Making Steel Green

Impression of the H2 Green Steel plant at Boden in Sweden currently under construction. It should be producing DRI steel with near-zero CO₂ emissions in 2024.                 Image credit: H2 Green Steel

In Hydrogen: The Not-So-Green Fuel of the Future, I attempted to dispel some of the hype coming from governments who see Hydrogen as the solution to their ‘net-zero CO₂ emissions’ targets. You can see the attraction: split pure water (H₂O) into its constituent elements by passing a current generated by ‘renewable’ sources through it, store the Hydrogen, and then turn the latter back into electricity using a fuel-cell and pure air. At laboratory-scale it works. And with no harmful chemicals in or out at any stage. The harsh reality of power-grid scale operation, however, involves the creation of both CO₂ and a number of very nasty by-products. Even the storage and distribution of Hydrogen is problematic. Confining the gas inside a pipe or container is rather like the chemical equivalent of herding cats; the gaseous molecule H₂ is physically much smaller than that of Natural Gas (Methane) CH₄, easily getting past seals designed for the latter.

Well, so much for clean energy production and storage; can anything be done about the CO₂ output of two extremely important industrial processes, steel making and the production of Portland cement? Between them, steel and cement account for 14% of the world’s greenhouse gas. Let’s talk about steel making first.

Making steel the old-fashioned way

Steel is almost pure Iron (Fe) with trace amounts of Carbon (C) and other metals which give it the particular mechanical characteristics required. But first, we need a process to turn raw iron ore dug out of the ground, into something approaching pure metallic iron, which is then converted to steel. Millennia ago the ancients found that a dry mix of ore, charcoal and limestone heated in a furnace to over 1000°C, yielded a liquid that could be poured into moulds where it solidified into cast iron. Centuries ago, the Blast Furnace was invented (Fig.1), enabling continuous production of iron with more consistent quality using the same ingredients. The blast furnace is a very sophisticated, but mechanically simple machine through which solid raw materials (the ‘Charge’) introduced at the top are subjected to complex chemical processes as they move down, finally emerging at the bottom as high-carbon molten metallic iron plus separated impurities, the latter also molten, and known as ‘Slag’. The chemical reactions are triggered by heat from superheated air (the ‘Blast’), and material moves through the furnace driven by nothing more than the force of gravity.

The Chemistry of a Blast Furnace

What follows is a description of the chemical reactions that take place inside a blast furnace. Even this is simplified, but it’s included so that a discussion of ‘green’ alternatives can be seen in context. The traditional method of turning ore into metallic iron has been refined over a couple of hundred years, and it’s not going to be easy to come up with a better, non-polluting alternative in a much shorter time frame. Look up Blast Furnace on the Internet and you will find many diagrams, some detailed, others over-simplified, but all illustrating the same basic structure and operating chemistry. Now try looking for Shaft Furnace which can work with Hydrogen gas as the reducing agent rather than Carbon Monoxide. The multiplicity of different designs suggests that the green alternative is still very much in the experimental phase with many problems yet to be ‘ironed’ out.

A blast furnace converts raw iron ore, consisting of various oxides of iron mixed with silicon dioxide (mainly sand), into high-carbon ‘Pig-Iron’ and easily removed Calcium Silicates. The key process is one of Reduction, that is removal of Oxygen from the ore compounds, usually Fe₃O₃ (Hematite) or Fe₃O₄ (Magnetite), leaving pure Fe (Iron). The primary reducing agent used is Carbon Monoxide gas (CO).

Making Carbon Monoxide

Right down in the lower part of the furnace where the charge is subjected to an over 1000°C air blast, the following reaction takes place:

C + O₂ → CO₂ + Heat – Exothermic reaction which produces heat.

The heat generated by this reaction with the Carbon in the coke raises the temperature to about 2000°C, which triggers a second reaction:

CO₂ + C ↔ 2CO – Endothermic reaction which uses heat.

This second reaction needs applied heat to work, reducing the temperature of the charge at this point to about 1700°C. That’s OK but if the temperature falls too far, the reaction reverses, hence the double-ended arrow, and the generation of CO ceases.

So, in total, the Carbon is oxidised by the Oxygen (O₂) in a forced draught of superheated air to create Carbon Monoxide (CO) gas. Efficiency can be improved by injecting pure O₂ into the air stream. Originally, the form of carbon used was Charcoal: wood that has been pyrolyzed or ‘cooked’ in an airless kiln to drive off all the volatile components. Later it was discovered that the residue, called Coke, left from making Town Gas from Coal using the same technique could be used instead. Not only was it better than charcoal, but also widely available as a waste product from all the new domestic gasworks springing up around the country. With the switch to Natural Gas for domestic heating, iron and steel works now make their own coke in coking ovens. Some of the gas produced is recovered to heat the air blast though.

Turning Iron ore into pure Iron

A number of reactions are needed to achieve the goal of pure Iron, each taking place at different temperatures corresponding to particular levels within the charge column (Fig.1).

Starting Point: Preheating

Starting at about 200°C at the top of the ‘Stack’, the charge is dried and preheated to about 400°C as it moves down.

Moving Down: Indirect Reduction

The zone between 400 and 800°C is where the CO rising with the air blast reduces Hematite to Iron in three reactions:

3Fe₂O₃ + CO → 2Fe₃O₄ + CO₂ – Hematite to Magnetite (exothermic)   (1)

Fe₃O₄ + CO → 3FeO + CO₂ – Magnetite to Wüstite (endothermic)   (2)

FeO + CO → Fe + CO₂ – Wüstite to Iron (exothermic)   (3)

Exothermic reactions generate enough heat to compensate for the endothermic one. The maximum temperature at this level (800°C) is too low to melt anything so the charge is still solid. This means that a lot of the ore remains unreduced.

Moving Down: Direct Reduction

Things are a lot hotter in this zone, 800 to over 1200°C, and that enables different reducing reactions to take place:

3Fe₂O₃ + C → 2Fe₃O₄ + CO – Hematite to Magnetite (endothermic)   (4)

Fe₃O₄ + C → 3FeO + CO – Magnetite to Wüstite (endothermic)   (5)

FeO + C → Fe + CO – Wüstite to Iron (endothermic)   (6)

They are all endothermic, so all the necessary heat comes from the burning coke. A lot of other compounds containing elements such as Manganese, Phosphorous and Silicon are also reduced endothermically at this point.

What about the limestone?

While the charge is still in the Direct Reduction zone, the limestone finally does something useful. Iron ore contains a lot of impurities, called ‘Gangue’, such as sand (mostly Silicon Dioxide SiO₂) which needs to be separated from the molten Iron in the Hearth. The limestone consists mainly of the chemical Calcium Carbonate CaCO₃ which when ‘burned’ in the furnace turns into Calcium Oxide CaO and CO₂ gas:

CaCO₃ → CaO + CO₂

Yes, I’m afraid this reaction releases even more of the dreaded greenhouse gas into the atmosphere. CaO, also known as Quicklime, is very caustic, and reacts vigorously with the Silicon Dioxide impurities turning them into Calcium Silicate:

CaO + SiO₂ → CaSiO₃

How is that useful? The Calcium Silicate, also known as ‘Slag’ is less dense than the molten Iron so it floats on top, making its removal relatively easy.

Moving Down: Direct Carburisation

Above 1200°C, Carbon diffuses directly into the Iron forming Iron Carbide or ‘Cementite’:

3Fe + C → Fe₃C

Some Indirect Carburisation with CO will also have taken place at around 900°C, releasing more CO₂ gas.

The solidification point of the iron is reduced by this infusion of Carbon, from 1536°C to about 1200°C. Hence the descending solid charge begins to melt at this lower temperature. So far, much of the charge material remains unchanged; reactions only taking place on the surface of the solid lumps. Once the charge begins to melt, all of it is exposed and the conversion to Pig Iron and Slag is completed above the Hearth.

Solid Iron

The chemistry doesn’t end here, as depending on Silicon content and rate of cooling you get either White Pig Iron where the Carbon remains as Cementite, or Grey Pig Iron, where the Cementite decomposes back into Iron, and Carbon in the form of Graphite. The Grey is shipped off to foundries to make Cast Iron products, while the still-liquid White gets its Carbon level reduced in an Open Hearth Furnace, or a modern Basic Oxygen Furnace producing Crude Steel. And more CO₂.

Let’s Go Green(ish)

Cleaning up steel production is not a new idea: anyone who has visited a steelworks in action will have left awestruck by the heat, noise, dirt and the sight of white hot liquid metal bubbling and steaming like molten lava in a volcano. Can a blast furnace be made zero-carbon?

Build a Better Blast Furnace?

Not really, this ‘machine’ for recovering pure iron from mined ore has been developed and improved over several centuries, but CO₂ production remains central to its operation. However, that operation can be made more efficient and consistent by pre-processing the input materials – the coke and iron ore:

1. Crush the raw ore into a coarse powder. Do the same with the coke.
2. Separate the ore powder from the Gangue powder by Froth Flotation, or if the ore is Magnetite, by Magnetic Separation.
3. Mix the cleaned ore with the coke together with some additives.
4. Funnel the mixture onto a moving grate heated by a gas flame. The coke grains burn at high-temperature ‘welding’ all the particles together into a solid block. This process is called Sintering.
5. Cut the block into pieces about 15mm across and use these to charge the furnace.

The similarly-sized and shaped sintered pieces help ensure an even gas flow through the furnace stack. The fine grains of coke lie in close proximity to the ore particles improving the consistency of reduction throughout the charge. Finally, the much-reduced quantity of Gangue may eliminate the need for the limestone component altogether. None of these improvements will reduce the amount of CO₂ produced, but there will be a marked reduction in the amount of energy consumed per ton of iron with less wastage. Worthwhile, but no nearer to the goal of making green steel.

The Midrex® Process and Directly Reduced Iron (DRI)

In 1966 a new iron-making process was invented which derives the reducing gas CO from Natural Gas instead of coke and feeds it into a simplified blast furnace known as a Shaft Furnace (Fig.2).

Although it looks a bit like a blast furnace in that it’s a firebrick-lined steel tube, there are some very significant differences:

• It operates at a much lower maximum temperature, less than 1000°C, and there is no melting phase.
• The input consists of iron ore pellets or lump ore. Pellets are made from ore that has been crushed and cleaned using the same processes as described above for Sintering. The ore powder is mixed with a binding agent (clay) and water, formed into rough 15mm pellets and then baked to harden them. No coke is involved.
• The hot blast of gas fed into the shaft’s Reduction zone is not air, it’s a mixture of Carbon Monoxide and Hydrogen produced externally by a catalytic convertor called a Reformer. Now comes the clever bit. The Reformer turns a mixture of Natural Gas (CH₄) and the furnace exhaust gas into a mixture of CO and H₂ (see below). CO reacts with the Iron ore to create Iron and CO₂ as before. But the Hydrogen also reacts with the ore to create Iron and H₂O (water vapour). This means that the shaft furnace is a lot ‘greener’ than a blast furnace.
• After being ‘reduced’, the now hot iron pellets move down into the Transition zone to be cooled, not heated further and melted. Natural Gas is fed in at this point to cool the iron down so that it gets discharged at the bottom barely warm. The Carbon component of the gas is transferred to the iron in a carburization process, while the Hydrogen moves upward to supplement the input reducing gas.
• The furnace output consists of solid ‘Sponge Iron’, so-called because of its appearance – cracked and riddled with holes. It’s also known as Cold Directly Reduced Iron (CDRI).

The Chemistry of a Shaft Furnace

The reduction of the iron ore with Carbon Monoxide is the same process used in the blast furnace Indirect Reduction: reactions (1), (2) and (3) above. The reactions with Hydrogen are remarkably similar (see below). But first, both reducing agents must be made.

Making Carbon Monoxide and Hydrogen

These are the chemical reactions taking place in the Reformer, converting Natural Gas and furnace exhaust gas into the reducing gases:

CH₄ + CO₂ → 2CO + 2H₂

CH₄ + H₂O → CO + 3H₂

2CH₄ + O₂ → 2CO + 4H₂

CO + H₂O → CO₂ + H₂

CH₄ → C + 2H₂

An efficient process that makes use of waste heat, CO₂ and H₂O from the furnace.

Reduction

The reduction process with H₂ is very similar to that with CO, except that the by-product is water vapour, not CO₂.

3Fe₂O₃ + H₂ → 2Fe₃O₄ + H₂O – Hematite to Magnetite (exothermic)

Fe₃O₄ + H₂ → 3FeO + H₂O – Magnetite to Wüstite (endothermic)

FeO + H₂ → Fe + H₂O – Wüstite to Iron (endothermic)

Some carburisation also takes place in the reduction zone:

3Fe + 2CO → Fe₃C + CO₂ – Carburising Reaction (exothermic)

3Fe + CO + H₂ → Fe₃C + H₂O – Carburising Reaction (exothermic)

Transition

Cold Natural Gas is blown into a zone immediately below the Reduction zone. This performs two functions: it brings about rapid cooling of the iron pellets, while boosting their carbon content.

3Fe + CH₄ → Fe₃C + 2H₂ – Carburising Reaction (endothermic)

Why the rapid cooling before discharge? If the hot iron is discharged and then left to cool slowly in air, it will re-oxidise (rust) which will render the reduction process somewhat pointless!

Cold Directly Reduced Iron (CDRI)

So, this is the product of the above process: pellets of reasonably high-carbon Sponge Iron at a temperature of about 50°C. Unfortunately, these sponge pellets are not much use for anything. The main advantage of CDRI is that it can be stored temporarily without oxidising too much. At the very least, they need to be melted down and reformed into solid ingots of cast iron. For this purpose, an Electric Arc Furnace (EAF) is required.

Hot Directly Reduced Iron (HDRI)

Given that a steelworks is most likely to want the sponge iron converted to usable metal as soon as possible, it makes sense to skip the cooling phase and dump the still-hot pellets into a nearby EAF immediately. This usually involves a conveyor belt system that can stand the heat.

Hot Briquetted Iron (HBI)

The most popular output of a DRI shaft furnace is iron in the form of hot briquettes. HDRI is compressed mechanically into dense blocks about the size of your hand. This format is a useful compromise, because by eliminating all the cracks and holes present in the sponge iron, it won’t oxidise as quickly when allowed to cool in air and stored. On the other hand, it’s still very hot when fed immediately into an EAF for further processing. Less energy is consumed reheating it.

Let’s Go Really Green

The DRI process with Natural Gas was developed to recover iron more efficiently than the traditional Blast Furnace and Coal method:

• The shaft furnace is simpler in its design and operates at half the temperature.
• Being still solid, the output is a lot safer to handle.
• A lot of the exhaust gas is fed back into the Reformer which creates the reducing gas.
• Replacing Coke ovens with a Natural Gas Reformer makes for a very much cleaner environment – drastically reducing the amount of dust and dirt.
• The amount of CO₂ produced is reduced by about 50%.

Back in the 1960’s, nobody (apart from a few far-sighted climate scientists), saw CO₂ as anything other than a harmless by-product essential for plant life, and ultimately all animal life. Essential because those green plants ‘inhale’ it and ‘exhale’ life-giving O₂. In the 1980’s, its ‘green’ credentials were further enhanced when the fitting of catalytic converters to car exhausts became compulsory in the UK. Now we know that you can have too much of a good thing….

Carbon Capture

The only way to stop blast furnaced-based steelmaking from increasing the level of CO₂ in the atmosphere is to use Carbon Capture technology (CCUS). The trouble is, it’s hugely expensive relative to the value of the steel and so far, very few steel plants in the world have adopted it. A lot of research and early-stage development is going on, summarised in the State of the Art: CCS Technologies 2024 report of the Global CCS Institute. Though it seems doubtful that any of this will be enough to meet current climate change targets.

DRI H₂

So, the short-term answer is to stop creating CO₂ in the first place. A glance at the chemical reaction formula for DRI above will suggest an obvious path to ‘zero-emission’ steel production: dump the Reformer and just use Green Hydrogen as the reducing gas [1],[2]. Hence in theory at least, the only substance emitted from the shaft furnace, apart from iron, will be hot water vapour (H₂O). The basic layout of a fully-green steel plant is shown in Fig.3.

On paper it looks great. No Carbon emissions because coal/coke is not used and there is no CO in the reducing gas. It does mean that no carburisation takes place and so the iron has a near-zero Carbon content. If necessary, it can be added in the EAF melt process.

The problem is generating the Hydrogen: the only green method involves the electrolysis of pure water, and its slow. Add an electrolyte such as Sodium Chloride (salt) to speed things up, and instead of Oxygen as a by-product you get Chlorine and Sodium Hydroxide. Nasty, and definitely not green. And yet one of the first large scale DRI H₂ plants built by H2 Green Steel at Boden in Sweden will start up in 2024. They must have solved the electrolyser problem somehow. I was wondering how the ‘Green Grid’ could possibly cope with the electrical load imposed by the EAF. But then I realised that the Swedes have access to copious amounts of renewable energy that’s not subject to the vagaries of the weather, or the onset of nighttime darkness: Hydroelectricity.

Electrolytic Processes

There are a number of other ideas around for green iron-reduction processes, such as Hydrogen-based Fine-Ore Reduction (HYFOR), Flash or Suspension-based Ironmaking, Plasma Direct Steelmaking and Electrolytic Processes. All are in the development phase, and all may prove impractical, mainly due to the cost of green electrical energy. The most compelling for its simplicity is an Electrolytic Process called Molten Oxide Electrolysis (MOE) from an MIT spinoff company Boston Metal.

The MOE process involves dissolving iron ore in a solvent of Silica (SiO₂) and Quicklime (CaO) at 1600°C. A (large) electric current is passed between submerged electrodes causing the migration of negatively-charged Oxygen ions to the positively-charged anode. Oxygen gas then bubbles off. Positively-charged Iron ions move to the negatively-charged cathode where they collect as elemental iron which pools at the bottom of the cell ready to be siphoned off. In other words, an Electrolyser does all the work eliminating the need for both Reducing gas and for a Blast/Shaft furnace! What’s not to like? Two things:

At one time, the only suitable material with which to make the anode was Carbon. You can see what’s coming. Yes, the C bonds with the O₂ to yield CO₂. In fact, just as much of the gas is added to the atmosphere as with a conventional iron-making process. And people have been trying for years to come up with a satisfactory alternative. In 2013 an anode material consisting of a mixture of steel and chromium was finally developed that could withstand the chemical environment, pass the high electrical current, but didn’t turn the Oxygen into Carbon Dioxide. It’s not ideal though and degrades rapidly with minor changes in the electrolyte chemistry and current distribution. It’s still a work in progress.

The second flaw is economic: it takes about 2 MWh of electrical power to make one tonne of iron. Electricity, especially green electricity is expensive.

For the moment, there is no Carbon Tax on CO₂ emissions so both Blast Furnace and Methane DRI steel plants have no financial incentive to change their ways. The introduction of a carbon tax might just see a big change in the fortunes of the MOE process.

Conclusion

Targets for halting the rise of the planet’s average atmospheric temperature will only be met by an immediate cut in the world-wide emission of the greenhouse gas Carbon Dioxide. That won’t happen because too many industrial processes such as electricity generation, steel and cement making depend upon the burning of fossil fuels. Nevertheless, the attempt must be made to find alternative ‘green’ processes before vast areas of the world become uninhabitable. Pretending that we can ‘just stop oil’ risks sending us all back to a pre-industrial revolution standard of living.

In the area of steel-making, the process that holds the most promise is Hydrogen DRI [3], and a number of pilot plants are nearing completion.

Meanwhile in the UK….

Despite a remarkably upbeat report [4] from UK Steel in 2022, the situation in the UK looks pretty chaotic. Most of the remnants of the British steel industry are owned and operated by two overseas conglomerates, the Chinese, British Steel (Jingye Group) and the Indian, Tata Steel. Both want to end primary steel production in the UK by shutting down the ageing blast furnaces at their respective sites in the Midlands and at Port Talbot. That leaves the government in a serious bind: the closure of these furnaces helps with the UK greenhouse gas targets, but it will likely result in many job losses in current unemployment black spots. Both companies have offered to replace the blast furnaces with ‘green’ EAFs – providing the UK taxpayer pays for them. Another concern for the UK is the strategic importance of primary steel making; EAFs do not work with iron ore, just pig iron and scrap metal. Even if the deals go ahead, there will still be a large reduction in the workforce.

One might ask the question: ‘Why not build Hydrogen DRI furnaces instead?’ I suspect cost is a big factor, but there is also the fact that few, if any, have been running long enough to confirm their viability as a source of genuinely green steel. The Swedish firm H2 Green Steel is taking that risk, but then as I mentioned above, they have access to 24/7 hydroelectric energy. The UK does not, and our only other source of reliable green power is from ageing nuclear power stations which are about to close.

It seems to me that a Small Modular Reactor (SMR) would be the ideal solution to the challenge of providing a reliable source of green electrical power to the electrolyser(s) and EAFs of a Hydrogen DRI steelworks. Any surplus electricity could be fed into the national grid or sold to other industries local to the site. Since Rolls-Royce has been developing them in the UK for years, shouldn’t they be a funding priority for the UK government, given the pressing need to meet the net zero-carbon target?

Latest Research: DRI and the Hydrogen Storage Battery

Just after I finished writing this article, news of some very interesting research grabbed my attention.  As I mentioned in the introduction above, a big issue with the Hydrogen Economy is storing the stuff produced when the wind is blowing and the sun shining, so it can be used when the weather is grey and still. The snag is that the electricity available from solar panels to power electrolysers is at a maximum when the demand for the Hydrogen is at its lowest. And vice versa. Storage of gas made in the summer for use during the following winter is essential. Researchers at ETH Zurich have come up with an answer to the storage problem with some clever lateral thinking. They realised that the chemistry of using Hydrogen to reduce Iron ore to iron and water is reversible, and a rechargeable Hydrogen ‘battery’ is a practical proposition [5].

The battery consists of a steel vessel filled with powdered iron ore, say Magnetite (Fe₃O₄). It’s ‘charged’ by blowing in hot Hydrogen which converts the ore to pure iron (Fe) and water vapour (H₂O) just like in a shaft furnace:

Fe₃O₄ + 4H₂ → 3Fe + 4H₂O       Charging the Battery using hot Hydrogen

When the water vapour disappears from the output, it indicates that the battery is fully charged. The input and output valves are closed sealing in pure dry iron with some residual Hydrogen gas. The absence of Oxygen means that the battery will not ‘self-discharge’ over time.

To recover the Hydrogen, hot steam is forced into the vessel oxidising the Fe back to Fe₃O₄ while releasing H₂ gas:

3Fe + 4H₂O → Fe₃O₄ + 4H₂       Discharging the Battery using hot steam

It sounds too good to be true because the process is hugely inefficient in its use of energy. But with large amounts of ‘free’ wind and solar energy available it might just be worth it to avoid the losses incurred with storing gas.

Further Reading

[1] Assessment of hydrogen direct reduction for fossil-free steelmaking, Cleaner Production, December 2018

[2] Detailed Modelling of the Direct Reduction of Iron Ore in a Shaft Furnace, Materials, Sept 2018

[3] Global green hydrogen-based steel opportunities surrounding high quality renewable energy and iron ore deposits, Nature Communications, May 2023

[4] A Vision for the Future of UK Steel Production, UK Steel, 2022

[5] Safe seasonal energy and hydrogen storage in a 1:10 single-household-sized pilot reactor based on the steam-iron process, Sustainable Energy & Fuels, 2024

If you're stuck for something to do, follow my posts on Twitter/X. I link to interesting articles on new electronics and related technologies, retweeting posts I spot about robots, space exploration and other issues. Click on Bill Marshall at the top of this article to see my back catalogue of DesignSpark articles.