Reassessing Steel: How Falling Cement Use Alters Future Projections

My recent online conversation with Scott Norris, a structural steel expert and Director of Engineering Solutions at Steelcon, prompted me to reexamine some assumptions I had previously made about global steel demand projections. Scott, whose practical experience within the steel industry grants him clear insights into current operational constraints and near-term realities, challenged my earlier views, particularly regarding how steel production will evolve alongside cement use in the coming decades.

Initially, my projections assumed a continued robust, if slower, growth trajectory for global steel demand through mid-century, driven by persistent global infrastructure expansion in developing countries. But now, reflecting on emerging trends within the cement sector and deeper scrutiny of underlying demand drivers, I am increasingly convinced that I have overstated steel’s long-term growth potential.

Cement and steel are closely intertwined, with approximately half of global steel demand tied directly to construction and infrastructure—precisely the sectors driving cement consumption. I’d performed my initial steel research and projection before I made it to cement. In the first half of 2024 I spent weeks digging through cement with all of the alternatives to its use that existed commercially, in policy and in the lab before projecting significantly falling cement demand through the century.

Six months after I’d published my projection, the World Cement Association published a key whitepaper agreeing that global cement demand will soon peak and then steadily decline as we move toward mid-century, driven by a combination of completed urbanization cycles in developed economies and shifting building methods globally.

China’s unprecedented build-out over the past few decades is rapidly drawing to a close, and the likelihood of another country matching China’s massive scale of cement-based infrastructure growth appears exceedingly low. In both steel and cement, China represents half of global demand and production, its population is 20% of the world’s and its economy is among the top three largest, counting Europe as a whole. If its volumes are declining and US and European volumes are flat or declining, then it’s difficult to see global steel demand increasing. India’s economic growth, for instance, is heavily service-driven, and the country’s existing urban infrastructure already limits the potential for a China-like construction boom. Similarly, other developing regions such as Southeast Asia, while certainly expanding, seem unlikely to follow a trajectory anything like the scale of China’s past three decades.

Norris rightly points out that the developing world, India and parts of Southeast Asia specifically, will nonetheless experience growth in steel demand. He notes that India alone is targeting significant growth in steel production, aiming to double its output to around 300 million tonnes by 2030 and potentially reaching 350–500 million tonnes by mid-century. Southeast Asian nations, notably Vietnam and Indonesia, are similarly positioned to experience steel demand increases over the coming decades. Norris’ argument rests primarily on near-term infrastructure commitments: many countries are actively building new blast furnace-based steel plants, which will lock in decades of conventional steel production. These blast furnaces typically operate for 40 to 50 years, meaning facilities constructed today could realistically continue running into the 2080s without policy-driven retirements.

Note that anywhere I ascribe a perspective or opinion to Norris, it is based on my interpretation of his comments and posts of projections that I reviewed, and may not accurately capture his point of view. Any errors are mine.

While Norris’s perspective underscores important practical constraints and near-term realities, I still question whether these projected increases will ultimately manifest at the magnitude currently anticipated. Much hinges on the extent to which the global economy embraces alternative building practices, material efficiency, and new structural technologies.

Cross-laminated timber, for instance, has emerged as a genuinely viable and rapidly growing alternative to conventional steel and concrete structures, and for increasingly tall buildings, with a 50 story tower under construction in the USA. Timber construction offers substantial benefits in material efficiency, requiring significantly less mass per unit of structural strength. This directly translates to reduced steel and concrete use in foundations and structural elements, potentially undercutting traditional steel demand forecasts significantly. Already, timber-framed buildings are gaining traction globally, supported by their substantial embodied carbon advantages over concrete and steel structures.

If 50% of demand for steel is in infrastructure and buildings, China’s demand is declining, new buildings pivot increasingly to CLT, western building is slow with more renovations and the developed world leverages its forestry resources for more CLT rather than building increased blast furnaces, then it’s difficult to project an increase in steel demand. Rather, we might be at peak steel today.

Another important consideration is the question of steel scrap availability. Norris identifies what he terms a “scrap valley”, a projected shortage of recycled steel sufficient to meet electric arc furnace (EAF) demand until around 2045, when he projects a scrap tsunami. He bases that on an excellent approach of considering the median age of scrapping across segments in decades, then applying that production years to arrive at a range of available scrap for the coming years through 2050. He notes that current projections from the World Steel Association anticipate that even by mid-century, scrap-based steel production will account for only half of global steel output, agreeing with his projections, necessitating continued reliance on virgin iron sources such as direct reduced iron (DRI).

My own research, however, points to potentially enormous untapped scrap sources tied up in fossil fuel infrastructure that is rapidly becoming obsolete due to the global energy transition. For example, in the United States alone, fossil fuel pipelines contain enough embedded steel to supply around four years of the entire nation’s steel demand. With 65% of global shipping being of bulks that are in structural decline — coal, gas, oil and raw iron ore — the current ships and many of the ones just being commissioned are going to be huge sources of scrap. Ships already see one of the highest levels of scrap reclamation for steel when decommissioned, and the current peak of bulkers, with 50% of them steaming to China’s factories, are going to be heading to the scrapping beaches.

This substantial volume of steel will become available as fossil fuel infrastructure is increasingly retired, dismantled, and repurposed. As such, while Norris and the IEA consider a looming scrap shortfall that prevents significant growth of electric arc furnaces, I look at it from a cross-domain perspective and see a much lower constraint on scrap availability..

Reevaluating these sector-specific factors further reinforces my emerging conclusion: the earlier assumptions about continued steel demand growth need significant moderation. Transportation infrastructure, another major steel consumer, faces fundamental shifts as many countries move away from extensive road networks toward mass transit and rail, which are steel-intensive but inherently more efficient and less expansive than highway construction.

Similarly, automotive steel demand growth may flatten considerably as electric vehicles emphasize lightweight materials — aluminum, composites, and specialized high-strength alloys — that gradually reduce steel content per vehicle. Likewise, the energy sector’s steel demand faces a complex transition. Fossil fuel-based infrastructure — oil rigs, pipelines, coal power plants, ships — historically provided significant steel demand. But with fossil infrastructure diminishing, steel-intensive renewable energy infrastructure, primarily wind turbines, and grid expansions, may initially offset some steel demand. Yet, eventually, once renewable build-out stabilizes, this segment’s demand could level off or decline moderately as well.

Machinery and industrial equipment will remain robust sources of steel demand, driven largely by global industrialization, technological transitions, and the shift toward automated manufacturing processes. Yet even here, efficiency improvements and alternative material substitutions—such as aluminum alloys and composites—may curb the magnitude of growth. The steel used in heavy industrial machinery remains challenging to replace, but incremental improvements in efficiency and emerging materials could gradually slow demand growth. The net effect across all sectors, then, is a moderated steel demand outlook. It won’t decrease as rapidly as coal, but my projection is altering to a decrease in absolute volumes over the coming decade.

This adjusted steel demand projection carries implications for global decarbonization pathways. A moderated demand trajectory eases the path toward achieving a fully decarbonized steel industry. Reduced overall volumes mean the reliance on virgin iron production technologies, such as hydrogen-based DRI or molten oxide electrolysis, can remain manageable rather than needing massive scale-up, while increased scrap availability further accelerates the shift toward recycling-based EAF production.

These dynamics sharply reduce the necessity for transitional or interim technologies such as natural-gas-based DRI combined with carbon capture, technologies Norris highlights as crucial bridging solutions. Instead, we might realistically expect the steel sector to transition more quickly and directly toward fully renewable-based methods, leveraging abundant scrap steel and renewable-powered hydrogen production earlier and more widely than current conventional industry projections anticipate.

As always with my projections, it’s important to remember a small handful of things. The first is that I don’t claim to be right, merely less wrong than most. The second is that these are coarse scenarios with very large error bars. They are intended to help us think about the future as well as policy and strategy. The third is that because I look across multiple domains, my projections inform one another in a way that perhaps specialists like Norris might not as readily achieve.

My dialogue Norris underscores the essential value of continually revisiting and challenging assumptions. Professional exchanges like ours illuminate gaps, opportunities, and emerging realities that refine our understanding of complex industry dynamics. The evolution of steel demand is not merely a question of market trends or technological readiness, but a reflection of deep structural changes in global development patterns.

At heart, cement and steel demand trajectories remain fundamentally connected, and as we adjust our expectations downward for global cement consumption, it becomes increasingly apparent that steel demand forecasts require similar recalibration. Moving forward, continued critical evaluation of these interlinked assumptions is essential to accurately projecting—and ultimately driving—global decarbonization efforts.