See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/327962750

Direct Reduction: Transition from Natural Gas to Hydrogen?

Conference Paper · September 2018

CITATIONS READS

11 3,633

2 authors:

Marc Hölling Sebastian Gellert

Hochschule für Angewandte Wissenschaften Hamburg ArcelorMittal
31 PUBLICATIONS   217 CITATIONS    3 PUBLICATIONS   13 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Usage of Hydrogen in Steel Production View project

Lecture Materials for Thermodynamics View project

All content following this page was uploaded by Marc Hölling on 29 September 2018.

The user has requested enhancement of the downloaded file.

HÖLLING, MARC, Prof. Dr.-Ing. 1
GELLERT, SEBASTIAN, Dr.-Ing. 2

Direct Reduction: Transition from Natural Gas to Hydrogen?

Abstract

The ambitious goals of the EU in the field of climate protection and reduction of CO2-
emissions are a huge challenge for the European steel industry. Today, the standard route for
production of high quality steel grades is the BF/BOF route, which is almost working at its
physical optimum. Alternative production routes must be developed to transform steel
production in a post-fossil and CO2-free industry as demanded by politics. Since reduction of
iron ore has the highest energy demand in steel production, a CO2-free reduction agent is
needed. A possible replacement could be hydrogen, if it is produced by electrolysis with
renewable energies. Wind and solar energy will then become the new basis for steel
production.
This paper gives a concept of a hydrogen-based production of DRI. The hydrogen
demand is calculated to be in the range of 600 Nm³ per ton of iron and the total electricity
consumption will be about 3.31 MWh/tH2BI. If compared to classical production of DRI, the
energy costs will increase by factor 5.5, which is a strong obstacle for a realization of such a
process under current conditions. Nevertheless, the new process should be investigated in
detail, in e.g. pilot plants, to be prepared when market conditions or regulations may change.

Keywords

Hydrogen, Direct Reduction, DRI, CO2-Emissions

1. Introduction

Today, the most climate friendly way to produce high quality steel grades in industrial
scale is the combination of a DRI plant with an EAF. One installation of such a process can be
found at ArcelorMittal Hamburg GmbH, with an annual production of about 1 Mio. tonnes
per year of billets. Figure 1 gives an overview of the production process and the direct and
indirect CO2-emissions in each step. The total CO2-emissions are 823,000 t/a.
If such a process should be transformed into an almost CO2-free process, then all CO2-
emissions have to be looked at. Indirect emissions, which encounter for approx. 50% of the
emissions, are mainly linked to electricity consumption. Since the energy sector of Germany
is currently under a strong transformation to renewable energies, the emission factor should
become zero in the future and indirect emissions will be of no importance anymore. The direct
emissions in the rolling mill are linked to the reheating furnace, which could rather easily be
changed to induction heating. When based on renewable energies, this induction heating will
also be CO2-free.

1
Hamburg University of Applied Sciences, Faculty of Life Sciences, Department of Process Engineering
2
ArcelorMittal Hamburg GmbH, Process Technology
Fig. 1: Production process of the ArcelorMittal Hamburg GmbH including production
figures and direct/indirect emissions in the year 2016

The emissions of the DRI plant are more challenging, since natural gas is used for a
chemical reaction and not just as an energy source. Thus, a simple change from fossil to
renewables is not that straightforward as in induction heating. To solve this problem,
hydrogen from electrolysis could be used as an alternative reduction agent.

2. Hydrogen-based reduction of iron ore: the H2BI-process

In DRI plants, up to 60 vol-% of hydrogen are used for the reduction of iron ore. The
Circored-process of Outotec [1] was even running on pure hydrogen. Thus, the fundamental
proof of concept of a hydrogen-based reduction is already given. A layout for a hydrogen-
based DRI-plant using a shaft furnace is given in Figure 2, which is based on [2]. Since the
new process is running without natural gas, no efficient cooling can be achieved and the
cooling zone will be replaced by hot briquetting. To make clear that this product is different
from HBI it will be called H2BI for Hydrogen-Hot-Briquetted-Iron (with zero carbon
content).
Hydrogen is produced by electrolysis (1), for which an efficiency of 75% (based on
HCV = 3.54 kWh/Nm³) is assumed. It is mixed with the recycle flow of the process (2) and
heated in two stages. The first preheating uses a part of the top gas of the process, which has
to be purged to avoid accumulation of inert gases like nitrogen or carbon dioxide. The second
heating step up to 900 °C is done by electricity (3). Since hydrogen is produced with an
efficiency of only 75%, it is more efficient to use electric heating instead of combustion of
hydrogen. The mixture of H2/H2O enters the shaft furnace and will reduce iron ore to metallic
iron. A temperature of 700 °C is assumed for briquetting of H2BI as well as for the chemical
equilibrium. The top gas is leaving the shaft furnace at a temperature of 350 °C (4). After heat
recovery, it will be cleaned and cooled in a washer (5). Since the oxidation-product of the
reaction is water, it can be easily removed by condensation in the washer. After purging, the
gas is compressed and internally reheated (6) before mixing with the new make-up hydrogen
(2). A similar layout is given by Midrex in [3].
Fig. 2: Layout of the H2BI-process. Numbers in brackets are explained in the text.

To identify the minimum hydrogen demand, the overall chemical reaction must be taken
into account.

Fe 2 O3 + 3 H 2 ↔ 2 Fe + 3 H 2 O (1)

For the production of two moles of iron, three moles of hydrogen are needed, which can
be transformed to a (minimum) specific hydrogen demand of 600 Nm³/t of iron. Since H2BI
does not consist of pure iron, about 7 wt-% of gangue and a metallization degree of 95% have
to be considered. Thus, the iron content of H2BI is only in the range of 88 wt-% and the
minimum hydrogen demand for H2BI can be calculated to 528 Nm³/tH2BI3.
For a more detailed look on the H2BI-process the shaft furnace has to be investigated,
which can be divided in the upper zone, eq. (2), and lower zone, eq. (3).

Fe 2 O3 + H 2 ↔ 2 FeO + H 2 O (2)
FeO + H 2 ↔ Fe + H 2 O (3)

The gas utilisation is limited by the chemical equilibrium in the lower zone of the shaft
furnace, eq. (3). The corresponding equilibrium constant K can be obtained from [4],

 1953  xH O xH O
K=
exp  − + 1, 0221 = 2 = 2 (4)
 t + 273,15  xH2 1 − xH2O

with t being the temperature in °C, xH2O and xH2 the mole-fractions of water and hydrogen. The
equilibrium composition of the gas phase for the reduction of wustite (FeO) is shown in
Figure 3. The reduction gas enters the shaft furnace with a water content of approx. 5 mol-%.
With an equilibrium temperature of 700 °C the composition in the lower zone is 27 mol-% of

3
The theoretical hydrogen demand can be corrected for H2BI with the iron mass content of 88 wt-%. Thus the
theoretical demand will be 600 ∙ 0.88 Nm³/tH2BI = 528 Nm³/tH2BI.
Fig. 3: Equilibrium composition for water and hydrogen in the presence of wustite
(FeO) and iron (Fe) [4].

water in the presence of wustite. The change in mole-fraction in the lower zone is accordingly
∆xlow =22 mol-% .
The conversion in the upper zone, eq. (2), is not limited by an equilibrium and can be
determined by stoichiometry. By comparison of eq. (2) and eq. (3), the hydrogen demand for
the reduction of hematite (Fe2O3) is half as much as for the reduction of wustite (FeO).
Therefore, the increase in the water content in the upper zone, ∆x up , will be half as much as in
the lower zone, ∆xlow .

1
∆ xup =∆ xlow =
11 mol-% and ∆xtotal =
∆ xlow + ∆ xup =
33 mol-% (5)
2

The outlet composition of the shaft furnace will be about 38 mol-% of water and
62 mol-% of hydrogen. With the change in composition of 33 mol-% and a specific hydrogen
demand of 528 Nm³/tH2BI, the minimum volume flow at the inlet of the shaft furnace can be
derived to be about 1,600 Nm³/tH2BI.
For a more detailed analysis of the H2BI-process a simulation model is needed, in
which component and enthalpy balances are solved. The following parameters were chosen
for the simulation:

• Inlet temperature in the shaft furnace of 940 °C

• Pressure at compressor outlet of 2.6 bar(abs)
• Water content at compressor outlet of 7.7 mol-% (psat (60 °C) = 0.2 bar)
• Purging rate of 10 % of the volume flow after the washer

The results are shown in Table 1. A hydrogen demand of 635 Nm³/tH2BI is calculated,
which is about 20 % higher than the theoretical demand due to the purging rate of 10%. The
volume flow into the shaft furnace (3) is in the range of 1.600 Nm³/tH2BI, as determined above.
Table 1: Results of the simulation model for the H2BI-process

(1) (2) (3) (4) (5) (6)

t in °C 25 188 940 350 60 290
xH2O in mol-% 0.0% 4.7% 4.7% 38.4% 7.7% 7.7%
xH2 in mol-% 100.0% 95.3% 95.3% 61.6% 92.3% 92.3%
spec. Vol. in Nm³/tH2BI 635 1,628 1,628 1,653 1,103 993

With the efficiency of 75% for the electrolysis, a specific electricity demand of
3.0 MWh/tH2BI is found. Additionally, 0.23 MWh/tH2BI are needed for electrical heating from
540 °C up to 940°C as well as 0.08 MWh/tH2BI for auxiliary equipment like blowers, pumps
and conveyers. In total, the H2BI-process has an electricity consumption of 3.31 MWh/tH2BI,
which has to be provided by wind and solar power. For a small DRI-plant with a production
of 80 t/h, like at ArcelorMittal Hamburg, this would result in a demand of 265 MW, which is
in the range of an off-shore wind park.

3. Economics of the H2BI-process

For the economics of the H2BI-process, two aspects have to be considered: Investment
costs for the electrolysis as the most expensive equipment and electricity costs.
Even though costs for electrolysis have been decreasing over the last years, it is still a
rather expensive technology. For example, Voest Alpine started a project using a 6 MW
electrolysis with project costs of about 18 Mio. € [5]. Thus, the current specific costs are in
the range of 3 Mio. € per MW. For a replacement of the Hamburg DRI-plant with 80 t/h, the
needed capacity for electrolysis would be 240 MW, resulting in 720 Mio. € only for the
hydrogen production. This is about three times the investment costs for a classical DRI-plant
with 80 t/h using natural gas.
The electricity costs for the H2BI are not straightforward, since a H2BI-plant should be
in operation at least 7000 h/a and not just in times of high renewable production. As a starting
point, on-shore wind is the cheapest renewable energy with a guaranteed compensation of
about 80 €/MWh. For the backup of the renewable energies in times of low production,
additional 20 €/MWh are assumed, resulting in electricity costs of 100 €/MWh for reliable
renewable energy. The conversion costs for H2BI due to electricity would be about
330 €/tH2BI, which is about 5.5 times as much as a conventional DRI-plant using natural gas.
The current economics give a very poor perspective for the new H2BI-process. There
are strong disadvantages in investment costs as well as in running costs, which cannot be
compensated by sales prices or by abatement costs for CO2. Thus, a realization of a full-scale
H2BI-process with e.g. 80 t/h is rather unlikely.
Nevertheless, a concept of a CO2-free steel production is given with this H2BI-process
and could be investigated in a pilot plant to gain first insights in this new technology. If
energy costs as well as investment costs will decrease along the road to a post-carbon
industry, this could be the starting point for a full-scale installation and the technology could
be ready in time.
 4. Research topics in hydrogen-based reduction

The concept for a hydrogen-based reduction process in a shaft furnace is quite clear and
it is backed by a vast experience in operating DRI-plants with 60 vol-% of hydrogen and even
100 vol-% in the Circored process. So the question is: Why could a pilot plant be useful?
Some points to be investigated are:

• Effect of H2BI with zero carbon content in the EAF (yield, refractory wear,
efficiency). Due to the lack of carbon, H2BI will have a negative impact on the
EAF, which has to be considered in the overall assessment.
• Reaction kinetics for the hydrogen-water-system in combination with iron
ore pellets. Studies on the kinetics of reduction of wustite in a CFB give a wide
range of activation energy, even though diffusion can be neglected [6]. Thus, the
reaction kinetics could be analysed in-situ in a pilot plant.
• Flexibility of the H2BI-process. For the future energy system, flexibility will
become more important due to fluctuating electricity production. A pilot plant
could be used to identify the potentials for load changes and “balancing power”.
• Accumulation of inert gases. The pellets fed to the shaft furnace will contain
air, so nitrogen might accumulate in the process and purging is needed. The
purge rate and effects of inert gases could be evaluated.

5. Conclusions

Strong efforts are necessary to reach the ambitious targets on climate protection in the
EU. The steel industry could become almost CO2-free by changing from fossil to renewables
including renewable hydrogen. A possible concept for a hydrogen-based reduction process
was shown in this paper with a specific electricity demand of 3.31 MWh/tH2BI. Unfortunately,
the conversion costs of this process are tremendously high and it is very unlikely for any
company to invest in a full-scale facility.
Nevertheless, the next step on the way to H2BI could be pilot plant with a production
rate of e.g. 5…10 tH2BI/h. This capacity would be big enough to capture all relevant effects
and the production is sufficient to operate an EAF for a certain time with H2BI. Thus, the
understanding of the H2BI-process could be deepened and the effects on EAF-performance
would be seen.

References

[1] S.A. Elmquist, P. Weber and H. Eichberger, Operational Results of the Circored Fine Ore
Direct Reduction Plant in Trinidad, stahl und eisen 122 (2002), issue 2, p.59
[2] M. Hölling, M. Weng and S. Gellert, Bewertung der Herstellung von Eisenschwamm
unter Verwendung von Wasserstoff, stahl und eisen 137 (2017), issue 6, p.47
[3] J. Ripke, J. Kopfle and C. Ravenscroft, Ironmaking with an H2 energy source, Steel Times
Inernational (2018), issue 1, p.31
[4] F. Oeters, A. Saatci, Stoff- und Wärmeumsätze bei der Reduktion von Eisenerzen,
Stahleisen, Düsseldorf (1985), p.53
[5] H2Future, Voest Alpine (2017), www.voestalpine.com
[6] J. Oh and D. Noh, The reduction kinetics of hematite particles in H2 and CO atmospheres,
Fuel 196 (2017), p.144

View publication stats