- H2 changes heat balance, permeability, and iron morphology
- Pellet specification is key controlling parameter in H2-DRI
Hydrogen is widely presented as the inevitable decarbonisation answer for steel. On paper, it looks elegant: replace carbon with hydrogen, emit water instead of CO₂, and keep the rest of the flowsheet largely intact.
In real plants, hydrogen behaves very differently.
Hydrogen is not just another reductant. It changes reaction heat balance, gas flow behaviour, bed permeability, iron morphology, and even downstream melting dynamics. These second-order effects (often ignored in policy papers and pilot announcements) are what determine whether a hydrogen campaign runs stably or descends into operational chaos.
Hydrogen reduction – fast chemistry, unforgiving physics
Hydrogen reduction reactions are fast, but they are also strongly endothermic in critical stages. As iron oxides reduce, heat is absorbed locally, creating temperature dips that alter gas density, flow patterns, and reaction kinetics inside the reactor.
At the same time, hydrogen reduction tends to produce very fine, highly porous metallic iron early in the reduction sequence. This porous “sponge iron” morphology dramatically increases surface area and contact probability between particles.
The combined effect is subtle but powerful: local cooling alters flow → morphology changes permeability → permeability shifts gas distribution → gas distribution feeds back into temperature and reaction rate.
This coupling is why operators often describe hydrogen as “overactive”. The issue is not just speed; it is that hydrogen pushes the system into an entirely different thermofluid-metallurgical regime.
Sticking and clustering problem
In hydrogen-based direct reduction, sticking is not mysterious. It is a metallurgical consequence of fresh metallic iron forming at particle surfaces, creating metallic bridges between pellets, lumps, or fines. These bridges are strengthened by:
- Elevated temperatures that promote sintering and neck growth
- Iron whisker formation under certain reduction conditions
- Pellet chemistry and induration that influence pore evolution
Hydrogen exacerbates this because it forms porous metallic iron earlier than CO-rich syngas. Porous iron increases contact area. Increased contact accelerates bridging. Bridging reduces permeability. Reduced permeability causes gas channeling. Channeling creates hot and cold spots, which further worsen morphology. Once this loop starts, the reactor becomes self -destabilising.
Channeling and segregation: Silent campaign killer
When permeability breaks down, the reactor stops behaving as a single unit. It effectively becomes several parallel reactors operating at different temperatures and reduction states.
The consequences are severe:
- Under-reduced core zones with falling metallisation
- Over-reduced channels generating fines, hot spots, and sticking
- Misleading top gas analysis that averages out local extremes
This is why hydrogen trials often appear “acceptable” in bulk data until instability suddenly accelerates. The real issue is not chemistry, it is morphology-permeability coupling, which traditional instrumentation struggles to detect early.
Microstructure matters
Hydrogen-reduced iron is often finer and more porous than CO-reduced material. While higher porosity can aid diffusion, it also introduces mechanical fragility, fines generation, and unpredictable behaviour during melting.
Pellet internal structure, gangue chemistry, and impurity distribution play a decisive role in how this porosity evolves. Under hydrogen-rich conditions, small differences in pellet design can translate into large differences in reactor stability. This is why hydrogen DR is far less forgiving than conventional DR, and why pellet specification becomes a reactor control parameter, not a procurement afterthought.
Why H2-DRI often consumes more power
The challenges do not stop at reduction.
In EAF or smelting environments, porous hydrogen DRI behaves differently:
- Lower effective density causes “floating” behaviour in slag
- Higher surface area increases re-oxidation risk at slag-metal interfaces
- Metal droplets take longer to coalesce and settle
The net effect is longer residence time and higher energy input to achieve full metal recovery. This is not a hydrogen chemistry penalty alone; it is a physical metallurgy and phase interaction penalty that directly impacts kWh/t. Ignoring this shifts decarbonisation costs downstream, where they quietly erode competitiveness.
A practical mitigation playbook for H2 campaigns
Plants that succeed with hydrogen do not chase maximum H₂ percentages. They design for stability first.
Temperature strategy: Don’t run H2 Like CO
Higher temperature may boost kinetics, but it accelerates sticking and sintering when fresh metallic iron is forming. Controlled temperature profiles and staged heat addition are essential.
Hydrogen staging
Instead of high hydrogen everywhere, successful campaigns use lower H₂ or mixed reductants during early iron formation and increase H₂ where permeability is already stable. This delays problematic morphology until the bed can tolerate it.
Burden engineering
Pellet chemistry, basicity, additives, induration, and even surface coatings influence sticking behaviour. In hydrogen campaigns, burden design becomes a primary control lever.
Gas distribution and pressure control
Hydrogen systems are more sensitive to channeling. Distributor design, recycle ratios, and pressure stability must be treated as first order controls.
Melt shop integration
Porous DRI changes slag practice, charging strategy, and residence time. Melting must be redesigned as part of the hydrogen system, not treated as a downstream constant.
This article is published by BigMint in collaboration with author Mr. R.V. Sridhar, Senior Independent Advisor, McKinsey & Co.
Leave a Reply