October 28, 2023: Green Steel
Good afternoon. Today we will look at some possibilities for reducing CO2 emissions from steel production, particularly green steel (produced from hydrogen).
Future Topics
But before we get into that, I want to solicit topic ideas. Don’t worry; I’m not running short of ideas. My list of topic ideas would last for years, even if I don’t add any new ones, which I obviously will. But I would like to know if there are any topics that readers would like for me to tackle, or alternatively, if there are past posts that were particularly interesting.
Green Steel
Before we talk about the hydrogen option for producing steel and other low-carbon options, let’s take a look at steel production in general.
In 2021, 1.9 billion tons of steel were produced worldwide, with more than half in China. Steel is an alloy of iron and carbon and has greater strength and fracture resistance than pure iron. There are numerous alloys—for instance, stainless steel is alloyed with chromium. Steel has been in use since ancient times, and Wootz (Damascus) steel, with originated in India, was renowned in the ancient world. Crucible steel developed in the Middle Ages. It was the Bessemer process, patented by Henry Bessemer in 1856, that was the first true inexpensive, mass production technique and that turned steel into a quintessential industrial material.
Iron appears in nature as iron oxide, and to make useful steel, it is necessary to remove the oxygen through a process known as reduction. This is typically done with carbon. Coal is turned into coke, which both provides the heat to melt the iron ore, and the carbon atoms to “steal” oxygen. The carbon dioxide created in this manner is a source of emissions that is inherent to modern steel production.
In modern times, the Bessemer process has given way to the blast furnace-basic oxygen furnace (BF-BOF) process, though the use of carbon as a reducing agent remains the same. As of 2020, 72% of world steel production was through the BF-BOF process, and the remaining 28% was through the electric arc furnace (EAF) process. An EAF, which typically takes scrap steel as a feedstock, melts the material with an electric arc. While an EAF typically has about a third of the lifecycle greenhouse gas emissions of BF-BOF, its use is limited by the amount of scrap steel economically available. An inherent source of EAF emissions is the graphite electrodes used for electrolysis. This is also a problem for Hall-Héroult process for aluminum manufacturing, and efforts to deploy inert anodes instead of graphite are ongoing.
Steel production is responsible for 3.6 billion tons of CO2 emissions every year, or about 7% of world greenhouse gas emissions. These figures can be broken into 3.1 billion tons for BF-BOF and 500 million tons for EAF. BF-BOF has lifecycle emissions of around 1.8 tons CO2 for every ton of steel produced.
Several solutions are suggested. Some just nibble around the problem. BF-BOF operates near thermodynamic limits and cannot be improved much. Recycling is all good and well, but there are practical limits. The same can be said for efficiency in steel usage. For a deep solution, it is necessary to employ new steel-making processes.
There are a few processes posited, and the leading process is known as green steel (I don’t like the term, but it’s what we have). The key idea is to use hydrogen instead of carbon as a reducing agent, so the result of reaction with iron oxide is water instead of carbon dioxide. The term “green steel” implies three things:
Hydrogen is used as a reducing agent,
The hydrogen is produced through electrolysis, rather than the steam-methane reforming that most H2 comes from today,
The electricity to power electrolysis is sourced from low-carbon sources.
If these three conditions are met, then lifecycle emissions of steel are estimated at around 200 kg CO2 per ton of steel, which is a nearly 90% reduction over the BF-BOF status quo.
If H2 is produced instead through the more common SMR route, then H2 reduction offers only a modest improvement over current BF-BOF. SMR has an estimated carbon intensity of 21.9 tons per ton H2 produced, and 60 kg H2 are needed for each ton of steel, so that works out to a carbon intensity of 1.3 tons per ton of steel, compared to the 1.8 that is common today.
Both BF-BOF and EAF steel typically cost $600-900 per ton as of 2021, a price that was elevated due to commodity and supply chain issues. Last year, it was estimated that green steel would carry a carbon mitigation cost of $64-180 per ton, which is high but not outrageous; estimates of the social cost of carbon usually range from about $50 to $100 per ton. This would imply a cost premium of around $100-300 per ton for green steel.
For green steel to really take off, this cost premium has to be reduced, and first and foremost that refers to electricity costs. This paper estimates costs of a green steel plant. About two-thirds to three-quarters of total levelized costs of steel are operational, with most of that being electricity and iron ore feedstock. Of the remaining costs, most are capital costs, with a small portion being overhead (marketing, insurance, legal, etc.). About half of the capital costs are electrolyzers, which also have the potential for cost reduction.
The paper estimates that 3100 kWh of electricity are required for a ton of green steel. Industrial electricity prices in the United States range from 5 cents/kWh to 23 cents/kWh across states, though obviously a highly energy-intensive green steel plant will gravitate to where electricity is cheaper. If a power grid is dominated by cheap and abundant hydropower or nuclear power and can save 3 cents/kWh, that is $91/ton of steel saved. IRENA estimates that H2 electrolyzers can be reduced in price by 40-80%; so that might be another $50/ton on the steel price. These two things should put green steel within range of a modest carbon price, or perhaps no price at all.
Here is a map of green steel projects planned around the world for the next 10 years. Projects today are pilot or demonstration projects, with no full-scale plants yet in operation. But market research predicts that will change; if green steel is a $364.5 billion industry in 2032 as this report projects, and the price is $600/ton, then that is a third of current steel demand in 10 years. Some major projects under development today include H2 Green Steel and Hybrit.
Hydrogen reduction is not the only approach to low-carbon steel. A company called Boston Metals is developing a process of driving chemical creations through a process known as molten oxide electrolysis, bypassing the reduction phase entirely. Also called electrowinnowing, this process also has an estimated carbon intensity of 200 kg per ton of steel and would require less electricity, though it is not as well developed. Carbon capture from steel mills is another possibility, though also one that has not yet gotten beyond the niche phase. The International Energy Agency estimates that the carbon mitigation cost of CCS from steel is $40-100/ton.
Steel has been referred to as one of the “hard to abate” sources of carbon dioxide emissions. This view may be outdated in the near future. The prospects in the steel industry also demonstrate as clearly as any a basic truth about climate change mitigation: progress will come because of, not despite, an economy that is friendly toward innovation and investment.
Quick Hits
In a special Halloween edition of The Batch, Andrew Ng tackles some sensationalist fears around artificial intelligence. Ng argues that fearmongering, especially unsubstantiated fears that AI could cause human extinction, is doing great harm.
It has only been three weeks, but it is clear that Israel’s Operation Sword of Irons will yield valuable lessons for American military planners. War on the Rocks considers four questions that Americans should be considering. They are,
How was Israel, with a heretofore renowned intelligence apparatus, caught off guard by the October 7 Hamas attack?
Can Hamas, with much more rudimentary tools, successfully fight with the technologically advanced IDF?
Is urban warfare possible without massive death and destruction? So far this has not been demonstrated.
Can the post-combat phase be effectively planned? The United States has had major recent failures in postwar Iraq and Afghanistan, and it looks like Israel may have to go a similar route in Gaza.
Also from War on the Rocks, Michael Eisenstadt considers the floundering American approach to Iran’s nuclear program and how it can be brought back on track.