Economic Analysis of an Integrated Steel Plant Equipped with a Blast Furnace or Oxygen Blast Furnace

Abstract

The oxygen blast furnace with top gas recycling (TGR-OBF) ironmaking technology can reduce CO₂ emissions, especially when combined with carbon capture and storage technology (CCS). However, the successful commercialization of technology cannot be achieved without economic evaluation. This study applied the Box–Behnken design method and a Monte Carlo simulation-based risk analysis to assess the sensitivity of influencing factors affecting the net present value (NPV) of an integrated steel plant (ISP) and to predict the impact of variable market scenarios on the NPV of three ISPs. The results indicated that among the three ISPs, the conventional ISP (C-ISP) is the least profitable, followed by the ISP equipped with TGR-OBF and CCS (ISP-OBF-CCS), and the most profitable is the ISP equipped with TGR-OBF (ISP-OBF), which is at least CNY 0.392 Bn and CNY 1.934 Bn more profitable than the ISP-OBF-CCS and C-ISP respectively. Under the current Chinese carbon trading policy and the cost of CCS, CCS technology does not make a company profitable. This study explored an approach for analyzing ISP economic applicability under uncertain markets, which can be used as a reference for the development of alternative processes for steel production.

1. Introduction

The carbon emissions of the steel industry account for about 7% of the total global emissions [1,2,3]. In China, carbon emissions from the steel industry are the largest in the manufacturing sector, accounting for about 15% of the country’s total emissions [4,5,6]. The application of an oxygen blast furnace with top gas recycling (TGR-OBF) in the ironmaking process can significantly reduce energy consumption and CO₂ emissions [7,8,9]. In our previous research, the energy consumption and CO₂ emissions of a conventional integrated steel plant (C-ISP) and integrated steel plant equipped with oxygen blast furnace (ISP-OBF) were calculated, and the advantages of ISP-OBF from a technical point of view were demonstrated [10]. However, the widespread application of a technology depends on its economic performance. To date, most economic analyses applied to ISP-OBF focused on fixed market prices [11,12]. These studies did not consider the impact of price fluctuations in investment construction, raw materials, product and carbon trading, etc. The above studies, while technically valid, did not provide an exhaustive evaluation given the stochastic character of the factors involved. The results of these studies make it difficult to assess profitability in uncertain markets. Therefore, appropriate economic evaluation methods need to be selected for use in uncertain markets. The Box–Behnken design (BBD) method and Monte Carlo method were recently used for an uncertain market economics analysis [13,14]. The BBD method is a kind of response surface methodology. It can establish a second-order surface model through continuous variables and then analyze the relationship between influencing factors and key indicators [15,16]. The BBD method has been widely used in chemical and other process industries in recent years. S. Ajebli [17] used Box–Behnken design (BBD) to optimize the activated carbon-produced process parameters. Oladayo Adeyi [13] used Box–Behnken design (BBD) to present a techno-economic assessment and uncertainty analysis of a scaled-up integrated process for bioactive extract production from Senna alata (L.) leaves. The Monte Carlo simulation is often regarded as an accurate stochastic method employed in uncertain economic analyses of engineering systems with many uncertain variables [18]. The economic evaluation combined with a Monte Carlo simulation minimizes uncertainty and provides a more accurate estimate of future returns [19]. Therefore, this study used a combination of the BBD method and the Monte Carlo method, based on the results of our previous studies, to investigate the influence of market uncertainty on the ISP economy under the alternative ironmaking process of a TGR-OBF. We did this in order to compare the economics of a TGR-OBF with the economics of blast furnaces from a whole-plant perspective. Firstly, three integrated steel plants were defined: a C-ISP, an ISP-OBF, and an ISP-OBF equipped with carbon capture and storage technology (ISP-OBF-CCS). Then, we fitted three binary regression models in relation to the key economic indicator with the main influencing factors based on the JMP (Trial Version Pro 16) software platform using the BBD method. Finally, sensitivity and uncertainty analyses, based on the Monte Carlo simulation technique, were performed on these models. This study explored an approach to analyzing ISP economics applicable under uncertain markets.

2. Models and Methods

2.1. Integrated Steel Plant and Its Key Economic Indicator

In order to study and compare the economy of a TGR-OBF with that of a BF from the perspective of the whole steel plant, three integrated steel plants were defined based on the operation data of the Baosteel steel plant and the TGR-OBF model established in our previous research [10]. The three ISPs were a C-ISP, an ISP-OBF, and an ISP-OBF-CCS. According to our previous research [10], an ISP consists of the steel manufacturing system and the energy system, as shown in Figure 1. For the three ISPs, the steel manufacturing system comprises coking, sintering, lime production, ironmaking (BF or TGR-OBF), steelmaking, and steel rolling processes. The purchased iron ore, pellets, coking coal, and coal are made into a hot rolled strip and metallurgical gas with the help of the energy system. In the analysis of a C-ISP, the Baosteel No. 2 blast furnace and its steel production process (located in Shanghai, China) were used as a reference case. We used the operating parameters of the BF and its steel production process in this study, as shown in Figures S1 and S2 (Supplementary Materials). We studied the economy of an ISP-OBF or ISP-OBF-CCS by replacing the blast furnace of the C-ISP with a TGR-OBF. Therefore, we used the TGR-OBF process model developed in our previous research [10] to calculate the operating parameters of a TGR-OBF for the same ore volume and fraction, the same hot metal production and composition, and the same production rate as the blast furnace. By combining the calculated TGR-OBF process model with the operation data of the Baosteel steel plant (without the BF process), the ISP-OBF or ISP-OBF-CCS were analyzed. The operating parameters of a TGR-OBF and its steel production process in this study are shown in Figures S3 and S4 (Supplementary Materials).

The net present value (NPV) is the most commonly used indicator for evaluating investment projects. This is because it represents the value that the project will create for investors and also enables project prioritization in situations involving several investment options. Therefore, the NPV served as a key indicator for evaluating the economy of an ISP in this study. In the economic analysis of the three ISPs, the following assumptions were made in this study:

1.

According to Baosteel’s data [20,21], the lifetime of the blast furnace and its corresponding steelmaking process major units for the first overhaul are approximately fifteen years. Therefore, the annual cash flow of the three ISPs for a fifteen-year time span were calculated, and the annual net cash flow equation of an ISP is shown in Equation (1). Therefore, the NPV value in this study is the sum of the annual net cash flow of an ISP over fifteen years.

$$NPV={\displaystyle \sum _{t=1}^n{(CF)}_t}{(1+i_c)}^{-t}$$

(1)

where (CF)_t is the net cash flow after tax of the year t, n is assessment period (years), and i_c is the discount rate.

2.

In the ISP-OBF analytic system, most of the top gas produced by a TGR-OBF is recycled into a furnace as recycled gas after CO₂ removal and preheating. Meanwhile, the removed CO₂ is emitted into the environment.

3.

In the ISP-OBF-CCS analytic system, the removed CO₂ was captured, transported, and stored underground. The electricity consumption of the CO₂ removal and capture was 120.65 kWh·t-steel⁻¹. The cost of transport and storage is mentioned in Section 2.2.

2.2. Main Influencing Factors and Their Levels

In order to identify the main economic influencing factors affecting an ISP to determine the scope of the NPV calculation, screening and assumptions about these factors were required. An ISP has many influencing factors. However, factors that have little influence on an ISP, such as personnel wages, water prices, scrap steel prices, electricity prices, etc., were predetermined according to the statistics of China’s steel industry [22]. In addition, under the premise of fixed output, influencing factors such as personnel quotas, water consumption, maximum productive capacity, iron ore consumption, and various taxes were nearly unchanged. Therefore, combined with steel industry statistics and Baosteel’s production data [20,21,23], parameters for the economic analysis of the ISP were identified in this study as shown in

Table 1. Data for an ISP’s economic analysis.

Parameter	Value	Parameter	Value
Construction period/a	2	Maximum productive capacity/t·a−1	3.97 × 106
Calculation period/a	15	Iron ore consumption/t·a−1	8.28 × 106
Production period beginning year/a	3rd	Water consumption/t·a−1	1.15 × 107
Design capacity reaching year/a	3rd	Steel scrap consumption/t·a−1	3.01 × 105
Benchmark yield of project/%	11	Corporate income tax rate/%	25
Personnel quota/person	3086	Proportion of welfare costs accounted for annual total wages/%	2
Personnel wages/CNY·a−1·person−1	80,000	Proportion of repair costs accounted for annual depreciation charge/%	10
Water price/CNY·t−1	5	Proportion of other manufacturing expenses accounted for annual depreciation charge/%	4
Scrap steel prices/CNY·t−1	3600	Proportion of sales expense accounted for annual sales revenue/%	0.5
Electricity price/CNY·kWh−1	0.52	Proportion of administration expenses accounted for annual total wages /%	1.5
Limestone price/CNY·t−1	140	Depreciation period of fixed assets/a	13
City maintenance and construction tax/%	1	Depreciation method	Composite life method
Education surcharges/%	3	Added-value tax rate/%	13

.

However, the coal consumption, electricity consumption, CO₂ emissions, and emissions allowances of the three ISPs were different. Based on the Baosteel No. 2 blast furnace and its steel production process production data, the coal consumption, electricity consumption, CO₂ emissions, and emissions allowances of the C-ISP could be obtained. By combining the calculated TGR-OBF process model with the operation data of the Baosteel steel plant (without the BF process), the coal consumption, electricity consumption, and CO₂ emissions of the ISP-OBF and ISP-OBF-CCS could be obtained. The CO₂ emissions in this study were the net CO₂ emissions of the ISP; their calculation refers to our previous research [10]. Then, the emissions allowances of the ISP-OBF and ISP-OBF-CCS could be calculated based on the CO₂ emissions [24]. For the ISP-OBF-CCS, the cost of CO₂ transport and storage was set at CNY 68 ·tCO₂⁻¹ based on the international average [25]. The details are shown in

Table 2. Economic analysis parameters for three ISPs.

Parameters	C-ISP	ISP-OBF	ISP-OBF-CCS
Coal consumption/t·a−1	2.29 × 106	1.73 × 106	1.73 × 106
Electricity consumption/kWh·a−1	7.41 × 104	2.42 × 105	2.42 × 105
CO2 emissions per ton of product/tCO2·t-steel−1	2.03	1.83	1.83
Emissions allowances/tCO2·a−1	8,279,248.05	7,463,558.58	7,463,558.58
CO2 transport and storage costs /CNY·tCO2−1			68
CO2 storage/tCO2·a−1			1,629,592.57

.

In addition, the cost composition of Chinese steel enterprises mainly included raw materials, energy and reducing agents, and construction investment costs. Among them, raw materials, energy and reducing agents, and construction investment costs accounted for 42.6%, 35.2%, and 14.2% of the total cost, respectively [23]. Thus, the main influencing factors affecting the economy of an ISP include iron ore price, coking coal price, and construction investment costs. The carbon price also served as an important factor affecting the economy of an ISP. The Chinese steel industry was included as one of the key industries in the national carbon market in 2017, based on the requirements of the China Development and Reform Commission’s “Key Work on Effective Launch of the National Carbon Emissions Trading Market” [22]. Almost all steel companies were included in the national carbon market, so the carbon price was also considered to be the main influencing factor.

Therefore, the iron ore price (P_o), coking coal price (P_co), construction investment cost (I_c), carbon price (P_ca), and hot rolled strip price (P_s) were used as the main influencing factors to carry out the economic assessment in this study.

In this study, raw materials and fuels were converted to iron ore and coking coal consumption because the raw materials and fuels for the three ISPs included pellets and pulverized coal in addition to iron ore and coking coal. In the three ISPs, raw material consumption mainly referred to iron ore and pellets, and fuel consumption referred to coking coal and pulverized coal. In order to reflect the impact of iron ore and coking coal price changes on the three ISPs and the economy, this study converted the consumed pellets into iron ore consumption according to the current price and also converted the consumed pulverized coal into coking coal consumption according to the current price. Hence, the total iron ore consumption and total coking coal consumption of the C-ISP were obtained. They were 2082.66 kg/t-steel and 576.03 kg/t-steel, respectively. The total iron ore consumption and total coking coal consumption of the ISP-OBF and ISP-OBF-CCS was 2082.66 kg/t-steel and 434.60 kg/t of steel, respectively. Detailed data are shown in Tables S1 and S2 (Supplementary Materials).

Generally, the carbon price in the carbon market was a response to the average social cost of GHG emission reductions in a country or region. China’s carbon prices and abatement costs were not fully correlated, and the phenomena of high volatility, large differences, and discontinuous trading in the process of pilot carbon trading had emerged. Therefore, referring to the current carbon price trend in China [26], the carbon price level was set at 1–200 CNY·t⁻¹ in this study. The construction investment of a C-ISP with a 5000 m³ blast furnace project was generally around CNY 18.85 billion, according to the Baosteel blast furnace process project [27]. However, the construction investment is influenced by changes in technology and market material (steel, cement, etc.) prices. Therefore, based on experience, the construction investment level of C-ISP was set at CNY 18–20 billion. For the ISP-OBF, the cost of retrofitting a blast furnace to a TGR-OBF was CNY 500 million, including the cost of furnace renovation, top gas recycling, and CO₂ capture retrofit and oxygen plant expansion [28]. Therefore, the construction investment level of the ISP-OBF was set at CNY 18.5–20.5 billion. For the ISP-OBF-CCS, the cost of CCS is mentioned above; the construction investment of the ISP-OBF-CCS was the same as the ISP-OBF, and its level was CNY 18.5–20.5 billion. At the same time, we assumed the iron ore price level, coking coal price level, and hot rolled strip price level based on the purchase price of Baosteel in the past five years [20] and the prediction of experts [21]. Finally, the main influencing factors and their levels of the three ISPs are shown in

Table 3. Main influencing factors and levels.

Factors	Levels
	C-ISP	ISP-OBF or ISP-OBF-CCS
Po/CNY·t−1	400–900	400–900
Pco/CNY·t−1	1100–1600	1100–1600
Ps/CNY·t−1	3500–4500	3500–4500
Pca/CNY·t−1	1–200	1–200
Ic/Bn CNY	18–20	18.5–20.5

.

2.3. Box–Behnken Experimental Design and Statistical Analysis

After determining the main influencing factors and levels, a Box–Behnken experimental design (BBD) was performed for these three ISP analytic systems. A total of 46 experiments with 6 central points were conducted for each ISP analytic system. Meanwhile, the main influencing factors were expressed at three levels. In a C-ISP analytic system, the main influencing factors were iron ore price (400 CNY·t⁻¹, 650 CNY·t⁻¹, and 900 CNY·t⁻¹); coking coal price (1100 CNY·t⁻¹, 1350 CNY·t⁻¹, and 1600 CNY·t⁻¹); hot rolled strip price (3500 CNY·t⁻¹, 4000 CNY·t⁻¹, and 4500 CNY·t⁻¹); carbon price (1 CNY·t⁻¹, 100.5 CNY·t⁻¹, and 200 CNY·t⁻¹); and construction investment costs (CNY 180 × 108, CNY 190 × 108, and CNY 200 × 108). Meanwhile, in an ISP-OBF or ISP-OBF-CCS analytic system, the main influencing factors were iron ore price (400 CNY·t⁻¹, 650 CNY·t⁻¹, and 900 CNY·t⁻¹); coking coal price (1100 CNY·t⁻¹, 1350 CNY·t⁻¹, and 1600 CNY·t⁻¹); hot rolled strip price (3500 CNY·t⁻¹, 4000 CNY·t⁻¹, and 4500 CNY·t⁻¹); carbon price (1 CNY·t⁻¹, 100.5 CNY·t⁻¹, and 200 CNY·t⁻¹); and construction investment costs (CNY 185 × 108, CNY 195 × 108, and CNY 205 × 108). The various experimental conditions of the main influencing factors, and corresponding NPV calculated as a response according to the BBD (46 runs) model, are shown in the form of design matrices in Tables S3–S5 (Supplementary Materials).

Following the Box–Behnken design matrix of the experiments, we performed the study and obtained the results. The JMP platform was applied for a regression analysis and analysis of variance (ANOVA). The expression of the fitted second-order polynomial model is shown in Equation (2):

$$Y=a_0+{\displaystyle \sum _{i=1}^n{a}_i}{x}_i+{\left({\displaystyle \sum _{i=j}^n{a}_{ii}{x}_i}\right)}^2+{\displaystyle \sum _{j=i+1}^{n-1}{\displaystyle \sum _{j=i+1}^n{a}_{ij}{x}_i{x}_j+\beta }}$$

(2)

where Y is the predicted response, which was the NPV in this study; a₀, a_i, a_ii, and a_ij are the regression coefficients; x_i and x_j are the input variables, which were P_o, P_co, P_ca, P_s, and I_c in this study; and β is the random error.

With the BBD, the influence of the main influencing factors (iron ore price, P_o; coking coal price, P_co; hot rolled strip price, P_s; carbon price, P_ca; and construction investment costs, I_c) on the NPV were investigated. In addition, ANOVA was used in order to test the significance of the model coefficients (p < 0.05). Moreover, a t-test and p-value were used to verify the importance of the regression coefficient. Finally, the adequacy of the model was determined by assessing the lack of fit.

2.4. Risk Analysis of an ISP

As a key economic indicator, the NPV was identified to determine the profitability of the three ISPs. Hence, the risk analysis was intended to form the certainty level of the NPV subjected to uncertain influencing factors. Besides this, in order to identify the main influencing factors that have tendencies to significantly perturb the NPV, a sensitivity analysis was necessary. Therefore, sensitivity and uncertainty analyses were performed for the three ISP analytic systems using the Monte Carlo simulation available in the JMP (Trial Version Pro 16. SAS Institute Inc., Cary, NC, USA) platform.

The iron ore price, coking coal price, hot rolled strip price, carbon price, and construction investment costs were the uncertain influencing factors that were taken into account for the sensitivity and uncertainty analyses. At the same time, the NPV was the forecast indicator. The construction investment was assumed to have a triangular distribution, and the remaining uncertain influencing factors were considered to be uniform distributions in the course of the Monte Carlo simulation, as shown in

Table 4. Uncertain factors distribution for Monte Carlo simulation of three ISPs.

Uncertain Factors	Distribution Pattern	Overall Distribution Range
		C-ISP	ISP-OBF or ISP-OBF-CCS
Pca/CNY·t−1	Uniform distribution	[1, 200]	[1, 200]
Po/CNY·t−1	Uniform distribution	[400, 900]	[400, 900]
Pco/CNY·t−1	Uniform distribution	[1100, 1600]	[1100, 1600]
Ps/CNY·t−1	Uniform distribution	[3500, 4500]	[3500, 4500]
Ic/Bn CNY	Triangular distribution	[18, 19, 20]	[18.5, 19.5, 20.5]

. Moreover, about 10,000 simulation trials were made to achieve low mean standard errors in the NPV.

3. Results and Discussion

3.1. BBD Based Models and ANOVA

Based on the JMP platform, the relationship between the key economic indicator and the main influencing factors was fitted by using the second-order polynomial equations for the three ISPs. The quadratic model equations are expressed by Equations (3)–(5) after removing the insignificant terms.

In a C-ISP analytic system:

$$\begin{array}{l}NP{V}_{C-ISP}=-409506584.9-34030170.83P_o-9425785.292P_c+16261275.35P_s\\ \begin{array}{cc}\begin{array}{cc}\begin{array}{cc}& \end{array}& \end{array}& \end{array}\hspace{1em}\hspace{1em}-1198614.731P_{ca}-1.870260497I_c+1.814890000\times {10}^{-11}{I}_c{}^2\end{array}$$

(3)

In an ISP-OBF analytic system:

$$\begin{array}{l}NP{V}_{ISP-OBF}=-252761851.7-34030170.83P_o-7116128.076P_c+16261275.35P_s\\ \begin{array}{ccc}\begin{array}{cc}& \end{array}& & \end{array}\hspace{1em}\hspace{1em}-1082849.976P_{ca}-1.888409365I_c+1.814890000\times {10}^{-11}{I}_c{}^2\end{array}$$

(4)

In an ISP-OBF-CCS analytic system:

$$\begin{array}{l}NP{V}_{ISP-OBF-CCS}=23113056578-34030170.83P_o-7116128.076P_c+16261275.35P_s\\ \begin{array}{cc}\begin{array}{cc}\begin{array}{cc}\begin{array}{cc}& \end{array}& \end{array}& \end{array}& \end{array}\hspace{1em}\hspace{1em}-857108.7084P_{ca}-1.888409359I_c+1.814890000\times {10}^{-11}{I}_c{}^2\end{array}$$

(5)

The main influencing factors—iron ore price (P_o), coking coal price (P_co), hot rolled strip price (P_s), carbon price (P_ca), and construction investment costs (I_c)—were the independent singular variables. The key economic indicator, NPV, was the predicted response.

Table 5. ANOVA results for C-ISP.

Source	Degree of Freedom	Sum of Squares	F-Value	Pro. > F
Po	1	1.1581 × 1021	1,791,336	<0.0001	Significant
Ps	1	1.0577 × 1021	1,636,131	<0.0001	Significant
Pco	1	8.8845 × 1019	137,430.8	<0.0001	Significant
Ic	1	2.2301 × 1019	34,496.67	<0.0001	Significant
Pca	1	2.2758 × 1017	352.0256	<0.0001	Significant
Ic × Ic	1	3.4370 × 1015	5.316600	0.0265	Significant

,

Table 6. ANOVA results for ISP-OBF.

Source	Degree of Freedom	Sum of Squares	F-Value	Pro. > F
Po	1	1.1581 × 1021	1,791,336	<0.0001	Significant
Ps	1	1.0577 × 1021	1,636,131	<0.0001	Significant
Pco	1	5.0639 × 1019	78,331.49	<0.0001	Significant
Ic	1	2.2301 × 1019	34,496.67	<0.0001	Significant
Pca	1	1.8574 × 1017	287.3106	<0.0001	Significant
Ic × Ic	1	3.4370 × 1015	5.316600	0.0265	Significant

and

Table 7. ANOVA results for ISP-OBF-CCS.

Source	Degree of Freedom	Sum of Squares	F-Value	Pro. > F
Po	1	1.1581 × 1021	1,791,336	<0.0001	Significant
Ps	1	1.0577 × 1021	1,636,131	<0.0001	Significant
Pco	1	5.0639 × 1019	78,331.49	<0.0001	Significant
Ic	1	2.2301 × 1019	34,496.67	<0.0001	Significant
Pca	1	1.1637 × 1017	180.0060	<0.0001	Significant
Ic × Ic	1	3.4370 × 1015	5.316600	0.0265	Significant

present the BBD-based ANOVA results for the three quadratic models. The F and p values were used to indicate the significance of each variable, with p values of <0.05 and high F values considered significant; p values greater than 0.05 and low F values were deemed insignificant. The ANOVA results of the three ISPs indicated that the three quadratic models and all the variables were significant. The coefficient of determination (R²) of 0.99999 in this study suggests that the three regression models have the best fit quality. Moreover, Figure 2 shows that the predicted and observed results were highly agreed upon based on the JMP platform. Therefore, the predicted values of the three quadratic models can be considered close to the response’s actual values.

Then, the contour plots and 3D surfaces showing the effects of variables (P_o, P_co, P_s, P_ca, and I_c) on the NPV were obtained from these three regression models. Figure 3 shows the relationship between the two variables (P_ca and P_co) and the response NPV when the three influencing factors are kept constant (I_c= CNY 1.9 Bn, Ps = 4000 CNY·t⁻¹, and Po = 650 CNY·t⁻¹). The red line is the contour of the response NPV, representing the corresponding values of the two variables (P_ca and P_co) at the same response value. The red area represents the area that does not meet the requirement. In this study, the contour NPV was CNY 0, and the red area represented the region where the NPV was less than CNY 0. From Figure 3a, we can find that in the C-ISP, if I_c = CNY 1.9 Bn, Ps = 4000 CNY·t⁻¹, and Po = 650 CNY·t⁻¹, regardless of the value of P_ca, there is a high possibility that the steel company will face a loss once the coking coal price is greater than 1400 CNY/t. However, from Figure 3b,c, the NPV of the ISP-OBF and ISP-OBF-CCS was definitely greater than zero for the same market conditions (IC = CNY 1.95 Bn, Ps = 4000 CNY·t⁻¹, and Po = 650 CNY·t⁻¹). This suggested that the profitability of the ISP-OBF or ISP-OBF-CCS was better than that of the C-ISP in the same market environment.

3.2. Sensitivity Analysis

The sensitivity coefficient was used to evaluate the degree of influence of the main influencing factors on the key economic indicator. It was equal to the magnitude of change in the key economic indicator divided by the magnitude of change in the influencing factor. Therefore, this study conducted a sensitivity analysis of the main influencing factors in the equation, based on the binomial equations in Section 3.1, to determine the extent to which different influencing factors affect the NPV.

Table 8. The sensitivity coefficients for three ISPs.

Influence Factors	Objective Indicator (NPV)
	C-ISP	ISP-OBF	ISP-OBF-CCS
Ps/CNY·t−1	94.91	21.44	25.00
Po/CNY·t−1	−32.11	−7.29	−8.50
Ic/CNY	−31.67	−7.54	−8.80
Pco/CNY·t−1	−18.51	−3.16	−3.69
Pca/CNY·t−1	−0.17	−0.04	−0.03

shows the results of the impact of the five influencing factors on the NPV in the three ISPs. The results showed that the hot rolled strip price (P_s) had the highest sensitivity coefficient. This was because the iron and steel production process is a raw material, fuel intensive, and relatively single product industrial process. Therefore, as the main product and main source of income of an ISP, the price of a hot rolled strip was the most important factor affecting the economy of an ISP. The sensitivity of each influencing factor to the NPV was ranked in the C-ISP as P_s > P_o > I_c > P_co >> P_ca. However, in the ISP-OBF or ISP-OBF-CCS, the sensitivity of the influencing factor to the NPV was ranked as P_s > I_c > P_o > P_co >> P_ca. These results were basically in agreement with the average cost components of Chinese steel enterprises [21]. Compared to the values of the C-ISP process, the I_c sensitivity values were higher in the ISP-OBF or ISP-OBF-CCS process. This indicates that the construction cost of the ISP-OBF or ISP-OBF-CCS process has a more significant impact on the NPV. Therefore, reducing the retrofitting costs of a TGR-OBF is key to its commercial success. The impact of the carbon price on the NPV was the lowest among the three ISPs. This is due to the current carbon trading policy being moderate for Chinese steel companies. The coking coal price (P_co) in the table had a significant impact on the NPV, while the commercialization of a TGR-OBF had the benefit of reducing coking coal consumption. Therefore, the commercialization of a TGR-OBF will contribute to the profitability of an ISP.

3.3. Uncertainty Analysis

Based on the three fitted equations, we used the Monte Carlo simulation of the JMP platform to analyze the influence of the main influencing factors of the three ISPs on the NPV in varying market conditions. The market was divided into nine scenarios: I, II, III, IV, V, VI, VII, VIII, and IX, in which Scenarios I, II, and III were downturn market situations; Scenarios IV, V, and VI were normal market situations; and Scenarios VII, VIII, and IX were overheated market situations, as shown in

Table 9. Uncertain factors distribution of three ISPs in different scenarios.

Scenario	Pca/CNY·t−1	Po/CNY·t−1	Pco/CNY·t−1	Ps/CNY·t−1	Ic/Bn CNY
					C-ISP	ISP-OBF or ISP-OBF-CCS
I	[1, 67]	[400, 667]	[1100, 1267]	[3500, 3833]	[18, 19, 20]	[18.5, 19.5, 20.5]
II	[67, 134]	[400, 667]	[1100, 1267]	[3500, 3833]	[18, 19, 20]	[18.5, 19.5, 20.5]
III	[134, 200]	[400, 667]	[1100, 1267]	[3500, 3833]	[18, 19, 20]	[18.5, 19.5, 20.5]
IV	[1, 67]	[667, 833]	[1267, 1433]	[3833, 4166]	[18, 19, 20]	[18.5, 19.5, 20.5]
V	[67, 134]	[667, 833]	[1267, 1433]	[3833, 4166]	[18, 19, 20]	[18.5, 19.5, 20.5]
VI	[134, 200]	[667, 833]	[1267, 1433]	[3833, 4166]	[18, 19, 20]	[18.5, 19.5, 20.5]
VII	[1, 67]	[833, 900]	[1433, 1600]	[4166, 4500]	[18, 19, 20]	[18.5, 19.5, 20.5]
VIII	[67, 134]	[833, 900]	[1433, 1600]	[4166, 4500]	[18, 19, 20]	[18.5, 19.5, 20.5]
IX	[134, 200]	[833, 900]	[1433, 1600]	[4166, 4500]	[18, 19, 20]	[18.5, 19.5, 20.5]

. As mentioned in Section 2.4, the construction investment had a triangular distribution, while the others had uniform distributions. In a downturn market, iron ore prices, hot rolled strip prices, and coking coal prices are usually affected by a depressed market, so they were all set at low levels here. Carbon prices may be subject to uncertainty due to policy regulations, so in a downturn market, we divided carbon prices into three scenarios (Scenarios I, II, and III). Construction investment, on the other hand, is hardly affected by the raw fuel market, so we considered it to be triangularly distributed within its level. In a normal market and an overheated market, the iron ore prices, hot rolled strip prices, and coking coal prices were all set at normal levels and high levels, respectively. The detailed scenarios and data are shown in Table 9.

The results of the Monte Carlo simulation are shown in Figure 4, which show the mean values and confidence interval under a confidence level of 95% (95% CI) for the NPV of the three ISPs under the nine scenarios. Detailed statistical data on the results of the Monte Carlo simulation are presented in Tables S6–S8 in the Supplementary Materials.

From Figure 4, the mean value of the NPV for the three ISPs in a downturn, regardless of scenario descending order, is ISP-OBF > ISP-OBF-CCS >> C-ISP > 0; the NPV of the three ISPs in the normal or overheated markets was ordered as 0>ISP-OBF>ISP-OBF-CCS>>C-ISP. This indicates that the ISP-OBF was the most profitable among the three ISPs, followed by the ISP-OBF-CCS, and the least profitable was the C-ISP. At the same time, under Scenarios IV-IX, all three ISPs were in a loss-making situation, and the least loss-making one was the ISP-OBF. This was because the steel industry is characterized by a long payback period with large investments. The models in this study were based on a 15-year calculation period, so a steel company should maximize equipment life if it wants to be profitable. In Scenarios I-IX, the ISP-OBF as the most profitable steel production route is at least CNY 0.392 Bn and CNY 1.934 Bn more profitable than the ISP-OBF-CCS and C-ISP respectively. This suggests that it would not be cost beneficial to apply CCS technology to the ISP-OBF in the nine scenarios. This was because the CCS technology transported and sequestered all the CO₂ emitted by the ISP. However, in the case of carbon trading, the Chinese government issued a certain amount of carbon credits based on the average carbon emissions of the ISP, and the ISP only had to buy CO₂ in excess of that amount, which was relatively small. Therefore, CCS technology was not an advantage in this study’s scenario. In addition, as the carbon price (P_ca) increased, the NPV values for the three ISPs decreased. Particularly, the change in carbon price had the largest impact on the economics of the C-ISP, as it had the largest carbon emissions. The ISP-OBF-CCS, in contrast, had the lowest carbon emissions due to the use of CCS, so the carbon price had the least impact on its economy.

4. Conclusions

In this study, we compared the economy of a TGR-OBF and BF from a whole-plant perspective by replacing the blast furnace in a C-ISP with the TGR-OBF model established in our previous study to form an ISP-OBF or ISP-OBF-CCS. When comparing economics, firstly, three integral steel plants were defined: C-ISP, ISP-OBF, and ISP-OBF-CCS. Then, we fitted three binary regression models in relation to the key economic indicator with the main influencing factors based on the JMP software platform using the BBD method. Finally, sensitivity and uncertainty analyses based on the Monte Carlo simulation technique were performed on these models. The study results indicated that the ISP-OBF was the most profitable among the three ISPs, followed by the ISP-OBF-CCS; the least profitable was the C-ISP. In Scenarios I-IX, the ISP-OBF, as the most profitable steel production route, is at least CNY 0.392 Bn and CNY 1.934 Bn more profitable than the ISP-OBF-CCS and C-ISP, respectively. The carbon price affected the NPV of an ISP. The higher the carbon price, the lower the NPV. However, under the current Chinese carbon trading policy and carbon transportation and storage costs, CCS technology did not make the company profitable.