Towards Green and Low-Carbon Transformation via Optimized Polygeneration System: A Case Study of the Iron and Steel Industry

Abstract

Polygeneration systems have significant potential for energy conservation and emission reduction and can effectively promote green and low-carbon development in energy-intensive industries, such as the iron and steel industry. However, its application faces the difficulty in technology selection under multiple objectives simultaneously, which is to determine the technology portfolio to achieve the synergy of energy conservation goals and air pollutant emission reduction goals, as well as ensure the economic benefits of the enterprises. This study investigated a case polygeneration system where the iron and steel plant are the core with four polygeneration paths and twenty polygeneration technologies. A multi-objective optimization model is developed to select the optimal technology combination of each polygeneration path under energy conservation, emission reduction, and cost control objectives, which is solved by the non-dominated sorting genetic algorithm-II (NSGA-II). The optimal results can reach significant energy conservation and emission reduction effects while obtaining economic benefits. However, synergistic and conflicting relationships among the objectives exist in both scales of iron and steel plants. The final decision scheme can achieve the mitigations equivalent to 15.9–27.1% and 16.3–42.6% of the energy consumption and air pollutant emissions of the steel enterprises with annual production of 3 Mt/a and 9 Mt/a, respectively. There are thirteen and twelve technologies that are selected as the final decision scheme in the polygeneration system in these two case enterprises. These findings demonstrate the significant roles the polygeneration system plays and provide critical insights and methodology in the technical selection of the polygeneration system.

1. Introduction

Over the past few decades, energy conservation and air pollutant emission control have been key environmental management issues for the world [1,2,3,4]. As the important source of energy consumption and pollution emissions, the industrial sector has long been the concern for energy conservation and emission reduction, especially for the iron and steel industry [5,6,7,8]. Iron and steel production is accompanied by a large amount of energy consumption and pollutant emissions (including SO₂, NO_x, PM, and other pollutants), so how to achieve low-carbon and clean production in the iron and steel industry has always been a concern of the government, enterprises, and scholars. Among many energy conservation and pollutant emission reduction technologies, polygeneration systems have gained attention due to their significant energy, environmental, and economic benefits [9,10,11,12].

In general, a polygeneration system for the industrial sector consists of two or more industries, and in which waste or by-products produced by some industries can be consumed by other industries [13,14], thus their overall environmental and energy impact could be reduced, and in some cases, additional economic benefit will also be created. Therefore, polygeneration systems are an important way to promote circular economy and enhance the efficiency of resource utilization.

There is a large amount of waste or by-products that will be produced in the production process of iron and steel [15,16], such as steel slag, waste heat, and waste pressure. These wastes or by-products can be used in the chemical, building materials, and power industries. In other words, extensive material and energy utilization pathways exist between the iron and steel, chemical, building materials, and power industries. For instance, coke oven gas and blast furnace gas in the iron and steel industry can be used to manufacture chemical products, blast furnace slag can be used to prepare building materials, and the waste heat and pressure generated from sintering, coking, and blast furnaces can be used to generate electricity.

The contribution of this study is to establish a comprehensive technical selection framework for the polygeneration system based on the chain of “supplier—waste/by-products—technologies—adopter”, and the trade-off of energy, environmental, and economic benefits was considered, as well as the trade-off between various environmental targets, such as SO₂ reduction and NO_x reduction. A multi-objective optimization model was developed in this study to address the above issues, and the iron and steel industry was taken as an example for analysis. The results of this study can provide insights for energy conservation and emission reduction in the iron and steel, power, chemical, and building materials industries, and these methods can also be applied to other industrial sectors.

The remainder of this study is organized as follows: Section 2 is the literature review. Section 3 introduces the methods and materials, including the polygeneration system simulation, multi-objective optimization model, and data processing. The results and discussion are presented in Section 4, while Section 5 reports the main conclusions of this study and some suggestions for the promotion of polygeneration systems in the iron and steel, chemical, building materials, and power industries in China.

2. Literature Review

Compared to the energy conservation and emission reduction system within a single industry, the system based on the polygeneration process is more complex and uncertain [17,18,19]. As a result, many studies have focused on the simulation and performance evaluation of polygeneration systems. Among them, techno-economic evaluation is a typical evaluation criterion since the economic effect is the primary motivation to build the system. For example, Kim et al. [20] and Lee et al. [21] had conducted techno-economic evaluation in a polygeneration system from coke oven gas and steel-mill off-gases, while similar evaluations can be seen in the system based on the LNG-regasification [22,23], catalytic hydropyrolysis [24], and agricultural residue [25]. In addition, energy efficiency is widely regarded as a vital evaluation indicator, especially in the polygeneration system of energy [26,27]. Shi et al. [28] have compared a custom coal-fired power plant with air-blown IGCC polygeneration ones and found that there is a 6% increase in energy efficiency after adopting the polygeneration system. Soutullo et al. [29] have taken three Spanish cities as examples to investigate the energy performance of the power–heat polygeneration plant in different ways. Environmental indicators have been considered in studies via a series of methods such as emission coefficient [30], life-cycle assessment [31,32,33], and environmental cost analysis [34]. These methods have proved the advantages of polygeneration systems via different evaluation criteria. However, since the evaluation could only be conducted among limited scenarios, these studies have usually focused on the current status of the system and thus cannot support the decision-makers to select the technologies in the polygeneration system and choose the optimal schemes.

To overcome this shortcoming, many studies have focused on the optimization and design of the polygeneration and have adopted multiple optimal methods. Among them, mixed-integer programming is a common model, usually taking best economic performance as the objective. Liu et al. [35] have built a mixed-integer nonlinear programming model to search for an optimal economic profit in an electricity–methanol polygeneration plant. Urbanucci [36] has summarized the advantages and limitations of mixed-integer nonlinear programming in solving the problems of polygeneration system designs. Other research has also adopted this method to optimize biomass co-generating technology selection [37], the selection of the contracted power from the grid within different renewable sources [38], and the building of microgrid polygeneration systems [39]. Though these studies have supported the improvement of polygeneration systems, they have solved single objective optimization problems and neglected other energy and environmental objectives as well as the trade-off relationship between them. In this situation, the optimal results they have obtained may have biases from the comprehensive optimal ones.

Some other studies have avoided this problem and have considered two or more objectives simultaneously. Khanmohammadi and Atashkari et al. [40] have investigated the optimal material and energy flows in a biomass polygeneration system under maximum exergy efficiency and minimum cost. Yeşilkaya et al. [41] have focused on the forest industry and designed a bi-objective model to seek for the lowest economic cost and highest ecological effects. These studies have transferred multi-objectives into single ones through different methods, such as the monetary and weighted sum method, with the relationship between objectives subjectively defined. Another type of multi-objective solution method is the searching algorithm, which also has wide applications. For example, both Ahmadi et al. [42] and Liu et al. [43] have adopted evolutionary algorithms in energy polygeneration system designs by combining energy, exergy, and economic objectives.

Compared with single emission reduction technology, polygeneration systems have many advantages in energy efficiency, economic benefits, and coordinated emission reduction, especially in the iron and steel industry, which is an energy- and emission-intensive industry. Through the cascade of waste or by-product utilization technologies, a polygeneration system can be formed with the core of the iron and steel industry, including the chemical, building materials, and power industries. However, it should be emphasized that the technical system of the polygeneration system is quite complex, and its quantitative evaluation needs to be based on system simulation. Meanwhile, it is necessary to manage the polygeneration system from the perspective of multi-objective optimization because the system involves multiple objectives and there are synergetic and conflicting relationships among them. To sum up, the simulation and analysis of the entire polygeneration system will be carried out in this study in order to quantify the benefits of different technologies and determine the optimal technology path under multiple objectives and constraints, as well as put forward the finial optimal scheme.

3. Method and Material

3.1. Polygeneration System Simulation

In this section, the material and energy metabolism paths of steel, chemical, building materials, and power industries will be identified first, and then the simulation of polygeneration systems will be modeled based on the bottom-up modeling method. The modeling process follows the chain of “supplier—waste/by-products—technologies—adopter”, where the iron and steel industry are the suppliers in this study and the chemical, building materials, and power industries are the adopters. The production of wastes and by-products in the iron and steel industry is shown in Figure 1 [44,45].

It can be seen from Figure 1 that there are eight types of waste/by-products in the steel industry, which are coke sensible heat, sintering waste heat, blast furnace gas residual pressure, coke oven gas, blast furnace gas, desorbed gas, blast furnace slag, and converter slag, respectively. According to the utilization approaches of these waste/by-products and the matching and applicability issues of raw materials, processes, and technologies in the polygeneration system, 4 polygeneration paths are designed in this study, and a total of 20 technologies (No. T1–T20) are identified in this study to address these waste/by-products [46,47,48], as follows:

• Path 1: Coke waste heat/Coke oven gas/Desorbed gas → T1–T9 → Power/Chemical products.
• Path 2: Sintering waste heat → T10 and T11 → Power.
• Path 3: Blast furnace gas/Blast furnace residual pressure → T12–T15 → Power/Chemical products.
• Path 4: Blast furnace slag/Converter slag → T16–T20 → Building materials. The specific information was shown in Figure 2.

(a)

Path 1: Polygeneration system of iron and steel industry, chemical industry, and power industry based on coking process. In this path, coke waste heat and coke oven gas generated in the coking process would be used for power generation and chemical product production, including coal tar, sulfuric acid, methanol, ethanol, and methyl methacrylate (MMA).

(b)

Path 2: Polygeneration system of iron and steel industry and power industry based on sintering process. In this path, sensible heat from the sintering process would be used for power generation and directly converted into coupling-driven kinetic energy.

(c)

Path 3: Polygeneration system of iron and steel industry, chemical industry, and power industry based on ironmaking process. In this path, the blast furnace gas, blast furnace top pressure, and other by-products generated in the ironmaking process would be used for power generation or the preparation of carbon monoxide (CO).

(d)

Path 4: Polygeneration system of the iron and steel industry and building materials industry based on ironmaking and steelmaking processes. In this path, blast furnace slag, converter slag, and other major solid by-products would be used to prepare building materials, such as cement, devitrified glass, and aggregate.

It can be seen from Figure 2 that a total of 9 technologies are included in polygeneration path 1, i.e., T1–T9. Similarly, there are 2, 4, and 5 technologies embodied in polygeneration path 2 (T10, T11), polygeneration path 3 (T12–T15), and polygeneration path 4 (T16–T20), respectively.

3.2. Multi-Objective Optimization Model for Polygeneration System

In this section, the energy, environmental, and economic benefits of each polygeneration path will be evaluated. However, these quantitative results cannot be directly used to guide the polygeneration path application because the issue of technology selection in the actual polygeneration system is not considered in these theoretical results. Given the output of waste/by-products and the scale economy issue, there must be a choice between various technical paths, which means only one or several technologies in the polygeneration system can be adopted as the final application scheme. Therefore, a multi-objective optimization model will also be developed in this section to select the optimal technology or technology combination of each polygeneration path so as to maximize the energy, environmental, and economic benefits and guide enterprises to adopt the optimal path to realize the comprehensive utilization of waste/by-products.

3.2.1. Variables

The variables are the utilization ratios of the technologies. Here, the utilization ratio of a technology is defined as the proportion of by-products that are utilized by the technology to their total generation amounts. Therefore, the variables should range between 0 and 1.

3.2.2. Objectives

In addition to ensuring the realization of single-dimension environmental goals, including energy conservation and pollution emission reduction, the synergistic and conflicting effects of various technologies on environmental objectives and the corresponding economic benefit are also considered in this study. Consequently, three optimization objectives are set in this study, including: (a) the maximum of energy conservation; (b) the maximum of pollutants emission reduction; and (c) the maximum of economic benefits. It should be emphasized that SO₂, NO_x, and PM will be converted into unified pollutant emission indicators according to their pollutant emission trading prices in order to simplify the calculation.

The energy benefit of polygeneration systems includes two parts: direct energy benefit and indirect benefit. The former refers to energy saved, or secondary energy generated in the way of energy utilization, while the latter refers to the energy consumption of resources. Production and processing can be reduced by reducing the resource consumption of the given sink industry under the resource utilization approach. Thus, the energy benefit of the polygeneration system can be calculated by Equation (1).

$$EC=CSO\times {\displaystyle \sum _{pro}{\displaystyle \sum _{bp}{\displaystyle \sum _{t}S{R}_{pro}}}}\times G{F}_{pro,bp}\times Rati{o}_{pro,bp,t}\times ({\displaystyle \sum _e{E}_{t,e}\times Eco{e}_e}+{\displaystyle \sum _{res}{E}_{t,res}\times {Ecoe}_{res}})$$

(1)

where $EC$ represents the energy benefit of a polygeneration system; $CSO$ is the crude steel output; $pro$ is the production process; $bp$ is the by-product; $t$ is the polygeneration technology; $e$ is the type of fuel; $res$ is the type of resource; $S{R}_{pro}$ is the process product coefficient of $pro$; $G{F}_{pro,bp}$ is the production coefficient of by-product $bp$ from the production process $pro$; $Rati{o}_{pro,bp,t}$ is the utilization ratio of by-products $bp$ from the production process $pro$ when applying technology $t$; $E_{t,e}$ is the conservation amount of fuel $e$ contributed by technology $t$; $Eco{e}_e$ is the conversion coefficient of fuel $e$ to standard coal; $E_{t,res}$ is the conservation amount of resource $res$ contributed by technology $t$; and ${Ecoe}_{res}$ is the energy consumption coefficient of resource $res$ in the production process.

The environmental benefits of polygeneration systems are mainly reflected in the reduction in three types of air pollutants in this study, which are SO₂, NO_x, and PM, respectively. Similar to energy benefit, environmental benefit can also be divided into direct and indirect benefit, where direct benefit refers to the pollutants emissions reduction from the polygeneration system with polygeneration technologies application, and indirect benefit refers to the pollutants emissions avoided by saving energy and resources with polygeneration technologies application. Therefore, the environmental benefit of polygeneration systems can be calculated by Equation (2).

$$E{R}_p=CSO\times {\displaystyle \sum _{pro}{\displaystyle \sum _{bp}{\displaystyle \sum _{t}S{R}_{pro}}}}\times G{F}_{pro,bp}\times Rati{o}_{pro,bp,t}\times (DE{R}_{t,p}+{\displaystyle \sum _e{E}_{t,e}\times E{F}_{e,p}}+{\displaystyle \sum _{res}{E}_{t,res}\times E{F}_{res,p}})$$

(2)

where $ER$ is the emission reduction amount of pollutant $p$; $DE{R}_{t,p}$ is the direct emission reduction effect of technology $t$ on pollutant $p$; $E{F}_{e,p}$ is the pollutant p mission factor of fuel $e$ during the production or combustion process; and $E{F}_{res,p}$ is the pollutant $p$ mission factor of resource $res$ during the production process.

The economic benefits of polygeneration systems are mainly determined by investment in fixed assets, operation and maintenance expenses, and the economic benefits of by-products. The investment in fixed assets refers to the equipment investment required by various polygeneration technology applications. The operation and maintenance expenses include the labor costs, material costs, and equipment depreciation required by polygeneration technologies to address the by-products. The economic benefit of by-products is contributed by energy and resources saved, as well as the symbiotic products. In short, the economic benefit of polygeneration systems can be calculated by Equation (3).

$$EB=CSO\times {\displaystyle \sum _{pro}{\displaystyle \sum _{bp}{\displaystyle \sum _{t}S{R}_{pro}}}}\times G{F}_{pro,bp}\times Rati{o}_{pro,bp,t}\times ({\displaystyle \sum _{bp}{B}_{t-bp}}+{\displaystyle \sum _e{B}_{t,e}}+{\displaystyle \sum _{res}{B}_{t,res}}-O{C}_t-F{I}_t\times A{R}_t)$$

(3)

where $EB$ represents the energy benefit of polygeneration system; $B_{t-bp}$ is the economic benefit of by-product $bp$ tackled by symbiotic technology $t$; $B_{t,e}$ is the economic benefit of technology $t$ through energy conservation; $B_{t,res}$ is the economic benefit of technology $t$ through resource conservation; $F{I}_t$ and $O{C}_t$ are the fixed investment cost and operation and maintenance (O&M) cost of technology $t$, respectively; and $A{R}_t$ is the annualized ratio of fixed investment in technology $t$.

As mentioned above, even for the same waste/by-product, different utilization approaches will create different energy, environmental, and economic benefits. Consequentially, determining the use approaches and amounts of various waste/by-products is the key to completing the optimization of the polygeneration system. Given this situation, in this study, the optimization variables in the polygeneration system are the utilization ratio of by-products $bp$ from the production process $pro$, i.e., $Rati{o}_{pro,bp,t}$.

3.2.3. Constraints

Three constraints are considered in the modeling process, as follows:

The constraint of waste/by-product utilization. In general, when the capacity of the iron and steel plant is given, the quantity of waste/by-products generated is also fixed. To be specific, the total utilization amount of a given waste/by-product by all technologies in the polygeneration system does not exceed its total production, as shown in Equation (4).

$${\displaystyle \sum _{t}CO{N}_{t,bp}} \le CSO\times {\displaystyle \sum _{pro}{\displaystyle \sum _{t}S{R}_{pro}}}\times G{F}_{pro,bp} , \mathrm{for} \forall bp$$

(4)

where $CO{N}_{t,bp}$ is the total utilization amount of by-product $bp$ by technology $t$.

The constraint of technological matching. The desorbed gas is generated by the tail gas of hydrogen production from coke oven gas, so the output of the desorbed gas should not be higher than 60% of the coke oven gas used for hydrogen production, as shown in Equation (5).

$$0 \le O{P}_{dg} \le Q_{cogh}\times 60\%$$

(5)

where $O{P}_{dg}$ is the output of the desorbed gas, and $Q_{cogh}$ is the amount of coke oven gas used for hydrogen production.

The logical constraint. The utilization ratio of each waste/by-product should be between 0 and 100%, i.e.,

$$0 \le {\displaystyle \sum _{t}CO{N}_{t,bp}}/(CSO\times {\displaystyle \sum _{pro}{\displaystyle \sum _{t}S{R}_{pro}}}\times G{F}_{pro,bp}) \le 100\%, \mathrm{for} \forall bp$$

(6)

3.2.4. The Multi-Objective Optimization Method (NSGA-II)

It could be found that there are multiple optimization objectives in the optimization model, and the correlation between them is significant. In addition, there are also many variables and constraints in the optimization model, and these variables follow a continuous distribution. Given this situation and referring to the previous studies [49,50,51], the non-dominated sorting genetic algorithm-II (NSGA-II) is developed in this study to solve the multi-objective optimization model.

NSGA-II is a kind of genetic algorithm. Its basic idea is to complete the coding of decision variables in a specific way and calculate the objective function of the initial population and the crowding distance of the population, which will be used as the basis for non-dominated sorting. Then, the better solution from the population will be selected as the parent, and the genetic, crossover, and variation operators are designed to generate new offspring. The above processes will be repeated until the maximum number of iterations is reached, at which time the final optimization results are output.

The non-dominated stratification method used by NSGA-II allows good individuals to have a greater chance of being inherited by the next generation. And the fitness sharing strategy makes the individuals on the quasi-Pamto plane evenly distributed, maintains the diversity of the population, overcomes the excessive reproduction of super individuals, and prevents premature convergence. The specific explanation and operation processes of NSGA-II have been presented in previous studies [49].

3.2.5. Performance Evaluation of NSGA-II

Two indexes are developed in this study to verify the performance of NSGA-II and ensure reliable optimization results, as follows:

(a) Index 1: Spacing distribution uniformity of solution set (spacing metric). This index is used to evaluate the uniformity of solution set distribution. In general, the more uniform the distribution of the solution set, the lower the possibility of the results falling into the local optimum, and the more the solution set can reflect the synergistic and conflict relationship between various objectives. The principle of this method is to calculate the standard deviation of the Euclidean distance between two adjacent points, and the lower the standard deviation, the smaller the difference in the solution set, as shown in Equation (7).

$$SM=\sqrt{{\displaystyle \frac{1}{N-2}}{\displaystyle \sum _{n=1}^{N-1}\left(d_n-\overline{d}\right)}}$$

(7)

where $SM$ is the spacing metric; $N$ represents the population size; and $d_n$ represents the distance between the $n$th and $n+1$th points and $\stackrel{-}{d}$ represents the average distance.

(b) Index 2: Mean value of solution set. The mean value of the solution set could reflect the convergence and overall change trend of the solution set simultaneously. On the one hand, if the solution set converges after a certain number of optimization generations, it indicates that the optimization result has become stable and reliable. On the other hand, the mean value of the solution set can also reflect the improvement of the solution set in the optimization process.

3.3. Data Processing

In 2020, the global crude steel production was about 1878 million tons (Mt), with more than 56% contributed by China [52], which means China is the largest steel producer in the world. Based on this, the iron and steel plants in China were selected as cases in this study. To be specific, two types of typical iron and steel plants in China are selected as cases for modeling analysis in this study. One is a medium-scale iron and steel plant with a capacity of 3 Mt/a, and the other is a large-scale iron and steel plant with a capacity of 9 Mt/a. The related parameters are listed in

Table 1. Main equipment and scale parameters of iron and steel plants with a capacity of 3 Mt/a and 9 Mt/a.

Technical Parameter				Working Time of Main Equipment (d)
Grade of lump ore	62%	Ratio of steel scrap	10%/8% 1	Sintering machine	330
Grade of sintering	55%	Ratio of molten iron	90%/92% 1	Blast furnace	350
Grade of pellet	63%	Coefficient of sintering machine	1.45 t/(m2·h)	Converter	320
Ratio of lump ore	10%	Coefficient of blast furnace	2.40 t/(m3·d)	Continuous casting	320
Ratio of sintering	60%/67.3% 1	Withdrawing speed of continuous casting	2.3–2.6 m/min	Rolling mill	300
Ratio of pellet	30%/23.7% 1	Comprehensive grade	58.1%

Note: ¹ 60%, 30%, 10%, and 90% are the parameters for iron and steel plants with a capacity of 3 Mt/a, while 67.3%, 23.7%, 8%, and 92% are the parameters for iron and steel plants with a capacity of 9 Mt/a.

.

In addition to the above data, industrial production data, material transformation data, technology data, and economic data are also collected in order to quantify the benefits of different polygeneration paths and determine the final optimization scheme. The production data are shown in

Table 2. Production data of the iron and steel industry.

Parameter	Value	Unit
Coke production/crude steel production	0.355	/
Sinter production/crude steel production	0.991	/
Pelletizing production/crude steel production	0.495	/
Iron production/crude steel production	0.938	/
Coke oven gas production	435.0	m3/t coke
Blast furnace gas production	1621.0	m3/t iron
Flash gas production	5940.0	m3/t iron
Converter gas production	100.0	m3/t steel
Blast furnace slag/flash slag production	0.310	t/t iron
Converter slag production	0.110	t/t tseel
Desorption gas output from hydrogen production	0.6	m3/m3 coke oven gas for hydrogen production

.

The numerical characterization of energy, environmental, and economic benefits of by-product utilization technologies in the iron and steel industry is the key to evaluating the benefits of polygeneration systems. For the same waste/by-product, different utilization approaches will present different energy, environmental, and economic benefits. Therefore, it is necessary to decide which technology to choose according to the utilization approach of by-products and determine the energy, environmental, and economic benefits of this technology compared with the reference benchmark technology. Based on the previous studies [53,54,55,56] and surveys, the numerical characterization information of by-product utilization technologies in the iron and steel industry is shown in

Table 3. Material transformation data, technology data, and economic data.

No.	Energy Conservation (kgce/t Product)	SO2 Reduction (g/t Product)	NOx Reduction (g/t Product)	PM Reduction (g/t Product)	Fixed Cost (CNY/t Product)	O&M Cost (CNY/t Product)	Technical Life (Years)	Benefits per Unit of Raw Material (CNY/t Product)
T1	12.65	10.00	29.90	1.90	130.00	18.89	25	34.64
T2	19.00	2.68	2.87	0.49	2.50	0.50	25	10.90
T3	45.70	20.00	24.00	4.00	2.50	0.50	25	5.40
T4	19.60	7.40	7.92	1.36	41.60	14.00	25	32.00
T5	94.62	6.40	1.20	1.60	22.42	63.72	25	120.84
T6	27.62	44.00	7.20	46.80	116.75	21.16	25	44.56
T7	5.70	13.00	14.00	12.50	1259.30	1219.10	14	1367.90
T8	12.23	7.80	7.20	1.60	12.23	6.93	25	9.18
T9	12.50	9.60	10.80	10.20	12.50	6.36	25	13.50
T10	76.75	56.00	67.00	12.60	164.45	93.25	25	120.84
T11	16.37	19.50	18.00	4.00	16.37	4.16	25	6.51
T12	23.08	17.80	12.00	14.60	23.08	12.37	25	25.85
T13	16.00	16.00	32.00	48.00	160.00	255.00	25	291.00
T14	16.52	33.00	97.72	47.43	162.56	104.42	25	150.00
T15	72.00	42.70	28.60	49.92	7.20	180.00	25	220.00
T16	33.94	29.60	26.30	107.00	93.94	442.50	25	530.00
T17	24.62	31.67	92.46	41.05	24.62	19.57	25	27.93
T18	3.89	20.60	14.70	20.30	3.89	12.84	25	30.00
T19	0.05	0.27	0.29	0.05	0.54	0.07	25	0.08
T20	0.05	0.00	0.02	0.00	0.02	0.02	25	0.11

.

In addition, to improve the calculation efficiency and optimization quality of NSGA-II, the related parameters also should be determined properly. There are six types of parameters involved in NSGA-II: population size, maximum iteration times, crossover probability, mutation scale, mutation probability, and division of reference points, which are set as 200, 200, 0.9, 0.1, 0.2, and 4, respectively [56].

4. Results and Discussion

4.1. Performance Evaluation Results of NSGA-II

For index 1 (i.e., spacing metric), the results of two types of iron and steel plants are shown in Figure 3. It could be found that the spatial distribution uniformity of solution sets for the two optimization issues increases in the first 100 generations and decreases in the next 100 generations and tends to a low value at the end of the optimization process, which indicates that the distribution of optimization results is uniform.

For the index 2 (i.e., mean value of solution set), the results are shown in Figure 4. It can be seen that for the two optimization issues, the mean value of solution sets in all dimensions has been improved to some extent, and convergence has been achieved after 150 generations. In conclusion, the evaluation results of the two indexes are reasonable, which verify the reliability of the NSGA-II proposed in this study.

4.2. Optimization Results

The optimization results of energy conservation and emission reduction in the iron and steel plants obtained by the NSGA-II are shown in

Table 4. The average optimization results of energy conservation and emission reduction in the iron and steel plants.

	Iron and Steel Plant with 3 Mt/a Capacity
	Energy Conservation (ktce)	SO2 Reduction (t)	NOx Reduction (t)	PM Reduction (t)	Economic Benefit (CNY 108)
Maximum	638.88	51.39	84.10	54.03	4.20
Minimum	557.31	39.70	48.32	39.94	2.59
Average	588.98	45.86	61.53	47.03	3.740
	Iron and Steel Plant with 9 Mt/a Capacity
	Energy Conservation (ktce)	SO2 Reduction (t)	NOx Reduction (t)	PM Reduction (t)	Economic Benefit (CNY 108)
Maximum	2051.71	156.65	200.29	185.02	13.75
Minimum	1584.24	116.70	117.40	127.52	9.79
Average	1786.95	133.02	147.90	145.50	12.33

. For the iron and steel plant with 3 Mt/a capacity, the average value of energy conservation, SO₂ reduction, NO_x reduction, PM reduction, and economic benefit of the optimization scheme are 588.98 ktce, 45.86 t, 61.53 t, 47.03 t, and CNY 3.74 billion, respectively. For the iron and steel plant with 9 Mt/a capacity, the five optimized indicators are 1786.95 ktce, 133.02 t, 147.90 t, 145.50 t, and CNY 12.33 billion, respectively. These results showed that the polygeneration system can achieve good energy, environmental, and economic benefits, and there is a synergistic relationship between energy conservation and emission reduction on the whole in the polygeneration system.

As shown in Figure 5, there are synergistic and conflicting relationships among energy, environmental, and economic goals in both scales of iron and steel plants, which are also closely related to the scale of iron and steel plants. More specifically, for the iron and steel plant with 3 Mt/a capacity, the goals of economic benefit, PM reduction, and energy conservation are synergistic, while the energy conservation goal is in conflict with the SO₂ and NO_x reduction goals. This is mainly because the energy consumption and cost of some SO₂ and NO_x reduction technologies are high in the middle-scale iron and steel plants. For the iron and steel plant with 9 Mt/a capacity, the goals of energy conservation, SO₂ and NO_x reduction are synergistic, which means compared with the middle-scale iron and steel plant, the large-scale iron and steel plant may have more advantages in energy conservation, SO₂ and NO_x reduction. However, the energy conservation goal is in conflict with the goals of economic benefit and PM reduction in this case. In short, it is necessary to make technology selection to avoid conflicts between different objectives when establishing the polygeneration system, and the scale of iron and steel plants should also be considered in this process.

4.3. Final Decision Schemes

As for the optimal variables, the average utilization ratio results of 20 technologies in the polygeneration system are shown in Figure 6. In general, the average utilization ratio of about half of the technologies in two steel plants with different production is similar. On the contrary, there exist some differences in the utilization of several by-products, such as coke oven gas (T2–T7), blast furnace gas (T12 and T13), and blast furnace slag (T16–T18), due to the scale effects of the technologies. This result demonstrates the necessity of considering the plant’s scale in technical selection.

The polygeneration system’s decision-making scheme is discrete, which means the above optimal results cannot be used directly to guide business decisions. Therefore, it is necessary to transfer the solution set to obtain the final decision scheme in order to select the polygeneration technologies from the perspective of iron and steel enterprises. When considering the scale economy, 13 technologies should be adopted by the iron and steel plant with 3 Mt/a capacity and 12 technologies should be adopted by the iron and steel plant with 9 Mt/a capacity, and the technology selection under various polygeneration paths is shown in

Table 5. Polygeneration technology selection under the final decision scheme.

Polygeneration Path	Iron and Steel Plant with 3 Mt/a Capacity	Iron and Steel Plant with 9 Mt/a Capacity
Path 1	T1, T2, T3, T4, T5, and T8	T1, T2, T3, T5, and T8
Path 2	T10 and T11	T10
Path 3	T14 and T15	T12 and T15
Path 4	T16, T17, and T19	T16, T17, T19, and T20

.

Based on the scheme and the technical scheme proposed in Table 5, various types of products can be produced by using the by-products or wastes from the iron and steel plant. To be specific, for the iron and steel plant with 3 Mt/a of capacity, 407 GWh of power, 31.5 kt of hydrogen (H₂), 935.9 kt of cement, 218.4 kt of devitrified glass, 42.1 kt of sulfuric acid (H₂SO₄), 47.3 kt of coal tar, and 263.0 kt of MMA will be output with the application of the above 12 technologies in the polygeneration system. It should be noted that methanol is also produced in the above process, but it is used as a raw material for MMA production. However, for the iron and steel plant with 9 Mt/a capacity, the technical scheme of its polygeneration system is different from that of the iron and steel plant with 3 Mt/a capacity, which also leads to differences in their final products. Under the optimized technical scheme, 1308 GWh of power, 338.8 kt of methanol, 1.5 million tons of cement, 300 m³ of liquified natural gas (LNG), 1.3 million tons of devitrified glass, 129.1 kt of H₂SO₄, 145.3 kt of coal tar, 807.1 kt of MMA, and 147.4 kt of aggregate can be produced through the polygeneration system with an iron and steel plant with 9 Mt/a capacity as its core.

Polygeneration systems can bring significant energy, environmental, and economic benefits by providing a large number of electricity and other industrial products.

According to the final technical scheme, the iron and steel plant with 3 Mt/a capacity can achieve energy conservation with 410.53 ktce, SO₂ reduction with 37.85 t, NO_x reduction with 68.34 t, PM reduction with 31.89 t, and the economic benefit is CNY 244.92 million. In other words, the energy and environmental benefits of the polygeneration system can reduce the energy consumption, SO₂ emission, NO_x emission, and PM emission of the iron and steel plant by 25.3%, 17.5%, 27.1%, and 15.9%, respectively. With the expansion of iron and steel plants’ capacity, the energy, environmental, and economic benefits of polygeneration systems are also improved. For the iron and steel plant with 9 Mt/a capacity, the energy conservation, SO₂ reduction, NO_x reduction, PM reduction, and economic benefit are 2124.49 ktce, 159.99 t, 206.67 t, 100.67 t, and CNY 1088.93 million, respectively, and the energy and environmental benefits of the polygeneration system are equivalent to reducing the energy consumption, SO₂ emission, NO_x emission, and PM emission of the iron and steel plant by 42.6%, 24.1%, 26.7%, and 16.3%, respectively.

The difference in the performance of the polygeneration system caused by the scale of iron and steel plants can be explained from two aspects. One reason is that the output of by-products or waste of iron and steel plants does not change linearly with their capacity. The output of coke oven gas and blast furnace gas in the iron and steel plant with 9 Mt/a capacity is more than three times that of the iron and steel plant with 3 Mt/a capacity, while the output of blast furnace slag and converter slag of the former is three times lower than that of the latter. The other reason is that the output of by-products or waste determines the optimal technology selection scheme of the polygeneration system. For instance, the utilization of residual heat and pressure is different for two iron and steel plants with different scales (T14 and T15 for the iron and steel plant with 3 Mt/a capacity, while T12 and T15 for the iron and steel plant with 9 Mt/a capacity). In addition, there are also differences in the products of the polygeneration system. For example, aggregate and cement will be generated at the same time by using converter slag if the capacity of the iron and steel plant is 9 Mt/a, while only cement will be generated if the capacity is 3 Mt/a.

At present, China’s iron and steel industry is continuing to carry out mergers and acquisitions, and the industrial concentration is further improved. Initially, these activities are affected by the national environmental protection goals; simultaneously, we believe that it is conducive to improving the energy conservation and emission reduction potential of the polygeneration system in the future based on the results of this study.

Another fact reflected by the above results is that polygeneration systems are an important approach to broadening the space for energy conservation and pollutant emission reduction in industrial sectors, which is of great significance for the industrial sectors to achieve the carbon emission control target and pollutant emission reduction target. In 2023, the crude steel production in China will have reached about 1.02 billion tons [57]. Based on the energy and environmental benefits of an iron and steel plant with 3 Mt/a capacity, in the case of widely promoted polygeneration systems in the iron and steel industry in China, it can save energy over 139 million tce (accounting for approximately 2.4% of China’s total energy consumption in 2023) per year and reduce SO₂, NO_x, and PM with 12.86 kt/a, 23.22 kt/a, and 10.83 kt/a, respectively. Most of these benefits can be achieved through policy guidance and incentives, which provides an important idea for the formulation of carbon and pollutant reduction policies in industrial sectors in China.

5. Conclusions

In this study, the iron and steel industry is selected as the by-product industry, while the chemical, building materials, and power industries are selected as the sink industries to simulate the polygeneration system, and four typical polygeneration paths are set up. And then, the technology structure system of the polygeneration process is modeled based on the bottom–up method, and the modeling process follows the chain of “supplier—waste/by-products—technologies—adopter”. Based on the above, a polygeneration process model including eight types of by-products and 20 technologies has been established. Apart from this, a multi-objective optimization model is developed in this study to select the optimal technology or technology combination of each polygeneration path under the constraints of energy, environmental, and economic objectives, as well as guide enterprises to adopt the optimal path to realize the comprehensive utilization of by-products. The main conclusions of this study are concluded below.

(1)

The energy conservation and emission reduction measures have an impact on multiple goals such as energy, environment, and economy during the application process. There are complex collaborative and conflicting relationships between different targets, and there is a risk of implicit transfer between various goals. The multi-objective optimization algorithm represented by the NSGA-II algorithm can well handle the above issues, which can provide various solutions for decision-makers, and decision-makers can formulate energy conservation and emission reduction strategies based on their preferences.

(2)

There are synergistic and conflicting relationships among energy, environmental, and economic goals in the process of applying polygeneration technologies in the iron and steel industry, which is also closely related to the scale of iron and steel plants. Therefore, it is necessary to make technology selection to avoid conflicts between different objectives when establishing the polygeneration system, as well as maximize the energy, environmental, and economic benefits for iron and steel plants with different scales.

(3)

Polygeneration systems can significantly improve the energy conservation and pollutant emission reduction potential of the iron and steel industry, which is crucial for achieving the pollution and CO₂ reduction targets of China’s industrial sector in the future. Polygeneration technologies can fully utilize waste or by-products in the iron and steel plants, such as coke oven gas, sintering waste heat, and steel slag, which will be converted into industrial products, including chemical products, electricity, building materials, and others. The independent energy conservation and emission reduction potential of a single industry is limited currently. And it has become a trend for the industrial sector to fully utilize by-products/wastes from different industries through the establishment of polygeneration systems to save energy and reduce emissions. In the future, it is necessary to focus on the multi-level utilization of energy in the industrial sector and the application of circular economy measures.

In conclusion, critical insights are provided in this study in the energy and environmental management of iron and steel through establishing an inter-industry polygeneration system with the iron and steel industry as the core and multi-objective optimization model. The results of this study can provide references for energy conservation and emission reduction management in the global iron and steel, chemical, building materials, and power industries.