Sustainable Flue Gas Treatment System Assessment for Iron and Steel Sector: Spherical Fuzzy MCDM-Based Innovative Multistage Approach

,


Introduction
As a result of the promises that nations have made under the Paris Agreement, the 2030 Agenda of the United Nations, and the Sustainable Development Goals (SDGs), study into the difficulties that are associated with the environment and climate change has become an attractive area [1]. The global green agreement was suggested to deal with the aforementioned issues by resuming the global commitments to converting the world into a wealthy and equitable society with a competitive and contemporary economy that is concerned with sustainability. The global green deal is aimed at addressing the challenges that have been described [2]. To this purpose, there is a need for a policy reformation regarding the supply of green energy across all industries [3], including the economy, production, consumption, transportation, large-scale infrastructure, agriculture, and food.
The steel industry is seen as an essential key symbol of the nation's total might, and it is an important core industrial sector that contributes significantly to the national economy. Steel is in high demand across a variety of Egyptian sectors, including construction, transportation, and consumer good manufacturing, as the country's economy continues to expand and its citizens enjoy higher standards of living. The rapid increase in demand for steel has encouraged several steel firms to make investments in the construction of new steel manufacturing facilities in Egypt. Egypt's iron production rose by the end of 2021 to its highest level in 10 years, as production rates reached 10.3 million tons of iron, according to data issued by the World Steel Organization [4]. The report indicated that Egypt ranked 20 th in the list of iron-producing countries globally, the first in Africa, and the third in the Middle East, after its production jumped 25.1% last year [4]. Egypt's iron production in 2020 amounted to approximately 8.2 million tons [4]. Figure 1 shows the development of iron production in Egypt from 2011 to 2021.
The iron and steel sector is a significant component of the foundation of the national economy. Despite this, the issue of emissions is a major problem facing this industry [5]. The steel industry produces millions of tons of nitrogen oxides, sulfur dioxide, and particles [5]. It can be said that the emissions of air pollutants produced by the iron and steel industry have surpassed those formed by the thermal power industry, making the iron and steel industry the greatest source of emanations produced by Egypt's industrial sector. In view of this, Egyptian government reports in recent years suggest the need to promote a transition to low emissions in the iron and steel sector. The Ministry of Environment and a number of other ministries dispensed joint opinions on supporting the application of low emissions in the iron and steel industry, which explicitly propose lowemission restrictions for a variety of pollutants in the iron and steel industry [6]. Due to augmented government regulation and public consciousness of environmental conservation, industries that wish to remain competitive in the global marketplace cannot bear to ignore environmental apprehensions. Businesses must voluntarily devise strategies to decrease the environmental influence of their products. Because of increasing consciousness and comprehension of sustainability and government initiatives in this field, industries can no longer bear to disregard sustainability in their operations [7].
Air pollutants in the iron and steel industry originate from the blast oven, coke kiln, sintering, pellet, and iron alloy furnace processes, among others [8]. The blast furnace's primary materials are sinter, pellets, and bulk ore. The emission of sulfur dioxide and nitrogen oxides per ton of pellets is 30-50% less than that of sinter; therefore, growing the ratio of pellets is a crucial method for reducing the emission of air pollutants [9]. The coke oven is heated by lean gas, also known as coke oven gas, which has low levels of sulfur dioxide and high levels of nitrogen oxides in the flue gas. Because of the leaking of coke oven gas, the flue gas from the coke oven includes many pollutants, including hydrogen sulfide, hydrogen cyanide, and ammonia [5]. In the coke oven business, one of the most prevalent challenges for pollution management is addressing the high concentration of nitrogen oxides as well as the presence of other complex pollutants. The production of iron and steel is a significant contributor to the release of carbon dioxide into the atmosphere [10]. A significant portion of the manufacturing of steel is devoted to the sintering process. Flue gas from sintering has a complicated composition, a broad temperature range, a high oxygen and moisture content, and a wide range of flue gas temperatures [11]. It is responsible for a significant portion of emissions. In order to ensure the continued and sustainable growth of the iron and steel industry in Egypt, treatment and purification are essential challenges that need to be addressed.
The present desulfurization, denitrification, dedusting, and other pollutant control units are connected in a straightforward sequence [8]. This connection is the primary factor in the success of the sintering flue gas pollutant control. Many pollution control units have varying degrees of success in terms of their effects on the environment, their economies, and their technologies. As a result, the performance of low-emission technology for sintering flue gas will vary in terms of its impact on the environment, the economy, and the technology. Steel manufacturers have been able to effectively modernize some established low-emission methods of sintering flue gas thanks to the ongoing research conducted by steel companies [12]. Even though significant progress has been made with low-emission retrofits, the process is still in the very important stage of promotion and implementation. Steel manufacturers have a pressing challenge at this decisive stage, determining how to choose the technology that will be most effective. A complete analysis of the system's performance in several different aspects is used to guide the selection of the most effective technique or approach. As a result, the development of a multiobjective decision assessment system for the low-emission technique of sintering flue gas is an absolute need.
Sustainable development is well-defined as "development that encounters present necessities without conceding future generations' capability to encounter their own requirements." This connotation integrated the crucial intertemporal aspect of human influence on the normal ecosystem [13]. The term "sustainability evaluation" refers to a kind of multidimensional assessment approach that is often used to determine the overall performance of systems with regard to their technical, economic, social, and environmental elements [14]. Wang et al. [15] developed a sustainabilityfocused assessment approach to identify the optimal sites for wind plants, including technical, economic, environmental, and sociopolitical aspects. Abdel-Basset et al. [16] evaluated the sustainability of appropriate geographical sites for photovoltaic farms through technical, economic, environmental, and sociopolitical dimensions. Sustainability has been prioritized in the selection of low-emission technologies due to society's commitment to a sustainable future. Specialists must institute significant measures for determining lowemission technologies in the environmental, technological, 2 International Journal of Energy Research and economic areas from a sustainability standpoint. To evaluate low-emission technologies, decision-makers must manage a large number of sustainability indicators, and careful clustering is required to categorize indicators into each sustainability class. In light of the steel sector's significance as an infrastructure industry for national economies, the primary goal of this study is to determine the sustainable lowemission technology for the steel industry using multicriteria decision-making (MCDM) methods. With time, the MCDM has garnered the interest of the scientific community [17,18]. Methods of MCDM have been extensively implemented in numerous domains [17,18]. It has developed novel methodologies to aid specialists in evaluating and choosing the optimum alternative among competing qualitative and quantitative considerations. However, it is rarely used in evaluating sustainable flue gas treatment systems for the iron and steel sector. The problems posed by MCDM are characterized by a lack of accuracy and an abundance of ambiguity. The fuzzy MCDM approach makes use of fuzzy numbers to manage and assess imprecision and ambiguity [19]. Due to the fact that humans think in an instinctual manner, there are no precise solutions to the decision-making issues that occur in real life. Various strategies and ideas have been created to address this issue as a direct result of the necessity to compact with ambiguity in the context of real-world challenges [19]. The assessment of flue gas treatment systems requires various conflicting indicators, including operating cost, ease of implementation, and climate change potential, all of which must be stable using MCDM methods. Therefore, MCDM is an appropriate tool to assist decision-makers to assess and select the most appropriate and sustainable flue gas treatment systems. In addition, a combination of various evaluation techniques helps researchers obtain more accurate findings, which are closer to actual values.
The uncertainty model is determined by the amount of data that is available. In the traditional MCDM approach, decision-makers provide their thoughts in the form of a crisp value. When hazy and unclear information is included in the process of decision-making, these crisp values are sometimes inadequate to handle the challenges that arise while making decisions in the real world. Fuzzy set theory has been utilized so that ambiguity and complexity may be dealt with [20]. The most recent expansions of spherical fuzzy sets (SFSs) are the membership, nonmembership, and hesitation degrees [21]. The purpose of these additions is to expand the domain area for the assignment of doubtful degrees such as membership and hesitancy. If given the opportunity to choose the level of power, we would go with the spherical fuzzy set. This research will offer a spherical fuzzy (SF) MCDM technique with the goal of improving the efficiency with which vague and imprecise information is managed.
To the best of the authors' knowledge, very little literature has been performed to evaluate flue gas treatment systems in the iron and steel sector in generic, especially by applying the fuzzy MCDM approaches. This study presents a SF-MCDM approach that considers uncertainty in decision-making by applying spherical fuzzy numbers (SFNs). The suggested methodology adopts two techniques of decision-making, which are the CRiteria Importance through Intercreteria Correlation (CRITIC) method [22] and the COmbinative Distance-based ASsessment (CODAS) method [23]. They are implemented under a spherical fuzzy environment. The SF-CRITIC method is applied to evaluate the three main aspects of sustainability and the twelve subindicators that have an impact on the evaluation and selection of the best sustainable flue gas treatment system in the iron and steel sector. Furthermore, the SF-CODAS is used to rank the substitutions selected for the study.
All in all, the primary contributions of this study are outlined below.
(i) To suggest a CRITIC-CODAS approach of the determined challenges based on a spherical fuzzy environment to cope with the unpredictability inherent in decision-making, this is the first study to develop a MCDM approach consisting of CRITIC and CODAS methods for evaluating flue gas treatment systems in the iron and steel sector  3 International Journal of Energy Research (ii) The proposed approach that was described may be used to get dependable answers in situations when there is a lot of uncertainty (iii) The organizational repercussions of this study have the potential to provide important guidance to the iron and steel industry assessment and sustainability sector as well as to decision-makers and investors in other industries (iv) A comparative and sensitivity analysis is applied to demonstrate the reliability, stability, and validity of the developed approach The remainder of this study is structured as follows. In Section 2, a literature survey on the iron and steel industry, sustainability, CRITIC method, and CODAS method is presented. Section 3 discusses the assessment aspects and indicators of flue gas treatment systems in the iron and steel sector. Section 4 gives the preliminaries for SFSs and develops the SF-CRITIC-CODAS approach. The application of the suggested approach to flue gas treatment systems is illustrated in Section 5. Section 6 introduces some concluding remarks and future directions.

Literature Review
In this section, some literatures associated with the field of this study are presented. This section was divided into three parts. The first part reviews the literature related to the iron and steel industry and reducing emissions in this industry. The second part reviews the sustainability literature, processes, and implementation in industries. The third and final part reviews the literature on the CRITIC and CODAS methods.
2.1. The Iron and Steel Industry. In this part, some studies related to the iron and steel industry, and its technologies are presented. Liu et al. [12] developed a study for creating a life cycle sustainability assessment approach of four lowemission technologies. Their approach adopted a MCDM tool for selecting the optimal treatment technology for sintering flue gas. Their approach takes into account three dimensions of sustainability, namely, environmental, economic, and technological. Nguyen et al. [7] presented a paper for determining the sustainable supplier for the steel manufacturing industry. Their approach adopted three MCDM methods, namely, data envelopment analysis (DEA), SF analytic hierarchy process (SF-AHP), and SFweighted aggregated sum product assessment (SF-WAS-PAS). The SF-AHP method is applied to evaluate the criteria and determine their weights, while the SF-WASPAS method is used to evaluate and rank the selected suppliers. Zhang et al. [24] introduced research using numerical methods into the process of sintering iron ore using a combination of flue gas recirculation sintering and fuel-stacked distribution sintering. Zhu et al. [5] presented a study on technology that can regulate many processes and pollutants simultaneously, with the goal of achieving very low emissions in the iron and steel sector. Hadjela et al. [25] introduced a paper on the optimization of low alloy steel straight turning by the use of MCDM techniques in conjunction with the Taguchi method. Liu et al. [26,27] introduced research on China's iron and steel industry's technological road plan towards the most effective decarbonization development possible. Cui et al. [28] presented a paper on an analysis of the life cycle impacts of low-temperature treatment for sintering flue gas emissions in the steel sector. Liu et al. [26,27] introduced a paper on emissions of air pollutants caused by the sintering process used in China's iron and steel sector, as well as the possibility of their reduction. Based on these studies and other literature, it can be said that the field of the iron and steel industry and the reduction of emissions resulting from it are attractive to researchers and academics. According to studies, the main purpose, which most studies seek, is to reduce emissions in this industry.

Sustainability in Industries.
In this part, some studies related to the application of sustainability in industries are presented. Environmental sustainability seeks to enhance human welfare by safeguarding natural capital, such as the atmosphere or soil, and concentrating on reducing carbon footprints, packaging pollution, water use, and their overall impact on the environment [29,30]. For industrial decarbonization and the accomplishment of the net zero goal, it is essential to make strides toward reducing carbon emissions and offsetting carbon [31]. Also, sustainability includes several aspects such as social, economic, technological, and environmental sustainability [32]. Romero and Linares [33] proposed a framework for analyzing the conception of the operational aspects of energy sustainability. Their approach adopted a MCDM tool as the main framework. Hezam et al. [34] introduced a combined decision support approach for the sustainability evaluation of bioenergy production methods. Their approach adopted two MCDM methods: the Stepwise Weight Assessment Ratio Analysis (SWARA) for determining criterion weights and the complex proportional assessment (COPRAS) for ranking bioenergy production methods. Their approach takes into account sustainability factors including environmental, social, economic, and technological aspects. According to Candan and Cengiz Toklu [35], their approach takes into account several aspects of sustainability, including the environmental, economic, and social aspects. Alao et al. [36] developed a paper for selecting a suitable waste-to-energy method under uncertainty. Their approach adopted an MCDM tool that combines two MCDM methods: the fuzzy AHP and the fuzzy Multiple Objective Optimization on the basis of Ratio Analysis plus Full Multiplicative Form (MULTI-MOORA). They took into account four aspects of sustainability, which are the environmental, social, economic, and technical aspects. International Journal of Energy Research methodology that includes one of its methods, the CRITIC technique. Their approach was employed to assess the obstacles preventing the full implementation of Industry 4.0 to achieve sustainable digital transformation. Specifically, Fermatean fuzzy CRITIC was applied to evaluate and determine the weights of the criteria. Haktanır and Kahraman [37] introduced a hybrid MCDM framework that combines two MCDM methods, namely, the CRITIC and the REGIME. Their approach has been applied in determining suitable wearable health technology. Their approach has been implemented under a picture fuzzy set environment. In particular, picture fuzzy CRITIC has been applied for evaluating and determining the criterion weights. Ali [38] developed a hybrid MCDM methodology that integrates two MCDM techniques, namely, the CRITIC and the Measurement of Alternatives and Ranking according to the Compromise Solution (MARCOS). His approach has been utilized for selecting the optimal smartphone under SFSs. In this regard, the SF-CRITIC method has been used for assessing and determining the criterion weights. Also, the CODAS technique has been utilized in several literatures. Raheja et al. [39] developed a paper for conducting a simulating analysis for the evaluation of the worst polluted cities. They employed an MCDM methodology comprising of the AHP and CODAS techniques. They applied the CODAS method for ranking the selected cities. Jafarzadeh Ghoushchi et al. [40] presented a study for assessing the barriers to clean energy development. Their approach adopted an MCDM tool that combines two MCDM methods, namely, the Stepwise Weight Assessment Ratio Analysis (SWARA) and the CODAS. In particular, they applied the SF-CODAS in ranking the alternatives determined. Their approach has been conducted in the SF environment.

CRITIC and CODAS
To summarize, a hybrid approach consisting of two MCDM methods, the CRITIC and the CODAS, is presented. The hybrid approach is conducted under a spherical fuzzy environment. The SF-CRITIC is used to determine the weights of the three main aspects of sustainability and their twelve subindicators. Furthermore, the SF-CODAS is applied to evaluate and rank selected flue gas treatment systems in the iron and steel sector.

Sustainable Assessing Aspects
This section presents the main aspects of sustainable assessment and subindicators for each main sustainable aspect. Sustainable flue gas treatment systems for the iron and steel sector are evaluated through three sets of sustainability indicators: the economic aspect (ECA 1 ), the environmental aspect (ENA 2 ), and the technological aspect (TEA 3 ) [12,28]. The economic aspect is divided into two subindicators, capital cost (CAA 1_1 ) and operating cost (OPA 1_2 ). The environmental aspect includes five subindicators, which are climate change potential (CCP 2_1 ), ecological impact (ECM 2_2 ), toxicological risk (TRA 2_3 ), land occupation (LAO 2_4 ), and resource exhaustion (REE 2_5 ). The technological aspect contains five subindicators, namely, technical maturity (TEM 3_1 ), reliability and flexibility (REF 3_2 ), waste utilization potential (WUP 3_3 ), system operation and maintenance difficulty (SOM 3_4 ), and pollutant removal efficiency (PRE 3_5 ). The main aspects and their subindicators were selected by reviewing the literature and opinions of experts. These aspects have been approved by experts and those with experience in this field. These aspects and their indicators are applied to determine the most appropriate and sustainable flue gas treatment system for the iron and steel sector. Figure 2 presents aspects of the problem, which are the main objective, evaluation aspects, their subindicators, and the selected alternatives. The main aspects and their subindicators are mentioned as follows.
3.1. Economic Aspect (ECA 1 ). The economic aspect discusses the fundamental influences affecting the selection of a flue gas treatment system, such as life cycle costs, which vary from system to system [12]. Also, this aspect refers to operating costs, budget availability for maintenance work, and the extent of support provided to these systems. The economic aspect plays an important role in evaluating flue gas treatment systems in the iron and steel sector. Two subindicators are included in this aspect as follows.
3.1.1. Capital Cost (CAA 1_1 ). This indicator refers to the total costs that are spent to establish a flue gas treatment system, including the cost of equipment, land, and installation, i.e. from the planning stage to the operation stage.

Operating Cost (OPA 1_2
). This indicator refers to the costs required for start-up and implementation, including raw materials, labor costs, and energy consumption. It also indicates the budget spent on equipment maintenance and replacement parts for systems used in flue gas treatment.

Environmental Aspect (ENA 2 )
. This aspect refers to the factors that must be taken into account when evaluating the environmental impact of the life cycle of the used systems [41]. Environmental subindicators were developed to measure the impact of choosing flue gas treatment systems on the environment. This aspect includes five subindicators, namely, climate change potential, ecological impact, toxicological risk, land occupation, and resource exhaustion.

Climate Change Potential (CCP 2_1
). This indicator refers to climate changes as a result of emissions and pollutants resulting from the life cycle of operating flue gas treatment systems. It is worth noting that choosing a sustainable system will help reduce emissions by a large percentage, which may be present in the processes of energy production and raw materials and their processing.

Ecological Impact (ECM 2_2
). This indicator refers to environmental problems and pollutants that occur as a result of activities related to flue gas treatment systems, including atmospheric pollution and ozone depletion. During the construction and operation stages, dangerous materials such as lubricants, paints, cleansing materials, and other chemicals may be released in large quantities, which may have an effect on the land and groundwater [42]. 5 International Journal of Energy Research 3.2.3. Toxicological Risk (TRA 2_3 ). This indicator indicates that the toxic substances found in chemicals are present in the environment as well as the food chain, posing a risk to the health of aquatic creatures as well as to human health.

Land Occupation (LAO 2_4
). This indicator refers to the changes that occur as a result of land use policies, whether agricultural or urban, and the shift in land use policies. It refers to the availability of allocated lands without the existence of complex procedures for their use in establishing sustainable systems. Accordingly, nonarable land should be used and away from urban areas.

Resource Exhaustion (REE 2_5
). This indicator refers to the gradual depletion and depletion of resources that serve human life. Resources are in several forms, including mineral and water resources. This indicator expresses the importance of developing resource use laws that are designed to slow uncontrolled resource extraction and to create a more sustainable resource base for future generations of resource use.

Technological Aspect (TEA 3 )
. This aspect is considered the most important and one of the main aspects for evaluating flue gas treatment systems in the iron and steel sector, as there are considerations about the technological aspect of the systems used [12]. Also, a set of subindicators was selected to measure the technological aspect and the extent to which sustainability was achieved during the operation process. This aspect includes five subindicators, namely, technical maturity, reliability and flexibility, waste utilization potential, system operation and maintenance difficulty, and pollutant removal efficiency.

Technical Maturity (TEM 3_1
). This indicator refers to the level of technological maturity and reliance on straightforward and locally applicable systems, related to industrial capabilities, such as equipment, spare parts, specifications, and technical standards, in addition to developing related Selection the best low emission system for flue gas treatment in the iron and steel sector.  International Journal of Energy Research technical capabilities. Additionally, it shows whether the technology has already achieved the maximum level of theoretical effectiveness or whether there is still space for development.

Reliability and Flexibility (REF 3_2 )
. This indicator refers to taking into account the likelihood of an incident happening during the actual procedure. The indicator makes it possible to perform a comprehensive analysis of the probabilities of different technological emergencies, as well as an analysis of the potential causes of incidents and the risks that could be posed by incidents. Additionally, this refers to the capacity of the system to bounce back and make adjustments in the face of danger. The indicator investigates whether or not the system is able to fix itself in the event that something goes wrong.

Waste Utilization Potential (WUP 3_3
). This indicator calculates the potential economic advantage that could be derived from recycling the refuse materials that are produced by the technologies. During the desulfurization and dedusting processes, each of the technologies will generate waste resources. These waste resources have the potential to be used as infill material in the subgrade or marketed for use in industrial construction materials.

System Operation and Maintenance Difficulty (SOM 3_4 ).
This indicator refers to the degree of difficulty involved in the operation, and maintenance of low-emission technologies varies according to the degree of complexity of the system implementations and the general retrofits of the technologies. Additionally, it refers to the budget that is depleted in the process of conserving the necessary machinery, equipment, and extra accoutrements.

Pollutant Removal Efficiency (PRE 3_5
). This indicator refers to the effectiveness with which the technologies can remove particulate matter (PM), sulfur dioxide (SO 2 ), and nitrogen oxides (NO x ) from the air. It is the methodological parameter that is utilized the most frequently in order to evaluate the various options available for representing the contaminant elimination potential.

Materials and Methods
In this section, the methodology recommended for this study for the assessment of flue gas treatment systems in the iron and steel sector is presented. This section is split into two parts. The first part introduces some concepts and preliminaries related to SFSs. The second part offers the steps of the proposed hybrid methodology CRITIC-CODAS for evaluating the key aspects and their subindicators, determining their weights, and arranging the selected alternatives based on SFS theory.

Preliminaries.
In this part, some main definitions, concepts, and operations associated with spherical fuzzy sets are given. SFSs are generalizations of Pythagorean fuzzy sets (PFSs), intuitionistic fuzzy sets (IFSs), and neutrosophic sets (NSs) that give a broader range to professionals. Consequently, the definition of SFSs is mentioned as follows.
Definition 1. A fuzzy set P, defined in reference Y, is in the form of the following equation [21]: where the function μ P ::Y ⟶ ½0, 1 designates the degree of membership of an item to the sets P, ʋ P ::Y ⟶ ½0, 1 designates the degree of nonmembership of an item to the sets P, and π P : Y ⟶ ½0, 1 designates the degree of hesitant of an item to the sets P, with the condition in the following equation: Definition 2 (see [40]). Let P 1 = ðμ P 1 , ʋ P 1 , π P 1 Þ and P 2 = ðμ P 2 , ʋ P 2 , π P 2 Þ be the two SFS numbers; q is a fixed number greater than zero. The mathematical procedures of these two SFS numbers are implemented as follows: Definition 3 (see [43]). Let P 1 = ðμ P 1 , ʋ P 1 , π P 1 Þ and P 2 = ðμ P 2 , ʋ P 2 , π P 2 Þ be the two SFS numbers. The following essentials with the conditionq , q 1 , and q 2 > 0, on these spherical fuzzy numbers according to the Equations (7), ( (8)), ( (9)), ((10)), ((11)), and ( (12)), are done.

International Journal of Energy Research
Definition 4 (see [44]). Let P = ðμ P , ʋ P , π P Þ characterize an SFS number. The score value (S j ) and the accuracy value (A j ) of the number P are computed as follows: Negative or zero values may be obtained if the answers acquired based on the score and accuracy are incorrect. And the performance may have the same level of precision. Consequently, prioritization function (PF) is taken into account for SFS numbers according to Definition 5 (see [21]). The spherical weighted arithmetic mean (SWAM) operator with regard to w = ðw 1 , w 2 , ⋯, w n Þ with w i ∊½0, 1 and ∑ n i=1 w i = 1 as an Equation (17) is computed.
4.2. Evaluation Model SF CRITIC-CODAS. In this part, a proposed integrated hybrid approach is presented to evaluate and select the most suitable flue gas treatment systems in the iron and steel sector by combining two MCDM methods, CRITIC-CODAS. The applied hybrid approach is conducted under spherical fuzzy environment and using spherical fuzzy numbers (SFN). The suggested approach comprises three main stages. The first stage is related to exploring the issue and its aspects and deter-mining the participating experts. The selection of the key aspects and their subindicators affects the choice of the most appropriate systems chosen in the evaluation process. The second stage refers to the steps of evaluating the main aspects and subindicators using the SF-CRITIC technique. The third stage shows the steps of the SF-CODAS technique for evaluating and classifying the selected systems in the evaluation process. The steps of the applied framework can be seen in Figure 3. Here, the illustrative steps of the hybrid framework proposed for this study are mentioned.

Stage 1: Data Collection
Step 1. The problem is studied, and its details are determined with great accuracy. The main aspects affecting the solution of the problem and subindicators are identified. Some main standards are used to select the experts participating in the study, as shown in Table 1. Due to the paucity of pertinent literature on the subject in the context of the present study, the employed methodology must rely on the opinions of qualified specialists with the necessary knowledge and experience in the field [46]. In addition, considering the numerous identified factors and the time-consuming nature of MCDM methods, the chosen data acquisition and analysis strategy must be effective and efficient.
Step 2. The main aspects of sustainability and its subindicators are determined through an analysis of the relevant literature and previous studies, in addition to the opinions of the participating experts. C j = ðC 1 , C 2 , ⋯, C n Þ, with j = 1, 2, ⋯, n. Let w = ðw 1 , w 2 , ⋯, w n Þ be the vector set applied for defining the main aspects and their indicators weights, where w j > 0 and ∑ n j=1 w j = 1. Finally, the final list of available alternatives for use in the study is determined. The set A i = fA 1 , A 2 , ⋯, A m g, having i = 1, 2, ⋯, m substitutes, is assessed by n decision indicators of set C j = fC 1 , C 2 , ⋯, C n g, with j = 1, 2, ⋯, n.
Step 3. A set of linguistic variables and their equivalent SFNs is determined as shown in Table 2. These variables are applied by the experts involved in the process of evaluating and determining the weights of either the main sustainable aspects or the sustainable subindicators. Also, these variables are applied in evaluating and ranking the selected alternatives.
Step 4. Create the assessment matrix. The assessment matrix includes the alternatives and main aspects and their indicators according to Equation (19) by all experts to express their preferences for these indicators. Let experts =ðexpert 1 , expert 2 , ⋯, expert t Þ be a set of experts who expressed their evaluation report for each alternative A i ði = 1, 2 ⋯ mÞ against their indicators C j ðj = 1, 2 ⋯ nÞ. Also, the assessment matrix is constructed using the linguistic terms presented in Table 2 and then using SFNs presented in Table 2 according to the following equation: Combine the initial spherical fuzzy assessment matrix.
Remove the fuzziness of the evaluation matrix.
Determine the objective weights by the CRITIC method.
Stage 3: Spherical fuzzy CODAS method Calculate the normalized decision matrix.
Calculate the weighted normalized decision matrix.
Determine the negative ideal solution.

Compute the Euclidean and Hamming distances
Compute the relative evaluation matrix and the overall score of each alternative.
Determine the optimal substitute.
Discuss the sensitivity and comparative analysis. Figure 3: Research framework.
Step 5. Combine the assessments on main aspects and their indicators. The single expert assessments are combined by applying SWAM given in Equation (17). Here, w q introduces the importance of the q th expert.
Step 6. Convert the SFNs to crisp value by applying the PF according to Equation (16). Let Z = ½z ij m×n be the combined evaluation matrix as a score matrix in Step 1. Compute the normalized decision matrix for main aspects and their indicators according to Equation (21) and Equation (22) [22]. For positive (benefit) indicators of the evaluation matrix, For negative (cost) indicators of the evaluation matrix, Step 2. Compute the correlation coefficient ρ jk between main aspects and their indicators according to where x j and x k are the means of j th and k th indicators. x j is computed according to Equation (24) and likewise for x k .
Step 3. Calculate the standard deviation (STDEV) σ j for each main aspect and their indicators according to Step 4. Determine the amount of information for main aspects and their indicators according to  Step 5. Calculate the weights of main aspects and their indicators according to Step 6. Finally, the same steps for defining the weights of the key aspects are employed for the subindicators for each main aspect. Then, the global weights of the subindicators are determined by multiplying the weights of the main aspects by the local weights of the subindicators.

Stage 3: Spherical Fuzzy CODAS Method
Step 1. Here, the same assessment matrix of alternatives with subindicators ½x ij m×n is used as in Equations (19) and (20). Based on that, compute the normalized assessment matrix according to Equation (28) for benefit indicators BE and for cost indicators CO.
Step 2. Obtain the weighted normalized assessment matrix according to Equation (29). Here, ðw j Þ 1×n presents weight of j th indicator.
Step 3. Identify the negative ideal solution NS j for each indicator according to Step 4. Compute the Euclidean and taxicab distances of alternatives from negative ideal solution by applying the following equations: Step 5. Construct the relative assessment matrix ½h is n×n according to where s ∈ f1, 2 ⋯ mg and γ indicates a threshold function to identify the equality of the Euclidean distances of two alternatives.
Step 6. Calculate the assessment score of each alternative according to Equation (34). Finally, the alternatives are ranked from best to worst by descending order of their F i ratings.

Application
The main purpose of this section is to apply the proposed hybrid methodology steps. The section is separated into three chief parts. The first part offers an actual case study of the implementation of the recommended approach. The second part applies the steps of the proposed SF-CRITIC-CODAS approach. The third and final part discusses the results of the study.

Case Study.
Currently, the iron and steel industry represents an actual aspect and a basic pillar in our lives because of its direct impact on development. The global production of crude steel has more than doubled in the last two decades, as production increased to 1.86 billion tons in 2020 [47]. Despite the decrease in the energy needs required to produce each ton of steel during the past two decades, growth in production rates is due to the increase in demand globally, which led to a steady growth in the total energy ingested by this vital industry. Consequently, the percentage of greenhouse gas releases grows. Consequently, an increase in carbon dioxide emissions and greenhouse gases, in general, prompted many iron and steel manufacturers around the world to research and develop and experiment with new technologies in manufacturing processes, to face these challenges, by applying the principles of sustainability. The global iron and steel industry bears a great responsibility. This is so that it can be in line with the goals of the Paris Cli-mate Agreement to contain global warming below two degrees Celsius; this industry must reduce its carbon emissions by more than half by 2050, while continuing to reduce emissions until reaching zero emissions. In this regard, steelmakers have reduced energy use in recent years. Thus, carbon emissions have been reduced, through technological innovations are based on the development of production processes, but these efforts still need more and more to decrease the carbon footprint of this industry in the future. In this regard, the Egyptian government has implemented strict legal measures in order to protect the environment. In addition, many initiatives have been published, for example, the "Go Green" initiative in order to keep the environment and raise awareness of the importance of sustainability of environmental resources [48]. Egypt seeks to achieve the principles of sustainable industrial development, the advancement of the industrial sector, and the reduction of emissions, through sustainable environmental policies adopted in the field of iron and steel sector. The iron and steel industry is part of the existing industries in Egypt. Therefore, these industries were alerted to a sustainable transformation in the manufacturing process and to reduce the emission of toxic gases to preserve the environment. Thus, stakeholders seek

Application of the Hybrid Approach SF CRITIC-CODAS
Step 1. The problem was considered, and the necessary details were determined. The main objective of the research was determined to select the best sustainable flue gas treatment system in the iron and steel sector. Four experts were identified to participate in the study according to the rules set forth in Table 1.
Step 2. Three main aspects of sustainability were identified, namely, the economic, environmental, and technological aspects. These three main aspects were divided into 12 subindicators to include all the details needed to solve the problem. Subindicators are capital cost (CAA 1_1 ) and operating cost (OPA 1_2 ) in the economic aspect. The environmental aspect includes five subindicators, which are climate change potential (CCP 2_1 ), ecological impact (ECM 2_2 ), toxicological risk (TRA 2_3 ), land occupation (LAO 2_4 ), and resource exhaustion (REE 2_5 ). The technological aspect contains five subindicators, namely, technical maturity (TEM 3_1 ), reliability and flexibility (REF 3_2 ), waste utilization potential (WUP 3_3 ), system operation and maintenance difficulty (SOM 3_4 ), and pollutant removal efficiency (PRE 3_5 ).
Step 3. Nine linguistic terms and their equivalent SFNs are identified as shown in Table 2. For example, the term extremely low significance is represented by 〈0.10, 0.90, 0.10〉. These terms are used to assess the core aspects and their subindicators and to define the weight for each indicator. In addition, the four alternatives selected for the study are evaluated. The alternatives used are System 1 , System 2 , System 3 , and System 4 .
Step 4. The assessment matrix has been created between the four systems and the main aspects according to Equation (19) by the four experts to express their opinions for these aspects using linguistic terms in Table 1, as presented in Table 3. Similarly, the assessment matrix has been created according to Equation (19) Table 2, as offered in Table 4.
Step 5. The assessment matrices of the main aspects by the four experts have been gathered by applying SWAM operator given in Equation (17), as presented in Table 5. Also, assessments are collected using expert weights given in Table 2.
Step 6. The SFNs of the main aspect assessment matrix have been converted to crisp values by applying the PF in Equation (16), as a score matrix in Equation (20).
Step 7. The normalized decision matrix of the main aspects has been obtained by applying Equation (21) as presented in Table 6.
Step 8. The correlation coefficient between the main aspects has been calculated by applying Equation (23) as offered in Table 7.
Step 9. The standard deviation for each main aspect has been calculated by applying Equation (25) as offered in Table 7.
Step 10. The amount of information for the main aspects has been determined according to Equation (26) as exhibited in Table 7.
Step 11. The final weights of the main aspects have been calculated by applying Equation (27) as offered in Table 7 and Figure 4.
Step 12. The assessment matrix has been created between the four systems and the economic aspect's subindicators according to Equation (19) by the four experts to express their opinions for these indicators using linguistic terms in Table 1, as presented in Table 8. Similarly, the assessment matrix has been created according to Equation (19) by the four experts to express their opinions on these indicators using SFNs in Table 2, as presented in Table 9.     Step 13. The assessment matrices of the economic aspect's subindicators by the four experts have been combined by applying SWAM operator given in Equation (17), as presented in Table 10. Also, assessments are collected using expert weights presented in Table 2.
Step 14. The SFNs of the economic aspect's subindicator assessment matrix have been converted to crisp values by applying the PF in Equation (16), as a score matrix in Equation (20).
Step 15. The normalized decision matrix of the economic aspect's subindicators has been obtained by employing Equation (21) as presented in Table 11.
Step 16. The correlation coefficient between the economic aspect's subindicators has been calculated by applying Equation (23) as exhibited in Table 12.
Step 17. The standard deviation for each indicator of the economic aspect's subindicators has been calculated by applying Equation (25) as offered in Table 12.
Step 18. The amount of information for the economic aspect's subindicators has been determined according to Equation (26) as offered in Table 12.
Step 19. The final weights of the economic aspect's subindicators have been calculated by employing Equation (27) as offered in Table 12 and Figure 5.
Step 20. The assessment matrix has been created between the four systems and the environmental aspect's subindicators according to Equation (19) by the four experts to express    Table 1, as presented in Table 13. Similarly, the assessment matrix has been created according to Equation (19) by the four experts to express their opinions on these indicators using SFNs in Table 2, as offered in Table 14.
Step 21. The assessment matrices of the environmental aspect's subindicators by the four experts have been combined by applying SWAM operator given in Equation (17), as presented in Table 15. Also, assessments are collected using expert weights presented in Table 2.
Step 22. The SFNs of the environmental aspect's subindicator assessment matrix have been converted to crisp values by applying the PF in Equation (16), as a score matrix in Equation (20).
Step 23. The normalized decision matrix of the environmental aspect's subindicators has been obtained by employing Equation (21) as presented in Table 16.
Step 24. The correlation coefficient between the environmental aspect's subindicators has been calculated by applying Equation (23) as offered in Table 17.   Step 25. The standard deviation for each indicator of the environmental aspect's subindicators has been calculated by employing Equation (25) as offered in Table 17.
Step 26. The amount of information for the environmental aspect's subindicators has been determined according to Equation (26) as offered in Table 17.
Step 27. The final weights of the environmental aspect's subindicators have been calculated by utilizing Equation (27) as offered in Table 17 and Figure 6.
Step 28. The assessment matrix has been created between the four systems and the technological aspect's subindicators according to Equation (19) by the four experts to express their opinions for these indicators using linguistic terms in Table 1, as presented in Table 18. Similarly, the assessment matrix has been created according to Equation (19) by the four experts to express their opinions on these indicators using SFNs in Table 2, as offered in Table 19.
Step 29. The assessment matrices of the technological aspect's subindicators by the four experts have been combined by applying SWAM operator given in Equation (17), as presented in Table 20. Also, assessments are collected using expert weights presented in Table 2.
Step 30. The SFNs of the technological aspect's subindicator assessment matrix have been converted to crisp values by applying the PF in Equation (16), as a score matrix in Equation (20).
Step 31. The normalized decision matrix of the technological aspect's subindicators has been obtained by employing Equation (21) as presented in Table 21.   Step 32. The correlation coefficient between the technological aspect's subindicators has been calculated by applying Equation (23) as exhibited in Table 22.
Step 33. The standard deviation for each indicator of the technological aspect's subindicators has been calculated by employing Equation (25) as offered in Table 22.
Step 34. The amount of information for the technological aspect's subindicators has been determined according to Equation (26) as offered in Table 22.
Step 35. The final weights of the technological aspect's subindicators have been calculated by employing Equation (27) as exhibited in Table 22 and Figure 7.   Step 36. In the end, the global weights of the subindicators were obtained by multiplying the final weights of the main aspects by the weights of the local subindicators as shown in Table 23 and Figure 8.
Step 37. Here, the combined matrix of the four flue gas treatment systems is used with respect to all subindicators for all main aspects as presented in Table 24, for ranking the selected four systems by applying the CODAS method.
Step 38. The normalized assessment matrix has been calculated by employing Equation (28) as exhibited in Table 25.
Step 39. The weighted normalized assessment matrix has been obtained by utilizing Equation (29) Table 26. Then, the negative ideal solution has been determined by utilizing Equation (30) as presented in Table 26.
Step 40. The Euclidean and taxicab distances of four systems from negative ideal solution have been computed by applying Equations (31) and (32), respectively, as exhibited in Table 27.
Step 41. The relative assessment matrix has been constructed according to Equation (33) as presented in Table 28. The γ is set to be 0.2 according to the opinions of experts.
Step 42. The assessment score of each alternative (system) has been computed by utilizing Equation (34) as presented in Table 28. The four systems have been ranked according to the assessment score in Table 28, and the final ranking of the four systems is shown in Figure 9.
5.3. Finding Discussion. In this part, the results obtained from applying the proposed hybrid approach SF-CRITIC-CODAS are discussed. The results of the study are separated into twofold. The first fold deals with the findings of evaluating and defining the weights of the three main aspects and their twelve subindicators. The second fold includes the results of the arrangement of four flue gas treatment systems in the iron and steel sector used in the study.
Initially, the three main aspects of sustainability were assessed using SF-CRITIC method. The findings in Table 7and Figure 4 indicate that the environmental aspect is the most influential with a weight of 0.431, followed by the technological aspect with a weight of 0.326, while the economic aspect is the least influential with a weight of 0.244. According to the results of the main aspects of sustainability, the focus on the environmental aspect shows the importance of taking into account environmental indicators when establishing industries that are similar to the iron and steel industry. Therefore, the environmental aspect has a main role in selecting the appropriate systems for the iron and steel sector.
After that, the subindicators of the economic aspect were evaluated using SF-CRITIC method as displayed in Table 12 and Figure 5. The findings indicate that the capital cost indicator has the highest weight of 0.664, while the operating cost indicator has the lowest weight of 0.336.
Also, the subindicators of the environmental aspect were estimated using SF-CRITIC method as exhibited in Table 17 and Figure 6. The findings indicate that the ecological impact indicator has the greatest weight of 0.261, followed by the land occupation indicator with a weight of 0.208, while the resource exhaustion indicator has the least weight of 0.145.
Furthermore, the subindicators of the technological aspect were estimated using SF-CRITIC method as exhibited in Table 22 and Figure 7. The findings indicate that the waste utilization potential indicator has the greatest weight of 0.276, followed by the system operation and maintenance difficulty indicator with a weight of 0.254, while the pollutant removal efficiency indicator has the least weight of 0.130.    In this regard, the global weights were determined for all subindicators as shown in Table 23 and Figure 8. According to the findings, it was found that the capital cost indicator has the largest weight of 0.162, followed by the ecological impact indicator with a weight of 0.112, while the pollutant removal efficiency indicator has the least weight of 0.042.
Finally, the four systems selected for the study were evaluated and ranked using the SF-CODAS approach. The value of the γ parameter was set to 0.2 according to the opinions of the participating experts. Most of the literature uses this parameter γ as an initial value of 0.2. The evaluation and ranking findings in Table 28 and Figure 9 indicate that System 4 is the most sustainable and suitable system to be used in flue gas treatment in the iron and steel sector, followed by System 3 , while System 2 ranks last.

Sensitivity Analysis.
In this subsection, a sensitivity analysis is demonstrated on the obtained findings to prove their consistency and stability [39]. Some decision-making considerations are subjectively determined based on the insight into the issue by specialists and the magnitude of environmental hazards. Consequently, these parameters vary in accordance with the modelling context of the decisionmaking system. In this paper, the sensitivity analysis is divided into two parts. The first part revolves around the change in parameter γ in the CODAS method, which is set according to the personal opinions of the decision-makers or participating experts. The second part is related to making many changes to the weight of the environmental aspect.
According to the first part that is related to the change in parameter γ in the CODAS method, the weight of this parameter was changed in ten cases, from setting the value of the parameter γ to 0 to set the value of the parameter to 1. The value of the parameter is incremented by 0.1 each time. For example, in the first case, the parameter value is set to 0, and in the second case, the parameter value is set to 0.1. According to the results of the ten cases in Figure 10, it is clear that System 4 is the most suitable and suitable system for use in flue gas treatment in the iron and steel sector. In this regard, the results of the arrangement of the four systems are the same as the arrangement of the default case when the parameter γ value was 0.2.
The second part is related to changing the weight of the environmental side from 0.1 to 1. Accordingly, there are eleven cases of change in the weight of the environmental side as in Figure 11. Ten cases are assigned the weight of the environmental side in the first case with a value of 0.1 and the next case with a value of 0.2 and so on, until 1. In the case 11, the economic, environmental, and technological sides are given an equal weight of 0.333. The results indicate that System 4 is the best in all cases except when the weight of the environmental aspect becomes 0.9 and 1. Therefore, relying on only one aspect in the evaluation, which is the environmental aspect, System 3 is considered the most appropriate. Finally, all these changes, whether in weights or parameters used in the suggested hybrid approach, are to illustrate the reliability of the findings of the approach.

Comparative
Analysis. In this part, the problem of evaluating flue gas treatment systems in the iron and steel sector is solved by some MCDM approaches based on spherical fuzzy theory. Comparative analysis is achieved to confirm the validity, consistency, and effectiveness of the suggested hybrid approach. The proposed approach is compared with the SF-MARCOS [38] and SF-WASPAS [7] approaches to assess and rank the four systems identified in the study. Also, the weights of the main aspects and subindicators obtained using the SF-CRITIC method were adopted as final weights in the two methods used in the comparison as shown in Table 23. Thus, the arrangement of the four systems was carried out through the two methods used in the  Figure 12.
Accordingly, the results of the three approaches used in the comparison confirm that System 4 is the most appropriate and sustainable among the different systems used in the evaluation process, followed by System 3 . In contrast, the order of the least sustainable alternative differed. In the SF-WASPAS and SF-MARCOS approaches, System 1 was the least suitable and sustainable. While in the proposed SF-CRITIC-CODAS hybrid approach, System 2 is the worst or least sustainable of the available systems. According to the findings, it can be seen that the order of some alternatives has changed. Dissimilarities in the ordering of the substitutes can be attributed to the distinct scientific foundations of each framework.

Concluding Remarks
During the following years, the iron and steel sector witnessed steady growth commensurate with economic devel-opments, development plans, and the increase in the establishment of infrastructure in all facilities. Also, the iron and steel industry is one of the most energy-intensive industries in terms of energy consumption and the resulting carbon, as it represents 7% of global carbon emissions. By the end of 2020, the steel industry is at the top of the most difficult sectors to decarbonize, but the promotion of the role of clean hydrogen and many other new technologies may facilitate the task. According to the International Energy Agency, the direct carbon emissions of heavy industries in the Group of Seven major industrialized countries are about 6 billion tons annually, more than one-sixth of the global total. How to select the most sustainable low-emission method has consequently become a pressing matter that requires immediate attention.
In this research, a comprehensive methodology for the sustainability assessment of low-emission flue gas treatment systems in the iron and steel sector is presented. Four experts with experience in the field of steel manufacturing

22
International Journal of Energy Research and with academic experience were invited to participate in the study at the same time. Nine linguistic terms and their equivalent SFNs were identified for the participants to use in evaluating the main aspects and their subindicators and determining their weights. The three main aspects of sustainability used in evaluating flue gas treatment systems in the iron and steel sector are economic, environmental, and technological. The subindicators of these aforementioned aspects are capital cost, operating cost, climate change potential, ecological impact, toxicological risk, land occupation, resource exhaustion, technical maturity, reliability and flexibility, waste utilization potential, system operation and maintenance difficulty, and pollutant removal efficiency. A hybrid MCDM approach is applied under a SF environment. The approach includes two MCDM methods: the SF-CRITIC and the SF-CODAS. In the beginning, the SF-CRITIC technique is employed to assess the priorities of the three main aspects and the twelve subindicators and then to determine the global weights of the subindicators. Then, the SF-CODAS technique is employed to evaluate and rank the four systems selected for the study. The findings of the study designate that the environmental aspect is the most influential aspect in selecting the best sustainable flue gas treatment system in the iron and steel sector. Finally, sensitivity and comparative analyses were performed to prove the robustness and stability of the proposed hybrid approach. The results of the two analyses indicate that the ranking of the candidate methods is stable, with some minor changes in the ranking in a few scenarios. Spherical weighted geometric mean SWAM:

Abbreviations
Spherical weighted arithmetic mean STDEV: Standard deviation.

Data Availability
The data used to support the findings of this study are included within the article.

Conflicts of Interest
The authors declare that they have no conflicts of interest.