1. Introduction
The impact of human-induced climate change is evident in the 1.1 °C increase in global surface temperatures, with industrial emissions of carbon dioxide (CO
2) being a significant contributor [
1]. To reverse this trend, it is necessary to take actions aimed at reducing CO
2 emissions in the future.
The steel industry is one of the main contributors to CO
2 emissions, accounting for approximately 27% of global industrial CO
2 emissions [
2] and 5–7% of the total CO
2 emissions worldwide [
3,
4]. Its substantial environmental impact underscores the pressing need for effective mitigation strategies within this sector. Previous research has highlighted numerous strategies for mitigating or eliminating CO
2 emissions in existing industrial [
5,
6] and specific steelmaking processes [
4,
7,
8,
9].
One promising pathway in the steel industry involves the replacement of fossil coal, traditionally used as the reducing agent in iron ore reduction, with renewable energy sources such as hydrogen and/or electricity [
7]. This strategy, known as carbon direct avoidance (CDA), seeks to diminish the initial reliance on carbon as an energy carrier [
10,
11]. However, it is noteworthy that a drawback of CDA lies in its inapplicability for retrofitting existing plant sites adhering to the conventional blast furnace steelmaking route, coupled with the substantial quantities of renewable energy required for implementation. Another promising pathway extends downstream from the production plant and involves the implementation of the carbon capture and storage (CCS) strategy [
12]. This strategy focuses on capturing carbon emissions post-production, followed by their secure storage [
13]. The captured CO
2 is transported through dedicated pipelines or ships [
13], subsequently sequestered within geologically deep formations, including but not limited to, deep geological formations, saline aquifers and previously depleted oil and gas reservoirs [
12,
13,
14,
15]. CCS faces challenges related to effective carbon storage in underground geological formations, primarily due to the need for substantial local infrastructure and secure underground storage facilities, which are essential to accommodate the significant quantities of CO
2 involved in the industry [
14,
15]. In this paper, we investigated another promising pathway for reusing carbon and other components (like nitrogen) from exhaust gases. In the literature, the strategy for utilizing carbon is referred to as carbon capture and utilization (CCU) [
12,
14,
16] or as integrated carbon capture and conversion (ICCC) [
17,
18]. Both strategies center around the utilization of CO
2 as a feedstock. Typical applications of these strategies include the production of chemicals, syngas or fuels, as well as mineral carbonation, enhanced oil recovery and direct utilization, such as in the food industry [
16,
17]. CCU focuses on the reuse of CO
2 in specific applications, whereas ICCC adopts a more comprehensive approach that integrates CO
2 utilization across various sectors, aiming to derive economic benefits from the conversion. However, in this paper, we specifically explore the CCU concept for chemical production using steel mill exhaust gases, including the production of gasoline, polymers and fertilizers [
19].
The exhaust gases originating from blast furnaces, coke ovens and basic oxygen furnaces contain significant quantities of carbon monoxide, carbon dioxide, methane, hydrogen and nitrogen. These gases serve as primary raw materials for chemical production. Hydrogen (H
2) is recognized as the limiting reactant in this process, requiring the introduction of external supplements, such as renewable energy (RE), to increase its quantity [
9]. Considering these factors, it is imperative to establish a symbiotic relationship between the steel industry and the energy sector, recognizing that such collaboration is essential for assessing the hydrogen supply limitations identified in this process.
Numerous recent studies have primarily emphasized the techno-economical aspects of CCU approaches within the steel industry [
4,
20,
21]. Commonly employed technologies for the extraction of CO
2 from these exhaust gases and their subsequent utilization in the CCU approach include chemical or physical absorption and membrane processes [
22,
23,
24,
25]. Typical chemical products that can be synthesized from steel mill exhaust gases are methanol [
21,
26,
27,
28], higher alcohols [
29,
30] or urea [
31,
32,
33].
To evaluate the long-term feasibility and determine the optimal configuration of the CCU concept, the application of mixed-integer linear programming (MILP) is a commonly employed scientific methodology, as demonstrated by prior research [
8,
21,
34]. For this reason, a model-based mathematical approach with MILP is recommended to evaluate the long-term feasibility of the system [
35]. The objective of the MILP model is to find the optimal production pathway, which includes selection of technologies, design and time-dependent operation conditions. However, it is important to note that these models face challenges in acquiring comprehensive information necessary for robust decision-making processes concerning the plant network (
Section 2). MILP models are particularly suited for addressing uncertainties in future conditions, requiring the construction of hypothetical scenarios to depict potential future states. In these scenarios, a spectrum of future developments related to key factors in the political, economic, social, technological, ecological and legal (short: PESTEL) domains can be taken into account [
36,
37]. To find and select the key factors for this system and develop these specific scenarios in a transparent and consistent way, there are several scenario approaches discussed in the literature [
38,
39,
40,
41]. Kosow et al. [
38] identify three primary approaches in this field: trend extrapolation, which overlooks interactions between key factors; systematic formalized approaches, which consider interactions between key factors; and creative narrative approaches, which emphasize intuitive projections of key factors. Scenarios in the literature can vary in their definitions and development methods, making it essential to establish a clear definition at the outset.
In this study, a scenario is defined as a specific part of the future by considering relevant key factors rather than a comprehensive picture. Integrating individual factors shapes the domain of shared development for all these aspects. The expanding slice of future developments of these key factors is described with the scenario funnel in
Figure 1.
Different scenarios in
Figure 1 portray diverse future possibilities right from the beginning. To achieve this, potential projections of various key factors (a, b and c) are chosen and bundled into scenarios. It is important to clarify that the term ‘projection’ should not be conflated with a ‘forecast’, which asserts the actual probability of occurrence. Instead, it represents a hypothetical construct, implicitly alluding to the potential for alternative futures [
38].
Some approaches involved the constructions of scenarios to evaluate CCU concepts of steel mill exhaust gases. Stießel et al. [
43] have crafted a singular scenario, designated for the year 2030, with the principal objective of identifying cross-industrial process concepts for a CCU approach. The authors have predominantly concentrated on external influences to shape this scenario, placing an emphasis on formulating a CCU concept that thrives in specific eco-friendly operational conditions. Within this context, Schlüter et al. [
31] have undertaken an investigation into the process concept of utilizing exhaust gases from steel mills for chemical production. This study explores three distinct operating conditions, with a primary focus on internal technical perspectives. The results are subjected to analysis within the framework of time-dependent boundary conditions, thus enabling the identification of the factors that limit carbon binding. Furthermore, Sadlowski et al. [
44] delve into the ecological potential of flexible methanol production utilizing steel mill exhaust gases through the application of an MILP model. In their research, they have outlined scenarios underpinned by three pivotal factors: external hydrogen production capacity, power supply sources and storage capacities. Subsequently, the outcomes are scrutinized with a specific focus on the carbon binding potential inherent in this CCU concept. The authors have collectively emphasized the pivotal role that diverse future developments of key factors, such as hydrogen or power prices, have in the evaluation of CCU concepts. This underscores the compelling need for a thorough investigation aimed at establishing consistent scenarios.
Therefore, this study examined scenarios for a particular CCU concept targeting the production of methanol, acetic acid, ammonia and urea from steel mill exhaust gases. This CCU concept and its interaction with future scenarios has not been analyzed in previous research. Our primary goal is to develop different, transparent and consistent scenarios which are suitable as input data for the evaluation using an MILP model. What sets this approach apart from recent publications is its departure from singular perspectives or single scenarios in the assessment of CCU concepts. Instead, a novel approach is introduced, characterized by a comprehensive evaluation spanning various domains of key factors. This innovative scenario development approach distinguishes this work from previous studies, offering a wider and more intricate array of future scenarios. In essence, the scenarios generated here serve as valuable tools for evaluating the long-term feasibility of the CCU concept using an MILP model, making a distinct and noteworthy contribution to the field.
The structure of this work is as follows.
Section 2 outlines the key attributes to enable the optimization with an MILP model. In
Section 3, we present the comprehensive methodology employed throughout the scenario development process.
Section 4 presents the findings and outcomes of our scenario development efforts. The study concludes in
Section 5, followed by
Section 6, which offers a forward-looking perspective.
2. Scope and Characteristics of Scenarios in an MILP Model
The implications of the CCU concept are currently in a less developed phase, with ongoing determination of the technical plant layout, profitability and potential environmental impacts. As a foundational assumption, it is posited that all interconnections for the various technologies and exhaust gas conditions, such as reactors, storages, compressors, separators, etc., have been integrated into the MILP model, as demonstrated in prior research. The selection of the final products is also established at the decision point [
35]. A comprehensive model description for methanol production can be located in Sadlowski et al. [
44]. The model’s output includes the optimal pathway, encompassing the selection of technology, chemical products, plant capacities and operational strategies. This outcome is contingent upon the provided future states, represented in the form of scenarios.
The scenarios generated in this study enable MILP and serve as input data for a linear optimization model, featuring distinct properties that set them apart from conventional scenario development.
First, the study formulates scenarios based on evaluations across key factors in multiple domains (PESTEL). Considering diverse internal and external influences brings reliable results about the optimal future CCU concept. Secondly, a fixed reference year for the scenarios is established, including the entire operational life span of the CCU concept. However, owing to the intricacy of the MILP model, a reduction in one reference year for evaluating becomes necessary. Thirdly, all considered key factors must exhibit clear quantifiability, either through fixed numerical values or linear mathematical relationships, to ensure the quality of the deterministic optimization results. The exclusion of the challenging-to-quantify political, regulatory and social domains from the PESTEL approach aligns with the premise that unforeseen disturbances during the plant’s life span are to be minimized. It is important to note, however, that complete elimination of overlaps between individual key factors and other domains may not be entirely feasible. For instance, certain factors, such as a CO2 emissions limit, may have direct ecological implications on the concept despite their political/regulatory origin. Lastly, a software-based method examines the individual relationships between these factors. A Cross-Impact Analysis (CIA) is employed to identify the most consistent and plausible combinations of factors for alternative scenarios.
In essence, these unique scenario properties facilitate their utilization within the linear optimization model, ultimately enhancing the reliability and applicability of our approach.
3. Scenario Development Framework and Process
An existing framework for scenario development has been customized to incorporate functionalities specific to the scenarios. This adaptation draws inspiration from a five-phase model introduced by an approach from Gausemeier et al. [
39] and an eight-phase model from von Reibnitz [
40]. From these influences, a modified generic six-phase model for iterative scenario development is derived, tailored to the specific application of the CCU concept for steel mill exhaust gases.
Figure 2 provides an overview of the individual six phases of the modeling framework.
This adapted modeling framework, incorporating both exploratory and quantitative approaches, has been devised to generate the five scenarios within the technological, economic and ecological domains. In the subsequent sections, we will apply the six phases illustrated in
Figure 2 to the specific CCU concept.
3.1. Premise for the Scenarios
Scenario-specific assumptions for further considerations are defined within the premise. The definitions are supplemented by the boundary conditions to form a basis for the scenario development.
Table 1 shows a short description of the determined premise.
The time horizon of the CCU system is estimated as 25 years, with five years of construction and a 20-year life span—operating from 2030 to 2050. The target year is 2040, the middle of its life span. The maximum generation is restricted under German competition law prohibiting market dominance [
46]. The market share-based presumption provides a first indication of dominance where a company’s market share exceeds 40 percent [
45]. Therefore, the maximum chemical product quantity is 40 percent of its market volume. The overall market assumption in scenario development is based on the system boundaries of and the cross-border trade with European neighbors. The profitability assessment requires the revision of the future cash flows to be compared with the current capital value. The discount rate is assumed to be constant at 2% annually during the whole life span. The technical parameters like possible plant connections, efficiencies, reaction conditions, exhaust gas amounts, etc. are given from previous studies, project work and own calculations.
3.2. Key Factor Selection
First, the domains of influence are determined. For such environmental and energy scenario development, the PESTEL domains are suggested in practice [
36,
37,
39]. However, the social and political influences are excluded from our scenario boundary. The CCU concept can be sensitively affected by adverse social acceptance, which may give policymakers a false sense of security, even leading to a rebound effect [
47]. Nevertheless, they are not suitable for our scenario’s target. First, the factors from these aspects are often measured in a qualitative approach. For example, political inclination may function importantly in evaluating the feasibility but is formulated qualitatively rather than as exact values (e.g., left and right orientation). Secondly, the issues depend on subjective assessment. For example, the social acceptance and benefit of CCU concepts can be understood in totally different way. Based on these reasons, the technical, economic and ecological domains are determined as the investigation fields’ demarcation.
In a first step, 106 internal and external influencing factors from the domains are determined. The importance of the influencing factors is identified through influence analysis. A detailed explanation of the method can be found in [
38]. The influence analysis examines the relationships between the factors. All possible pairs of factors are measured on a four-level scale from 0 (no effect) to 3 (strong effect) regarding their mutual impact [
38,
39]. The factors with a high active sum are selected as key factors out of the 106 influencing factors since they have the biggest influence out of the overall system [
40]. This procedure leads to the final 24 key factors to set up further scenario development.
3.3. Reference Scenario Development
The reference scenario, often referred to as the “trend scenario”, operates on the assumption that no new measurements will be made by the target year [
38,
39]. The chosen scenario technique for this purpose is trend extrapolation [
38]. This scenario is referred to as “Business as usual” (
BAU). The current values are derived from well-established and reliable knowledge. Extrapolations in the target years (2040 for operating time and 2025 for investing time) are suggested by a meta-study of different reports about energy and chemical market development scenarios. If reliable development is unavailable, the assumptions are made through trend analysis. It is carried out by collecting historical data for as long as possible, and past trends are extended to the future [
38].
The time series of the electricity price in 2040 are derived from a forecasting model [
48]. This model assumes an energy-only market and calculates the operating plans of the power generation systems. The projected time series of the carbon footprint
CFt for future energy production and
REt share for 2040 is determined based on the
CFt in 2020 derived from historical data from AGORA [
49]. The future
CFt is calculated based on the hourly data from 2020 and the varied fraction of
RE share in the German grid mix according to Equation (1):
The exemplary results of the projected time series of the
CFt are shown in
Figure 3.
Figure 3a shows the dimensionless sorted annual
CFt of the year 2020 (
RE = 48%) and the projection to reference scenario of 2040 (
RE = 85%).
Figure 3b shows an example of a two-week period of the
CFt. The average value for the reference scenario is projected to ca. 110 g
CO₂/kWh
el. Electricity prices are calculated in a similar way.
It is necessary to specify fuel prices for natural gas and coal, as well as the CO
2 certificate prices to determine the marginal cost. The prices in the target year are taken from the European Commission [
50] and Bloomberg [
51]. H
2 price plays an important role in defining the potential of
RE and green electricity. The price is derived from the International Energy Agency (IEA) [
52].
The market prices of the chemicals are the biggest part of the revenue. Oxygen as a by-product from water electrolysis is also considered a part of revenue. The prices are assumed by the trend analysis based on the historical data from 2019 to 2021. Plus, the chemicals’ market volumes dramatically affect the size of the plant and expenditure as the maximum generation is regulated by 40% of the market share limit. They are taken from the IEA [
53].
The CO2 emission allowance characterizes the emissions from the steel mill’s exhaust gases. At 100%, it represents the endorsement of the existing state, wherein exhaust gas combustion and CO2 emissions in the atmosphere are accepted without alteration (no Cap). A Cap, such as of 90%, mandates a 10% emissions reduction. In the BAU scenario, no adjustments are presumed. However, a reduction in the allowance signals a compulsory decrease in CO2 emissions, thereby enhancing the CCU potential of the concept.
Table 2 shows the data of the most probable
BAU scenario. Data from the technology domain are shown as relative value (1 = no changes) and are derived from project internal communications and plant development reports. The H
2 generation is a crucial aspect of the system. Therefore, H
2 efficiency, the electricity required to generate the external carbon-free H
2, is considered as a separate factor from the overall energy requirement.
The target year of the capital expenditure (CAPEX) is set as 2025 following a five-year construction period. The CAPEX is calculated for each plant, including the gas conditioning, external H
2 production and chemical synthesis plants. The calculation is based on the capacity method [
63]. The CAPEX of a plant
Cb and its capacity
Sb is estimated based on the reference CAPEX
Ca and its capacity
Sa [
63,
64]. The reference data are taken from various techno-economic analysis studies and the
Ca is extrapolated to the target year of investing, 2025. It is extrapolated to the 2025 value by applying the chemical engineering plant cost index (CEPCI), as
i, to account for the inflation rate. The publication years of the studies are between 2006 and 2021. The original CAPEX
C0, capacity
Sa and CEPCI
i0 for all technical plants and years are used or derived from these studies. The CEPCI value for the year 2025
ia is determined through trend analysis from the last five years. Therefore, the reference CAPEX
Ca is calculated with Equation (2):
The CAPEX development of the hydrogen production plants is assumed to be lower in the future. According to [
65], it is assumed that the CAPEX for alkaline (ALK) and proton-exchange membrane (PEM) electrolysis will be reduced by 14% and 22.5% in next five years, caused by reduced manufacturing costs and assumed technological breakthroughs. Based on the updated
Ca to 2025, the
Cb is calculated via the capacity method in Equation (3), where
f is the degression coefficient for the economy of scale for chemical plants with a value from 0.6 to 1.0.
However, the final
Sb of the plant is not determined in the scenario development process. Therefore, the
Cb of the individual component is represented as the function within the possible installed capacity range of
Sbmin and
Sbmax. The
Cbmax of
Sbmax is where the exhausted gas utilization is maximized based on the market restriction. It should be noted that the
Sbmax of each plant is differently estimated depending on the final products due to the varied size of the market volume.
Cbmin is assumed to be 10% of
Cbmax. If there is lower than 10% of
Cbmax, it is not worth installing these plants because a significant emission reduction is required for the CCU concept.
Table 3 shows the range of
Sb and
Cb of a water gas shift (WGS) plant for each chemical as an example.
The Cb should be represented in a full range of the plants Sb regardless of the production route. In the case of the WGS plant, then the CAPEX is resulted in the function within the overall range of Sb from 2.33 to 376 kg/s.
If
f of the component is less than 1, like the WGS plant (0.82), the
Cb is a root-function. The MILP model requires linearity of
Cb and therefore should be revised as a linear approximated function
Cb,lin. The linearized functions’ maximum relative error tolerance from the original
Cb is set as 10%. If this does not match, an additional sampling point for piecewise linear approximation is considered until it reaches the <10% criteria. In the case of the WGS plant, two linear functions are generated with one piecewise sampling point and a maximum relative error of 7.4%.
Figure 4 presents the original
Cb function on the left side and the derived piecewise linear functions
Cb,lin1 and
Cb,lin2 on the right for the WGS plant.
Other plants’ CAPEX is calculated in a similar way. The maximum relative error is detected in ammonia synthesis plant as 9.4%. The range of CAPEX for each plant, regardless of the production pathway, is generated.
The operating cost (OPEX) is the expenditures incurred in the plant. It considers the variable, fixed and other costs of the system. The variable costs, including the raw and auxiliary materials, are calculated differently depending on the operating time and final production pathway. The fixed and other costs are structured based on [
66]. The projected OPEX in the
BAU scenario is estimated to be constant by the target year.
3.4. Future Projection of Key Factors
Each key factor is projected into the future in alternative states. Qualitative projections are created at first. This includes the possible projections of highly decreasing (↓↓), moderately decreasing (↓), constant (-), moderately increasing (↑) and highly increasing (↑↑). Not all projections make sense (e.g., decreasing projection of CO2 certificate price), so the number of varied projections differs between three and five depending on the key factor. As mentioned, each projection involves quantified numerical values. If the data from the BAU scenario are available, the projection is based on it. The variation rate from the current value to fixed projection of the BAU scenario is applied to other alternative projections identically.
It should be noted that some key factors involve deliberately exaggerated or passive quantification. The factors that contain exaggerated quantification are the “driving factor”. The extreme value of these driving factors brings a clear difference from other projections. On the other hand, the passive quantification is for the case that the value from the BAU scenario is over-predicted. The scenario which involves the projection may cause discord with other elements. Thus, they are quantified at a lower variation rate. Passive quantification makes the combination of the factors more consistent.
If the data from the reference scenario are unavailable, they are quantified based on the independently estimated assumption. For example, the H
2 efficiency has an improving rate of 5% for projection (↑) and 10% for (↑↑), according to [
52]. The factors, conversion efficiency and energy efficiency are assumed to be identical in improving rates of H
2 efficiency. It is not plausible to assume that they have greater improvement than external H
2 supplements because these plants are at a state-of-the-art level.
The investment cost is projected through the independently generated method. A scaling factor, “
s-factor”, is applied to the generated
Cb,lin function of each plant to switch the range of the CAPEX by multiplying itself. The
s-factor is derived from the water electrolysis’ CAPEX as it is available to obtain reliable data on future development. Plus, it can be compared with the current CAPEX as the
f is equal to 1—it is not affected by the varied size of the capacity.
Table 4 presents the assumed CAPEX of ALK and PEM in diverse future situations.
In the
BAU scenario, the specific CAPEX for ALK and PEM is estimated to decrease to 86% or 77.5% by the target year, respectively [
28]. In a pessimistic future, the CAPEX is assumed to be constant as the current value. A regular projection assumes a 50% higher decreasing rate of the CAPEX than the
BAU scenario. In the optimistic situation, the decreasing rate is doubled by the
BAU scenario. The variation rates of the future CAPEX from the
BAU CAPEX are the
s-factor. They are applied to all considered plants of the CCU concept depending on the scenario concept and the result of a cross-impact analysis. Through the process, the range of the component’s CAPEX is newly assigned for each scenario. Another external H
2 supplement option, methane pyrolysis (MP), is applied as an identical
s-factor with the ALK.
3.5. Scenario Formation
Based on the projections of key factors, the actual formation of scenarios takes place. The systematic formalized scenario technique of cross-impact analysis (CIA) is applied to ensure consistent combinations [
38,
39]. The CIA analyses the relationships between the key factors and the probabilities of the occurrence of future events by considering their direct and indirect mutual effects [
67].
A cross-impact matrix is first created, which assesses the conditional probability of specific projections if another future event has occurred according to the seven-level scale from −3 (Strong inhibitory influence) to 3 (Strong promoting influence) [
30]. After that, the concept of each scenario is developed. The scenario concepts focus on the state of the specific domain to be improved or regressed or the worst or the best operating situations. Based on the concept of the scenarios, the corresponding factors are fixed in a particular projection to fulfil the determined idea. Four different scenarios, excluding the reference scenario, are created. A brief explanation of the different scenario concept and targets is shown in
Section 3.5.1,
Section 3.5.2,
Section 3.5.3 and
Section 3.5.4.
3.5.1. CO2 Reduction and RE Share Target (RE-Boom)
The
RE-Boom depicts the best condition from the ecological perspective.
Table 5 presents seven key factors which are forced to demonstrate the scenario.
The key factors, CO2 certificate price and CF and RE share, are forced environmentally friendly. The chemicals’ market volume is fixed to be highly increased to remove the market restriction for more possible CCU production. The CO2 price is defined to be decreased to reduce the availability of direct CO2 sales options.
3.5.2. Technical Improvement and Market Booming (Market-Boom)
The scenario
Market-Boom set the perfect condition from the economic and technical perspectives.
Table 6 shows the eleven fixed key factors to fulfil the scenario concept.
The key factors related to the revenue are all fixed to be highly increased to maximize the profits. The factors in the technology class are defined to be highly advanced. However, the H
2 efficiency is estimated to be moderately increased to make a clear difference with the
H2-Boom scenario in
Section 3.5.4.
3.5.3. Energy and Market Crisis (Crisis)
The
crisis scenario projects the worst situation from the economic perspective. The concept refers to the current Ukraine conflict with energy and economic crisis.
Table 7 presents eight forced factors for the scenario concept.
The factors related to the profitability are all negatively assumed. The prices of raw materials go up immensely, and the revenue of the products is reduced substantially. Regarding the product’s market condition, it is evaluated from the perspective of the supplier. In other words, it is assumed that the chemicals market is in depression, so the supplier must sell the product at a lower price.
3.5.4. Hydrogen booming (H2-Boom)
The
H2-Boom focuses only on the best condition of H
2 generation from the technical aspects.
Table 8 shows six essential key factors to satisfy the scenario concept.
The H2 price is assumed to be reduced following a drop in generation cost due to highly increasing manufacturing costs and technology breakthroughs. It is aimed at the hydrogen subdomain from the overall technical domain.
3.6. Scenario Generation and Selection
Based on the formulated cross-impact matrix and the scenario concepts, the CIA is conducted to determine the most consistent combination. It tests all theoretically possible combinations to analyze their contradictions with the framework conditions. However, the generated matrix involves more than a billion possible combinations. The CIA, thus, can only be checked with algorithm-based software support. For this reason, the ScenarioWizard® v4.52 software is used. In addition, the economic key factors pertaining to the prices and market volumes of the investigated chemicals of acetic acid, methanol, ammonia and urea are bundled into a single key factor. This bundling helps streamline the assessment of the interactions and complexity among these key factors. This judgment is based on the observation that the scenario concepts typically do not entail a direct comparison of the superiority among various projections of the chemicals. Bundling does not affect the quantitative values for each factor but merely has an identical qualitative projection. As a result, the bundle containing the possible candidates to be a final scenario is generated for each scenario. In our case, eight options for RE-Boom, five for Market-Boom, two for Crisis and fifteen for H2-Boom are generated.
To select the final and most consistent combination out of the candidate’s pool, the Consistency Value (CV) and Total Impact Score (TIS) function as the evaluation criteria [
30]. The TIS means the sum of the impact scores of all selected scenario variants. The CV is the parameter to evaluate if the chosen combination of the factors is consistent. In the case of a positive or zero CV, the combination is accepted as consistent [
67,
68]. Based on the scenario selection criteria, the final scenarios are determined. All scenarios have a CV equal to 0 and the highest TIS out of the possible candidates, so they involve the most consistent combination.
5. Conclusions
This study aimed to establish coherent, credible and distinct scenarios to serve as input data for MILP models, facilitating the evaluation of the long-term viability of the proposed exhaust gas utilization concept. These five scenarios were created by incorporating both internal and external key factors based on three primary criteria. To construct the Business-As-Usual (BAU) scenario, these factors were projected forward to the target years of investment (2025) and operation (2040). Notably, the formulation of initial expenditure was outlined through the assessment of investment costs. A linear approximation approach for the CAPEX functions was employed to ensure compatibility with future use in MILP models while adhering to a maximum error tolerance of 10% from the original nonlinear function.
All 24 crucial key factors were projected into the target years, combining qualitative and quantitative elements to shape a hypothetical future trajectory. This process adheres to an objective methodology that relies on BAU scenario data rather than making random predictions. It also incorporates deliberate variations, both exaggerated and conservative, to make the evaluation of the CCU approach more apparent. In instances where BAU scenario data were unavailable, quantification was determined based on independently estimated assumptions.
Quantitative projections of investment costs were made possible through the application of a scaling factor (referred to as the “s-factor”). This s-factor, derived from the water electrolysis CAPEX, ranges from 0.71 to 1.29. It was then applied to other facilities to modulate the amplitude of the CAPEX functions, thereby introducing diverse economic scenarios. An s-factor greater than 1 signifies economically unfavorable situations due to higher initial expenditures.
This comprehensive analysis offers novel perspectives on the potential future trends of pivotal key factors, such as electricity prices, the share of renewable energy or the chemicals market trends, as the “Hot potato” key factors.
The concept for each scenario is predetermined prior to the scenario formulation process (
Section 3.5). In essence, scenarios are developed based on deliberately specified conditions, resulting in four distinct scenario concepts:
CO2 Reduction and RE Share Target (RE-Boom): This scenario envisions the most favorable ecological conditions, emphasizing CO2 reduction and a high renewable energy (RE) share.
Technical Improvement and Market Booming (Market-Boom): Here, the focus is on achieving optimal economic and technical conditions, with an emphasis on technical advancements and a thriving chemical market.
Energy and Market Crisis (Crisis): The Crisis scenario represents the most adverse economic situation, depicting a scenario of energy and market crisis.
Hydrogen Booming (H2-Boom): This scenario exclusively highlights the most advantageous conditions for hydrogen generation, considering both technical and economic aspects of the plants and market.
These scenario concepts provide a comprehensive spectrum of potential future conditions, each with its distinctive emphasis and perspective.
Based on the predetermined concept, we conduct a cross-impact analysis (CIA) to identify the most consistent combination of key factors within the established conditions. The computed combinations are regarded as highly reliable outcomes. Firstly, in terms of consistency, all scenarios exhibit the desired consistency value (CV) of zero and the highest total impact score (TIS) among the potential candidates. Secondly, determinations of the remaining factors’ projections through the CIA calculation are also plausible in describing the scenario concept within the scope of the investigation. Lastly, the results yield significantly different operational conditions that are meaningful. These findings build upon existing evidence of the reliably structured scenarios. The generated scenarios can now be directly applied in a linear optimization model to evaluate the prospective CCU concept.
In summary, the five scenarios we have generated provide a clear portrayal of potential situations for the CCU concept in the target years. These scenarios are founded on a complex interplay of factors from various impact parameters. Importantly, each scenario offers a consistent, plausible and notably distinct combination of key factors.