Steel and cement production, crucial for infrastructure and daily life, accounts for about 15% of global CO₂ emissions. Decarbonizing these industries by mid-century is vital for climate stability. Existing decarbonization scenarios heavily rely on large-scale deployment of supply-side technologies like carbon capture, utilization, and storage (CCUS) and hydrogen-based production (e.g., direct reduced iron (DRI) using green hydrogen). However, these strategies depend on significant infrastructure build-out (CO₂ transport and storage, renewable electricity, green hydrogen production), introducing considerable uncertainty. This study investigates the feasible supply of steel and cement within Paris-aligned carbon budgets, explicitly incorporating this infrastructural uncertainty. The historical data reveals two key challenges. First, the preparation period for deploying CCUS and green hydrogen production takes decades. Second, even with rapid growth, the green hydrogen supply may not reach desired levels by 2050. Current literature underestimates the risk of deployment failure, prompting this study to quantify global steel and cement supply under these conditions using a stochastic framework, accounting for mass balancing, rather than large-scale economic models. This approach allows explicit linkage between infrastructure deployment and steel/cement decarbonization, with transparent assumptions.

Existing decarbonization scenarios for steel and cement emphasize rapid deployment of supply-side technologies, particularly CCUS and hydrogen-based solutions. The International Energy Agency (IEA) projects significant emission reductions from CCUS. Academic literature also emphasizes CCUS and hydrogen technologies, highlighting the potential economic competitiveness of hydrogen-based steel under favorable conditions. However, a common challenge is the reliance on extensive infrastructure build-out, raising concerns about the feasibility and speed of deployment. While some studies acknowledge the long timelines for infrastructure development, the risk of deployment failures is often underemphasized. This study addresses this gap by explicitly modeling infrastructure deployment uncertainty within a carbon-constrained framework.

The study uses a stochastic optimization model based on physical mass balancing equations. The model maximizes global steel and cement production within Paris-compliant carbon budgets (1.5°C and well-below 2°C scenarios), considering zero-emissions infrastructure deployment (CCUS and non-emitting electricity) and scrap availability. The model incorporates uncertainty in infrastructure deployment by using historical data and IEA scenarios to define ranges for CCUS capacity and non-emitting electricity supply. A Monte Carlo simulation, employing uniform distributions for uncertain variables, runs the optimization 1,000 times to determine uncertainty ranges for steel and cement supply. Technological progress is assumed to align with established paradigms and industry best practices. The model accounts for various production processes (blast furnace, DRI-EAF using fossil fuels or hydrogen, and scrap-EAF) for steel and considers clinker substitution and waste material use for cement. Expected demand is compared against three IEA scenarios: Baseline, Net Zero, and Low Energy Demand scenarios.

The analysis reveals a significant shortfall in feasible steel and cement supply under Paris-compliant carbon budgets. Even with substantial technological advancements, feasible steel supply is estimated to meet only 55–85% (interquartile range) of baseline demand in 2050, while cement supply is projected to meet only 22–56%. The study highlights that current CCUS capacity falls far short of previous IEA projections. The required expansion rate for 2050 CCUS capacity is far greater than current construction plans. While hydrogen-based and recycling-based steel production show growth potential, they are limited by infrastructure availability and scrap limitations. The model estimates that these two routes could provide more than 60% of current steel supply in 2050. Although the minimum requirements to meet basic human needs are significantly lower than the feasible supply, fulfilling the demands of high-income countries and ensuring equitable distribution poses a challenge. Currently, high-income countries have far exceeded feasible per capita material stocks, while low-income countries fall below. Allocating feasible supply based on current usage shares reveals that construction and manufacturing sectors may only meet approximately 40% and 60% of expected baseline demand, respectively. Figure 2 demonstrates that the feasible supply falls drastically short of the expected demand for both cement and steel under a 1.5°C carbon budget. Figure 3 shows the breakdown of steel production by different processes under a 1.5°C budget. Figure 4 and 5 show comparisons of cumulative feasible supply versus minimum needs and feasible per capita stock versus current levels, respectively, highlighting the disparity in material consumption across income groups. Figure 6 shows the incompatibility between feasible supply and expected demand across construction and manufacturing sectors.

The findings emphasize the inherent uncertainty in relying solely on supply-side decarbonization strategies. The limited feasible supply necessitates a twofold approach: governments must accelerate infrastructure development, while industries must prepare for potential deployment failures. This could involve strategic investments in decarbonized production, shifting from selling materials as commodities to offering services, and improving material resource efficiency. The construction and manufacturing industries need to adapt to lower material availability through design changes, optimized material utilization, and extended product lifespans. Significant material reductions are possible (30–70% in construction and 35–65% in manufacturing), requiring both behavioral and technological changes. The study advocates for a 'forecasting supply and backcasting demand' approach instead of the traditional reverse, prioritizing understanding feasible supply and then optimizing demand-side actions. This method highlights the urgency of demand-side actions and quantifies the necessary level of resource efficiency to stay within carbon budget constraints.

This study demonstrates that relying solely on supply-side decarbonization for steel and cement is insufficient given the significant uncertainties in infrastructure deployment. Feasible supply, even with technological advancements, falls drastically short of expected demand. Therefore, governments need to expedite infrastructure development, and industries must prioritize resource efficiency and prepare for potential infrastructure deployment failures. A holistic approach focusing on both supply-side advancements and robust demand-side strategies is crucial to meet climate goals without jeopardizing essential services.

The model simplifies some aspects, such as technological progress and resource availability, using linear projections and established paradigms. The assumption of uniform distributions for uncertain variables may not perfectly capture the true distribution of infrastructure deployment. The analysis focuses on global scales and may not reflect regional variations in infrastructure deployment or resource availability. Finally, the model primarily focuses on physical mass balancing and does not explicitly incorporate economic factors that could influence the feasibility of different production pathways.