546 Journal of Chemical Engineering of Japan Copyright © 2020 The Society of Chemical Engineers, Japan Journal of Chemical Engineering of Japan, Vol. 53, No. 9, pp. 546–554, 2020 Multi-Objective Optimization Model for the Energy System of Electric Arc Furnace Steelmaking Considering the Cost and Carbon Dioxide Emission under Uncertainty MyungSuk Son, YoungWook Bin, In-Beum Lee and Suh-Young Lee Department of Chemical Engineering, Pohang University of Science and Technology (POSTECH), Pohang 37673, Republic of Korea Keywords: EAF, MILP, Energy-saving Techniques, Stochastic Multi-objective Programming, Carbon Dioxide A stochastic multi-objective mathematical model based on material and energy flows under the system uncertainty of specific electricity demand-related scraps characteristics and operation conditions was developed to simulate the elec- tric arc furnace steelmaking process and optimize carbon dioxide emission and cost. Considering several energy-saving technologies based on stochastic thermodynamic energy efficiency and electricity price, this paper suggests an optimal cost-saving and carbon-dioxide-emission-reducing strategy. The suggested model provides a trade-off relationship be- tween cost and CO 2 emission. To minimize CO 2 emissions, a tunnel preheater without additional fuel consumption was suggest In contrast, to minimize the cost of utilizing cost-effective fossil fuels instead of electricity as the system energy requirement, an oxy-fuel burner and shaft furnace type of preheater were proposed. This problem was formulated as a mixed-integer linear programming model. Introduction Industrial energy use has been growing strongly in recent decades. According to the statistical data published by the international energy agency, the industry sector accounted for 37% (157 EJ·yr -1 ) of the total global final energy use, and direct industrial carbon dioxide (CO 2 ) emission reached 24% (8.5 Gt CO 2 ·yr -1 ) of global emissions in 2018 (IEA, 2020). e iron and steelmaking industry is one of the most energy-intensive industry sub-sectors; it contributed 24% of the total direct industrial CO 2 emissions in 2018 (IEA, 2020). Following the environmental requirements of stricter climate change conventions, from the Kyoto Protocol to the Paris Agreement, various efforts have been made to reduce carbon emissions in the steel industry (Worrell et al., 2001; Jones, 2012; Cavaliere, 2019). Several CO 2 -reduction technologies and routes have been developed in the steel industry. While the conventional blast furnace (BF)–basic oxygen furnace (BOF) process is a ma- ture technology and accounts for 74% of steel production worldwide (Li et al., 2018), it requires expensive capital in- vestment and emits considerable quantities of pollutants, in- cluding CO 2 , SO x , NO x , and CO. erefore, alternative tech- nology such as the scrap-based electric arc furnace (EAF) process has attracted interest in terms of energy-saving and CO 2 emissions reduction. (Riesbeck et al., 2011; Lingebrant et al., 2012; He and Wang, 2017; IEA, 2020). e EAF steelmaking process, which supplies heat to the charging materials for melting via an electric arc, can produce all kinds of steel and features flexibility and smaller capabilities as advantages. e EAF process is also advanta- geous for the following reasons: (1) it requires less capital investment and a shorter installation period; (2) it is envi- ronmentally friendly as it emits less NO x and SO x because it uses electricity; (3) it recycles raw materials, i.e., scrap, and gives high metallic yield; and (4) it is easy to control the quality of the steel product (Lee and Sohn, 2014; Li et al., 2018; Dutta and Chokshi, 2020). Typically, a scrap- based EAF process is less energy-intensive in that it requires 4–6 GJ·t -1 of iron produced when using 100% scrap, where- as the BF–BOF routes require 13–14 GJ·t -1 of iron produced (He and Wang, 2017). Several studies have been conducted to estimate the electricity requirements for the EAF process (Köhle, 2002; Sandberg et al., 2007; Gajic et al., 2016; Carlsson et al., 2019). Köhle (2002) developed a multivariate linear regres- sion formula using the measurable empirical mean furnace values, such as the total charged scrap, alloys, slag formers, tapping steel temperature, power on/off time, consumed burner natural gas and oxygen injection, and added oxygen for post-combustion. Sandberg et al. (2007) developed a method to estimate the impurity (i.e., Cu, Sn, As) and alloy content (i.e., Cr, Ni, Mo) in scrap grades using partial least square regression. Gajic et al. (2016) developed a multilayer perceptron model using an artificial neural network which estimates the electricity demand based on the input material composition. Because each process in iron and steel production is closely related, one change in a single process can result in an unexpected outcome. erefore, the mathematical modeling of each process is crucial for an overall decision- Received on May 8, 2020; accepted on June 16, 2020 DOI: 10.1252/jcej.20we077 Correspondence concerning this article should be addressed to S.-Y. Lee (E-mail address: suhyoung@postech.ac.kr). Research Paper