three steps: fuzzification; aggregation and defuzzification. Neuro-fuzzy modeling is to use neural networks and fuzzy set theory to model practical systems. Neuro-fuzzy technology was developed that can automatically extract business rules from the neuro-fuzzy model. Mathematically, the fuzzy utility function is a more accurate measure on the consumption utility. The levels of prescribing the criteria function could be obtained by using a composition algorithm for three vectors: • vector – technical parameters’ vector (ti); • vector – economical parameters’ vector (ej); • vector – weight vector (p1). The execution of the EAF’s charge preheating modelling (CPM) was made both for a 10t EAF and for a 50t EAF. Food and Environment Safety - Journal of Faculty of Food Engineering, Ştefan cel Mare University - Suceava Volume XII, Issue 1 – 2013 ADRIAN IOANA, AUGUSTIN SEMENESCU, CEZAR FLORIN PREDA, EAF ENVIRONMENTAL IMPACT ANALISYS THROUGH MATHEMATICAL MODELING AND FUZZY LOGIC, Food and Environment Safety, Volume XII, Issue 1 – 2013, pag. 38 – 46 46 5. References [1]. A., IOANA, The Electric arc furnaces (EAF) functional and technological performances with the preheating of the load and powder blowing optimisation for the high quality steel processing, Ph D Thesis, POLITEHNICA University of Bucharest, 1998, pp. 21-187. [2]. A., IOANA, Ecological aspects and solutions in EAF’s line of work, The Danube river environment & education, Tulcea, nr. 3, 2005, pp. 17-20. [3]. A., IOANA, Technical – economical analysis options for the quality of the steel elaborated in the EAF, Int. Conf. IMT, Oradea, Proc., Fasc. Management and Technol. Eng., Vol. V (XV), 2006, pp. 132-135. [4]. A., IOANA, Environment & EAF, Proc. 5th IASME / WSEAS Int. Conf. on HEAT TRANSFER,THERMAL ENGINEERING and ENVIRONMENT - HTE’07, August 25-27, Vouliagmeni, WSEAS Press, Athens, 2007, pp. 51-54. [5]. A., IOANA, Managementul producţiei în industria materialelor metalice. Teorie şi aplicaţii., Editura PRINTECH Bucureşti, ISBN 978-973-758- 1232, 2007, pp. 17-232. [6]. N., MARDANI, Modeling the Impact of Environmental Variables on Type and Density of LandCcover, Carpathian Journal of Earth and Environmental Sciences, February 2012, Vol. 7, No. 1, 2012, pp. 137 – 148. [7]. A., IOANA, C., BĂLESCU, Management Elements of the Materials Costs, ISI Book of the „3rd International Conference on Management, Marketing and Finances”, [ISI-Web of Science/Science Citation Index Expanded], Huston, USA, 2009, pp. 71-75. [8]. A., IOANA, C., BĂLESCU, Optimization of steel elaboration in EAF through mathematical modelling, Junior Euromat Conf., Proc., Lausanne, 4- 8 Sept., 2006, pp. 85-87. [9]. A., IOANA, V., MIREA, C., BĂLESCU, Economic Processes Study through Fuzzy Logic, Journal of Economic Computation and Economic Cybernetics Studies and Research, no. 2., [ISSN 0424- 267X, ISI-Web of Science/Science Citation Index Expanded], Academy of Economic Studies Publishing, Bucharest, 2010, pp. 129-138. [10]. A., IOANA, V., MIREA, C., BĂLESCU, Analysis of Service Quality Management in the Materials Industry Using the BCG Matrix Method, Amfiteatru Economic Review, Vol. XI, Nr. 26, [ISSN 1582-9146, ISI-Web of Science/Science Citation Index Expanded], Bucureşti, 2009, pp. 270-278. [11]. M., CABEZA, M.B., ARAU, R.J., WILSON, C.D., THOMAS, M.J.R. COWLEY, A., MOILANEN, Combining probabilities of occurrence with spatial reserve design. Ecology, 2004, pp. 41. [12]. M.P., AUSTIN, Spatial prediction of species distribution an interface between ecological theory and statistical modeling, Ecol. Model., 2002, pp. 101- 118. [13]. A., IOANA, A., NICOLAE, Optimal working of EAF , Ed. Fair Partners, Bucharest, 2002, pp. 51-117. [14]. A., IOANA, C., BĂLESCU, Sustainable Development and Risk Management, Computers and Simulation in Modern Science, Vol. V, ISI Book WSEAS, [ISI-Web of Science/Science Citation Index Expanded], 2011, pp. 35-41. [15]. A., IOANA, A., NICOLAE, C., BĂLESCU, Elements of Electric Arc Furnaces’ Management, Metalurgia International nr. 7/2008, Vol. 13, [ISSN 1582-2214, ISI-Web of Science/Science Citation Index Expanded], Bucureşti, 2008, pp. 12-15. [16]. Gh., BOLDUR–LÃŢESCU, Gh., CIOBANU, I., BÃNCILÃ, Analiza sistemelor complexe, Editura Ştiinţificã şi Enciclopedicã, Bucureşti, 1982. [17]. B.V., BARNES, Forest ecology, John Wiley and Sons, INC., 1998, pp. 610-773. [18]. D., TALOI, C., BRATU, E., FLOIAN, E., BERCEANU, Optimizarea proceselor matalurgice, Editura Didacticã şi Pedagogicã, Bucureşti, 1983, pp. 15-127. [19]. D., MÃRGULESCU, I., VÃDUVA, Analiza economico – financiarã asistatã de calculator, Editura Tehnicã, Bucureşti, 1985, pp. 27-83. [20]. R., RISTIĆ, Environmental Impacts in Serbian Ski Resorts, Carpathian Journal of Earth and Environmental Sciences, Vol. 6, No. 2, 2011, pp. 125 – 134. [21]. B.B., SALEM, Application of GIS to biodiversity monitoring. Journal of Arid Environments, 54(1), 2003, pp. 91-114. [22]. D.A., SARR, D.E., HIBBS, M.N.A., HUSTON, Hierarchical perspective of plant diversity, Quarterly Review of Biology.80, 2005, pp. 187-212. [23]. A., IOANA, C., BĂLESCU, Environmental Study of the Formation of Evacuated Burnt Gases from a Steels Making Plant, Revista de Chimie Nr.5, [ISSN 0034-7752, ISI-Web of Science/Science Citation Index Expanded], Bucureşti, 2009, pp. 468- 471. [24]. R., DIRZO, E., MENDOZA, Biodiversity. Encyclopedia of Ecology, 2008, pp. 368-377. [25]. A., GUISAN, N.E., Zimmermann, Predicting habitat distribution models in ecology. Ecological Modeling, 2000, pp. 147-186. [26]. K.H., CHI, M.B., CHANG, S.H., CHANG, Measurement of atmospheric PCDD/F and PCB distributions in the vicinity area of Waelz plant during different operating stages, Science of the Total Environment 391 (1), 2008, pp. 114-123. [27]. Da Silva, MACHADO, J.G.M., BREHM, F.A., MORAES, C.A.M., DOS SANTOS, C.A., VILELA, Characterization study of electric arc furnace dust phases, Materials Research, 9 (1), 2006, pp. 41-45. [28]. C.A., PICKLES, Thermodynamic analysis of the separation of zinc and lead from electric arc furnace dust by selective reduction with metallic iron, Separation and Purification Technology 59 (2), 2008, pp. 115-128. [29]. F., SKOV, J., SVENNING, Predicting plant species richness in a managed forest. Forest Ecology and Management, 180(1-3), 2003, pp. 583-593.