There are many aspects of the energy transition where the solutions are clear, the technology is ready, and what is needed is to scale deployment. There is a need to massively build wind and solar; deploy electric vehicles; redesign our cities for shared mobility, mass transit, bikes, and pedestrians; and implement more stringent building codes to maximize efficiency and drive the electrification of heating.
But there are other sources of emissions where the solutions are not obvious, the paths to decarbonization are more complex, and the technology still needs to be developed. These are described as the “hard-to-abate” sectors and include shipping and aviation, as well as two of the basic building blocks of our modern world: steel and cement.
As sources of emissions, steel and cement production are not small: each represents around 7 percent of energy-related emissions globally. As such, any route to deep decarbonization will sooner or later have to confront these two materials.
A Need for Heat
While the production of steel and cement each involves distinct processes, they have one challenge in common: the need for continuous, high-temperature heat. Incidentally, this is also required for the production of glass and some chemicals.
And while there are inherent emissions in the steel- and cement-making processes, a large portion of the emissions associated with the production of these two materials comes from this need for high-temperature process heat and our current practices of getting this heat—which is consistently burning fossil fuels.
But not all technology can produce the temperatures that are needed. In an October 2019 report, the Innovation for Cool Earth Forum identified several potential sources of high-temperature process heat, finding that biodiesel, hydrogen, and electric resistive heat could all supply the temperatures needed for steel, cement, and glass production. But it also noted that these existing options “face challenges based on price, performance, and variability” and that more technology development is needed.
Among these three, hydrogen has probably attracted the most attention. However, not all hydrogen is low carbon. Existing hydrogen production is dominated by “grey” hydrogen made using the steam-reforming methane process. This is a relatively cheap way to get hydrogen but results in substantial emissions.
There are also plants that produce “green” hydrogen using electrolysis, where water is split into hydrogen and oxygen. This can be a zero-emissions process if it uses electricity from renewables, but the current process is not terribly efficient, is expensive, and is currently only being done at a relatively small scale.
The Riddle of Steel
In the end, solutions for low-carbon, high-temperature process heat only deal with one of the problems. Scrap steel can be simply melted in an electric arc furnace (EAF), but for the production of steel from iron ore, it is necessary to separate the oxygen from the iron in the ore, a process called reduction.
This has traditionally been done by burning layers of iron ore and coke—nearly pure carbon—in blast furnaces. The oxygen binds with the carbon in the coke, resulting in iron and carbon dioxide (CO2). And this process accounts for a big part of the emissions from steel production.
Around 5 percent of new steel is currently produced through a different process called direct reduction (DRI) in which the oxygen is removed without melting the steel, and this is typically followed by melting the steel in EAF. This is mostly done with syngas, a combination of carbon monoxide and hydrogen derived from natural gas, and is popular in nations where gas is abundant, and coal isn’t. This results in approximately 50 percent lower emissions but doesn’t fundamentally solve the problem.
In the snowy forests of Northern Sweden, near the Gulf of Bothnia, a plant is being built to explore one of the most promising routes to decarbonizing steel: hydrogen-based DRI. “From a technical point of view, hydrogen has been proven as very good agent for iron ore reduction in the laboratory for a long time,” states Martin Pei, chairman of the board of Hybrit Development AB, a joint venture of three companies to make fossil-free steel. “It is rather easy.”
Scaling it is another matter. The only plant to date to produce steel using hydrogen-based DRI at scale was a plant based on fluidized bed reactor technology in Trinidad, which began production in 2010 but shut down after multiple difficulties. “They produced quite a few tons, but they never made it fly,” explains Pei.
This new pilot plant in Sweden being built by Hybrit partners SSAB, LKAB, and Vattenfall will make steel from iron ore pellets using “green” hydrogen produced using electrolysis, with the resulting sponge iron melted in EAF powered by renewable electricity. The partnership’s pilot plant is scheduled to start next year as the first step toward a larger facility.
“Scaling up means to be successful for the whole process so that we can produce on a stable basis and at least running continuously for several weeks, day and night,” states Pei. He also notes that Hybrit’s process will use a shaft furnace for the hydrogen DRI, as is used in gas-based DRI, and not the fluidized bed reactor process that the Trinidad plant struggled with.
This means that the use of hydrogen is the only deviation from the existing, proven DRI processes.
The Cost Issue
There is still the challenge of cost, most notably the relatively higher cost of green hydrogen versus coke as a reduction agent. However, the location of Hybrit in Northern Sweden offers multiple advantages. First, the region has access to local, high-quality iron ore. Second, industrial-scale electricity is both cheap and relatively clean in Sweden. And, finally, there is the price on carbon emissions through the EU Emissions Trading Scheme, which could increase in the future.
“Without putting a cost on emissions, it would be very difficult to be economically competitive using Hybrit,” explains Pei. Hybrit estimates that with current prices—including the existing price of coal, emissions, and Swedish wholesale electricity—its cost to make steel will only be 20–30 percent higher than the blast furnace method.
What is further encouraging is that these cost assessments of hydrogen DRI are based on comparably nascent technologies for the reduction process and hydrogen production, for which the capital requirements are expected to come down as the industry scales up. In addition, the cost of clean electricity as an energy source is structurally decoupled from the cost of coking coal, so as the cost of renewables continues to decrease, hydrogen- and electricity-based steel-making processes will become increasingly competitive.
And while Hybrit is the farthest ahead with this idea, it is not alone. ArcelorMittal is investigating the use of hydrogen at its DRI plant in Hamburg, Germany, and says that it plans a pilot project but has not given a time line for this. Germany’s Salzgitter also reports that it is investigating the use of hydrogen as part of its decarbonization pathway.
“To decarbonize the steel sector, you are going to need new steel mills and not to use the old ones.”
But even if Hybrit and/or other steelmakers can get hydrogen-based steel to work at a larger scale and lower cost, they still face an uphill battle to transform the steel industry, as existing blast furnaces will need to be replaced. “The challenge in my mind is that we are going to have overcapacity in blast furnaces,” states Thomas Koch Blank, who leads work on decarbonizing steel and cement at Rocky Mountain Institute. “To decarbonize the steel sector, you are going to need new steel mills and not to use the old ones.”
Given that steel mills represent massive capital investments that are recovered over decades, this is not a small matter. But for those blast furnace-based steel mills that are kept online, there is still the option of installing carbon capture and storage systems.
But if you think steel is hard to decarbonize, consider the process of making cement. “It is even worse,” states Koch Blank. “You have carbon in the molecule of your raw material.” He is referring to limestone (CaCO3), the most common raw material for making portland cement. This limestone is ground into small bits (clinker) before being heated in kilns to a very high temperature to produce calcium oxide (CaO), releasing CO2 in the process.
The heat used for this process is an emissions problem on its own, but more than 60 percent of the annual emissions from cement production comes from the limestone itself. This also means that, like steel, methods to improve the efficiency of the cement-making process are limited in their CO2 reduction potential.
Some of the most promising solutions for reducing CO2 in cement-making involve using different materials. And while there are several alternatives, many of these have limits as well.
The Energy Transitions Commission (ETC) in its report series Mission Possible identifies minerals for making belite clinker, calcium sulphoaliminate, and carbonization of calcium silicates clinkers, as well as magnesium silicate–based cements as potential alternatives. The organization says magnesium silicate–based cement could be made without any process emissions but also notes that the minerals to form these chemistries are “much less available” than limestone.
Unlike these, ETC notes that pozzalan, a kind of volcanic rock, is much more abundant. The organization describes cement made from pozzalan feedstock and other alkali-/geopolymer-based cements as potentially the most promising and estimates that pozzalan-based cement could eliminate process emissions by 70 percent.
An Industry Built on a Chemistry
There is also the option of using less portland cement in the mix to make concrete, but both this and the use of alternative feedstocks means concrete with different characteristics. This would require architects and engineers to alter their approaches, and this may not be as easy as it sounds.
“The specific chemistry of the cement product is basically written into building codes,” explains Koch Blank.
“The specific chemistry of the cement product is basically written into building codes.”
There are also issues for builders. Portland cement sets in 12 hours, meaning that a concrete floor or beam poured one day will be solid the next, if not fully cured. Using another kind of cement that takes longer to set could affect building schedules, workflows, and, ultimately, project economics. But that is not even the biggest concern for the construction industry, especially for the many smaller construction firms.
“These companies are small enough where if anything happened to the building, they would go under immediately,” states Koch Blank. “The building code makes their construction insurable.”
Reuse and Reduce
The simplest means for lowering the emissions from both cement and steel is to reuse both to the maximum amount possible. An increasing portion of global steel demand is met by recycling steel in EAF, and the International Energy Agency expects most of the growth in steel demand to 2050 to be met this way. If these furnaces are powered using clean electricity, emissions can be nearly eliminated.
However, there are limits to this process. To make certain kinds of high-quality steel, you need a minimum portion of virgin iron in the mix. But more centrally, recycled steel will not meet all the demand in a growing market.
In the 21st century, the world population continues to increase, more people are moving to cities, and the standard of living is rising in the developing world. All of that means even more demand for steel and concrete.
For cement, the problem is even more difficult. Once cement is hydrated, it can only be used as fill or aggregate, meaning that there will still be a demand for new cement.
Another route to reducing emissions is dematerialization: the use of less material to achieve these same means. This is addressed in the companion article titled “Doing More with Less” and includes not only new designs but also the use of different materials.
ETC cautions against relying too much on any one approach to decarbonizing cement and steel. Achieving deep reductions will likely require a combination of low-carbon heat and different methods for eliminating process emissions—potentially including carbon capture and storage. Greater efficiency, more reuse of materials, and dematerialization processes will also need to play a role.
In the end, we have many of the solutions that we need. But the task is more complex, and the exact routes are less clear.