ITmk3
Process

Shoichi KIKUCHI1, Shuzo ITO1, Dr. Isao KOBAYASHI1, Osamu TSUGE2, Koji TOKUDA2

The ITmk3 process, which produces high-quality iron

(ITmk3 iron nuggets) from ore fines and coal fines, was
developed based on a unique concept of iron ore and
carbon composite technology. The development started
in 1996, followed by test operations at a pilot plant and
large-scale pilot plant. The first commercial plant was
constructed in the U.S. and has started producing iron
nuggets. This article outlines the history of the
development, the features of the process, the product,
and the future outlook.

Midrex Technologies, Inc.

1,900

BF

Liquid

FASTMET

1,400
1,200

1,000

1,400

ITmk3

1,600

Gas DR

1,600

Stable Fe-C system

1,800

800
0.0

Liquid

Liquid
Graphite 1,200

Graphite

1,000

Temperature ()

Research & Development Department, Iron Unit Division, Natural Resources & Engineering Business,

Temperature (K)

800
Graphite
0.5

1.0

1.5

2.0

2.5

600
3.0

3.5

4.0

4.5

5.0

C ()

Introduction
ITmk3 (pronounced as "[ai ti:] mark three") is a
rapid ironmaking process that includes reducing ore,
carburizing & melting iron and separating slag, all at
relatively low temperatures.
Referring to generation classes of ironmaking
processes, the first generation would be the blast
furnace (BF) process, which is the current
mainstream. The second is the direct reduction
ironmaking (DRI) process as typified by the MIDREX
process. ITmk3 (Ironmaking Technology Mark 3)
falls in the third generation category; it is a process
based on a concept totally different from those of
the earlier generations. In the ITmk3 process, a
series of reactions occur within about 10 minutes.
The reactions occur much faster in this process than
in BF and DRI processes. In a typical BF process, raw
materials dwell in a furnace for about 8 hours, while
in the MIDREX process, the materials stay in a
shaft furnace for 6 hours.

Fig. 1 Operational region of ironmaking processes

fuel coal. In the 1980s, various attempts were made

to develop a new process, including the smelting
reduction process. With this background, Kobe
Steel and its US subsidary, MIDREX Technologies,
Inc., developed a process called FASTMET, which
produces reduced iron by heating agglomerates
consisting of fine ore and coal.
In 1995, when developing the FASTMET process,
the companies found that metallic iron can be
separated from slag within ten minutes of heating.
This reaction principle was adapted for a new
ironmaking process, ITmk3. This paper describes the
details of the new process.
Fig. 1 is the iron-carbon phase diagram, which
compares the operational regions of ironmaking
processes, including ITmk3. The blast furnace (BF)
process produces pig-iron saturated with carbon at
a temperature around 1,500, while ITmk3
produces metallic iron with higher purity at a
temperature lower than that required for BF.

1. Developmental background
2. Basic research
Blast furnace processes require pretreatments
such as producing coke from coking coal and
preparing sinter from iron ore. Thus, an integrated
steel mill with a blast furnace must have a capacity
greater than 10,000 tonnes/day to be feasible. This
limits the types of resources, such as raw materials,
that can be used. The production may also lack in
flexibility. Direct reduction ironmaking processes
using natural gas, on the other hand, are limited as to
plant locations. DRI plants can be built only in areas
where natural gas is produced at a low cost.
Thus, the industry has focused on new processes
that utilize the abundant reserves of fine ore and
77

KOBELCO TECHNOLOGY REVIEW NO. 29 DEC. 2010

Kobe Steel began to study a new ironmaking

process in 1996. Using a tube furnace installed at
the Iron & Steel Research Center (Kakogawa Works,
Kobe Steel), the company found that metallic iron
grows rapidly, being separated from slag, at a
relatively low temperature. Fig. 2 shows this
phenomenon occurring in the tube furnace. After
this result was reported and became the subject of
repeated discussions among researchers, domestic
and overseas, the conclusion was reached that this
was an unprecedented phenomenon. The new
ironmaking process was called ITmk3.

Fig. 2 Inside view of tube furnace

Fig. 4 Test furnace at Surrey University

3min

5min

6min

9min

Fig. 5 ITmk3 reaction mechanism

Fig. 3 Hot samples of box-type furnace test

The company initiated collaborative research

involving universities and institutes, domestic and
overseas, and installed a box-type furnace at the
Iron & Steel Research Center to conduct experiments
on a larger scale. Fig. 3 shows hot samples
discharged from the box furnace. Metallic iron has
a bright orange color, while slag appears in dark
orange, showing they are clearly separated from each
other. Tohoku University and Tokyo Institute of
Technology conducted research to elucidate the
reaction mechanism of this new ironmaking
phenomenon1), 2). University of Surrey, in the UK,
conducted research on heating methods, using a
large multi-stage box furnace newly installed at their
site (Fig. 4). Max Planck Institute in Germany studied
the equation modeling of the reactions3).
The reaction between iron ore and coal remains
the same as that for general ironmaking and is
expressed as follows:
FexOy + yCO = xFe + yCO2 (1)
CO2 + C = 2CO (2)
C(s) C (carburized) (3)
Fe(s) Fe(l) (melt) (4)
Reactions (1) and (2) occur in the FASTMET
process. In the ITmk3 process, there are the
additional reactions (3) and (4), which separate
metallic iron from slag.

In this new ironmaking process, particles of iron

ore and coal exist next to each other in each
agglomerate. This is in contrast with the BF and
DRI processes, which use bulky materials such as
iron ore and coke, and may explain the fast reaction
of the new process to some extent. However, this
does not fully explain the rapidity of the reactions.
Another series of experiments, conducted at the
Kobe Steel Iron & Steel Research Center, confirmed
a similar phenomenon. In these experiments, typical
ores and coals, four types each, were combined
and were heated in a box furnace. A study was
conducted to clarify the effect of raw materials,
temperature and furnace atmosphere on the
reactions. Another study was conducted to elucidate
the reaction mechanism. This study involved
investigating the cross sections of agglomerates in
the stages of reducing, melting and slag-separating,
respectively. Fig. 5 shows sample results of the
cross-sectional study. During the first three minutes,
the agglomerate, consisting of fine ore and coal, did
not exhibit any significant change in appearance,
despite the reduction reaction that should have
occurred inside. After five minutes, the metallic iron
and slag started to partially melt and became
separated. In about six minutes, the entire
agglomerate started to melt rapidly, with metallic
iron being separated from the slag. After nine
KOBELCO TECHNOLOGY REVIEW NO. 29 DEC. 2010

78

minutes, they had separated completely. University

of Surrey conducted experiments to optimize this
process reaction by separately controlling the
reduction and melting. Separable control is a feature
of their multi-stage box furnace.
In the reaction mechanism, the rapid carburizing
phenomenon could not be explained by the
conventional theories of gas carburization or solid
carburization. Much time was devoted to elucidating
the new carburizing phenomenon. Unlike the shaft
furnaces used for DRI plants, the contribution of
gas carburization is considered to be small. A recent
study by Tohoku University found carburization
via liquid slag 2), which appears to contribute to the
rapid carburization along with solid carburization.
Using the box furnace at the Iron & Steel Research
Center, Kobe Steel tested over a hundred types of
raw materials, which verified the versatility of the
raw materials that can be used for the ITmk3
process. The tests included applicability tests on
low grade materials, such as iron ore containing a
large amount of crystal water, oil coke and upgraded
brown coal (UBC). They confirmed the applicability
of these low grade materials.
3. Application of new ironmaking method
3.1 Pilot plant
In parallel with the basic research, application
studies were conducted on the new ironmaking
process based on this unique carburization
phenomenon. Among the various processes
deliberated, granular ironmaking was chosen as the
most feasible approach. Designing of a pilot plant
began in 1998. Unlike box furnaces, in which
reactions occur under ideal conditions, the pilot
plant had many issues to be addressed. Process
development was conducted to resolve these issues.
Fig. 6 shows the process flow of granular

ironmaking. The process comprises

1) agglomerating iron-ore and coal (blue),
2) reducing and melting the agglomerates (red),
3) separating metallic iron from slag (green), and
4) treating exhaust and recovering heat (yellow).
The pilot plant employs a rotary hearth furnace,
which facilitates radiation heating, to exploit the
rapid reaction feature of ITmk3. In a smelting
reduction process, melting occurs before reduction,
during which FeO may corrode furnace refractory.
To avoid this, reduction and smelting technology
was adopted in which melting occurs after reduction
is completed.
In a rotary hearth furnace (RHF), the heating
combustion gas from a burner may adversely affect
the reductive atmosphere in the furnace. To resolve
this issue, the Mechanical Research Laboratory at
Kobe Steel conducted a simulation study based on
the computational fluid dynamics (CFD) to
optimize the RHF, in the designing stage, for gas
characteristics such as temperature distribution and
flow (Fig. 7).
There are many technical challenges in
continuously producing metallic iron by melting
raw materials on the hearth of a furnace. Once
melted and separated from slag, the metallic iron
is cooled, solidified and discharged out of the
furnace. This subjects the furnace hearth to repetitive
thermal stress in short cycles. The furnace hearth
must also resist corrosion caused by the slag

Fig. 7 CFD analysis result for rotary hearth furnace

Flue gas

Iron oxide
concentrate

Reductant
(coal)

Mixer

Heat
recovery
system

Air

Pelletizer

Dust
collector

Dryer

Burner
fuel
Separation

Iron nugget

Rotary hearth
furnace

Slag

Fig. 6 Process flow

79

KOBELCO TECHNOLOGY REVIEW NO. 29 DEC. 2010

reacting on the hearth. Several tens of types of

refractories were tested in order to identify the one
best suited for the purpose. Repetitive thermal
stress, involving rapid heating and cooling, was
applied using an experimental furnace owned by
a refractory manufacturer. A box furnace was used
to clarify the corrosion behavior of slag. The hearth
refractory was determined based on these
experiments.
Although important, the selection of the hearth
refractory does not resolve all the issues associated
with the continuous production of granular iron. The
key to the process is to maintain the hearth in good
condition during the entire process, including the
charging and discharging of materials to/from the
furnace. Studies using the box furnace made such a
technology available for the pilot plant.
A major difficulty was to design an apparatus for
retrieving granular iron from an RHF. Although
cooled and solidified, the granular iron is still at
an elevated temperature. Several methods were
contemplated. Having expertise in equipment for
handling hot iron, specialists at Kobe Steel's
Machinery and Engineering Company (currently,
Machinary Business) designed the discharging
apparatus. The Materials Research Laboratory of
Kobe Steel collaborated in the selection of materials
used for the apparatus.

Fig. 8 Small pilot plant in Kakogawa Works

Feed zone

In June 1999, within one year of the beginning

of process development and designing, Kobe Steel
started the construction of a small pilot plant, with
a capacity of 3,000 tonnes/year, at its Kakogawa
Works (Fig. 8). Experimental operation began at the
end of August 1999. In September, the operation
began to produce reduced iron. The plant was
confirmed as producing iron reduced over 90% as
originally planned. Then, the operating temperature
was raised to successfully produce granular iron for
the first time. In the beginning, operation continued
for about six hours a day, during which various
adjustments were made to the raw material mixture,
heat pattern, furnace atmosphere and retention time.
Improvements were also pursued for productivity
and the quality of the granular iron. During these
continuous operations, each lasting for about six
hours, there were no problems with the hearth-a
major concern-nor was there any significant
mechanical failure. Experiments continued at a good
pace (Fig. 9).
In late November, when the company started
continuous operations lasting for an extended period
of time, a pool of molten iron began to form on the
hearth and grew larger and larger. It was not possible
to cool and solidify this large pool of molten iron, a
problem that necessitated the development of a new
technology.
In April 2000, a new technology for maintaining
the hearth was developed through observing the
changes that occur on the hearth, completing the
first campaign of the pilot plant.
The second campaign operation began in the
latter part of 2000 with a view to receiving visits from
potential collaboration partners and operating for an
extended period of time.
This campaign succeeded in conducting a stable
operation that continued for the target number of
days, the result of a modified hearth maintenance
technology. The hearth was repaired periodically
while producing granular iron, which enabled
continuous operation. This approach has no
theoretical limits on continuity and advanced the

Reduction zone

Melting zone

Fig. 9 Inside views of RHF at small pilot plant in Kakogawa Works

KOBELCO TECHNOLOGY REVIEW NO. 29 DEC. 2010

80

3.2 Large pilot plant

In 2002, designing began for the large pilot
plant. MIDREX Technologies, Inc. collaborated in
designing the details. An RHF was the heart of the
pilot plant. There were many rotary hearth furnaces
for heating, with sizes up to 50m (diameter)7m
(hearth width), delivered to the market. A goal was
set for the first commercial plant to have a size
close to that described above (50m7m). This
would make the target capacity of the commercial
plant about 500 thousand tonnes/year, judging from
the productivity achieved by the small pilot plant.
The larger the pilot plant, the smaller the risk in
upsizing it to a commercial scale; however,
developing a large pilot plant may become too
costly. Therefore, care must be taken in deciding
the size of the pilot plant.
The following describes two factors emphasized
in determining the size of the rotary hearth furnace
used for the large pilot plant.
1) In the ITmk3 process, the secondary combustion
of the carbon monoxide generated in carbon
81

KOBELCO TECHNOLOGY REVIEW NO. 29 DEC. 2010

1.2
Small pilot plant
1
Total energy ()

application. During the campaign, data were

collected for the design and operation of a large pilot
plant that followed.
While conducting this experimental operation,
Kobe Steel was visited by potential partners, who
were given tours of the site. Among the visitors, the
most interested were those in the consortium
organized by Iron Range Resources, ClevelandCliffs Inc. (currently, Cliffs Natural Resources), the
largest mining company in North America, and
Steel Dynamics, Inc., the second largest EAF based
steel manufacturer. Iron Range Resources and
Cleveland-Cliffs Inc. had attempted to revitalize
the mining industry, while Steel Dynamics, Inc. had
been searching for a stable supply of iron source.
Their aims coinciding, they were in search of a
partner for collaborative development.
In March 2002, the decision was made to
construct a large pilot plant in the beneficiation
and pelletizing facility of Northshore Mining,
Minnesota, owned by Cleveland-Cliffs. Different
roles were allotted for the collaboration: Kobe
Steel to provide technology and to design the pilot
plant, Cleveland-Cliffs to supply iron ore and to
provide operators, Steel Dynamics to receive and
evaluate the product granular iron, and Iron Range
Resources to finance the project. Later, the U.S.
Department of Energy also provided financing,
recognizing the project as the development of a
new ironmaking technology with excellent energy
efficiency.

0.8
0.6
0.4
Large pilot plant

Commercial plant

0.2
0
0

100

200
300
400
Annual production (1,000t/y)

500

Fig.10 Production size vs. energy consumption

composite agglomerates significantly affects

energy efficiency and the reductive atmosphere
in the furnace. The energy input (furnace gas
volume) for producing a tonne of granular iron
was correlated to the production amount
(Fig.10). The capacity of the large pilot plant
was determined to be 25,000 tonnes/year,
which is at the folding point of the correlation
curve.
2) The key to this process is the maintenance
technology for the furnace hearth, and the
hearth width is an important factor. Considering
the hearth width of 7m for the commercial
plant, a conservative scale-up factor, from the
pilot to commercial, was considered to be a
factor of 3 or 4. Assuming a similar scale-up
factor for the small pilot plant in Kakogawa, the
hearth width for the large pilot plant was
determined to be 2m.
The furnace profile was determined using
the CFD model, which was further optimized
by feedback data from the operation of the
small pilot plant.
As it involves the development of a process, the
project must minimize the risk to equipment.
Because of this, the large pilot plant had to adopt
industrially proven equipment as much as possible.
Another prerequisite was that the equipment could
be used for the commercial plant. In the end, the only
apparatuses especially designed for this process
were the charging and discharging units for the
RHF. The rest were chosen from among those that
had been industrially proven.
The process flow for the large pilot plant is
basically the same as that for the small pilot plant,
except that, for separating iron from slag, the small
pilot plant adopts a batch-type process, while the
large pilot plant adopts a continuous process.
The construction of the large pilot plant began
in June 2002. The plant was completed and blown-in
in May 2003. The first day was spent in producing
reduced iron with a high metallic ratio, as in the

(a) Outside view of RHF

(b) Inside views of RHF (reduction zone)

(c)Inside views of RHF (cooling zone)

Fig.11 Rotary hearth furnace at large pilot plant

case of the small pilot plant. The production of

granular iron succeeded on the second day. The
large-scale demonstration followed and continued
for 15 months until August 2004 (Fig. 11).
The following four campaigns were conducted
during the demonstration operation:
1) optimizing the hearth maintenance technology
2) improving productivity and granular iron
quality
3) improving the unit consumption of fuel
(optimizing the secondary combustion)
4) testing performance
Various improvements were made during these
campaigns. The demonstration by the prototype
plant went smoothly for an extended period of time
with an equipment utilization of 91 to 94%. A minor
issue was the failure of a ceiling refractory caused
by the insufficient firing of support anchor tiles by
a refractory manufacturer.
Official environmental measurements, conducted
twice during these campaigns, confirmed the process
to be more environmentally friendly with higher
energy efficiency than conventional ironmaking
processes. The advantage was also confirmed by the
Minnesota Pollution Control Agency4) and the U.S.
Department of Energy.
The following summarizes the features of the
ITmk3 process.
- The process is simple.
- It allows the direct use of low grade materials (e.g.,
fine ore and fuel coal with neither sintering nor

Fig.12 Iron nugget product

coking).
- It is highly energy efficient.
- It has a low environmental load.
- It facilitates the adjustment of production by
starting and stopping.
- The facility cost is low.
- Most of the equipment involved in the process has
been industrially proven with high reliability.
- The plant is easy to operate, without a need for
handling molten iron, which is unprecedented in
ironmaking processes; so operators with mining
companies can run the plant in the same manner as
they run a pelletizing plant.
The granular iron produced by the large pilot
plant (Fig. 12) was received well by large EAF based
steel makers in the U.S. and was used as a raw
material for their steel products, such as sheet, plate
KOBELCO TECHNOLOGY REVIEW NO. 29 DEC. 2010

82

and special rods. The granular iron, with its ease of

melting in EAFs, increased the furnace productivity
by 5 to 8% compared with conventionally used pigiron. Thus the ITmk3 was demonstrated to be
superior, not only in its ironmaking process, but also
in its product, granular iron. The granular iron,
which is to be continuously charged into EAFs, will
further increase the productivity and energy
consumption of the steelmaking process.
The following are the features of granular iron.
- It has a 2 - 4% carbon content.
- It is a clean iron with no impurities such as copper
and nickel, which adversely affect the steelmaking
process.
- It has a large specific gravity.
- It is easy to melt with a low melting point and
high thermal conductivity.
- It is of an appropriate size for charging
continuously into steelmaking furnaces.
- It has a size that is easy to handle.
- Unlike other reduced iron and hot briquette iron
(HBI), the iron is completely reduced.
- Unlike other reduced iron and hot briquette iron
(HBI), it does not reoxidize and ignite a fire.
- Unlike other reduced iron and hot briquette iron
(HBI), it does not contain a gangue constituent.
In the U.S., granular iron is referred to as "iron
nuggets," after gold nuggets.
3.3 Construction of commercial plants
The success of the large pilot plant immediately
led to planning the construction of the first
commercial plant. The Mesabi Iron Range, the
largest ore mine in North America, lies from east to
west, about 100km north of Lake Superior. There
are many beneficiation/ pelletizing plants in this
area. One of them is the plant run by LTV Mine,
which closed in 2002 due to the steel recession.
With this prospective site in mind, we have begun to
work toward obtaining an environmental permit
to construct a commercial plant (Fig. 13).
In 2007, stimulated by the recovering demand for
steel, Steel Dynamics, Inc. decided to construct its
first commercial plant with a capacity of 500
thousand tonnes/year of granular iron to secure a
clean iron source. This project was co-financed with
Kobe Steel. Kobe Steel's role included the provision
of the process under license, engineering, major
equipment and instructors. Steel Dynamics, Inc.
agreed to accept all the granular iron produced and
use it for their own EAFs. The construction of the
commercial plant began in June 2007, and production
started in 2010 (Fig. 14).
The construction of another commercial plant in
83

KOBELCO TECHNOLOGY REVIEW NO. 29 DEC. 2010

Fig.13 Location of commercial plant

Fig.14 Commercial plant at Hoyt Lakes

Michigan, in the U.S., is being planned in

collaboration with Cleveland-Cliffs Inc. (currently,
Cliffs Natural Resources), one of the partners in
collaborative development. Efforts to obtain an
environmental permit are underway. Commercial
plants are also being planned in many other countries,
including Kazakhstan, India and Ukraine.
4. Business model
Granular iron manufacturing is a process
suitable for locations adjacent to ore mines. The
process converts iron ore, the major raw material,
into value additive granular iron. The operation is
similar to the conventional pelletization. The process
also removes unwanted oxygen and gangue minerals
contained in iron ore and pellets, allowing the
shipment to steel mills of a product consisting
essentially of iron. Unnecessary oxygen and gangue
minerals account for about 40% of ore weight. Thus
the new process decreases the shipping weight
and significantly reduces CO2 emissions during
shipment. The steel mills that receive the granular
iron can enjoy the benefit of granular iron that
generates no slag and decreases the cost associated
with the treatment of slag.

In the U.S., EAFs account for 63% of steelmaking,

significantly exceeding the amount made by the blast
furnace - converter method. However, the quality of
scrap, the major raw material for EAFs, has become
more degraded year by year, and securing clean iron
source has become a major concern for EAF based
steel makers. Clean iron source such as pig-iron and
HBI, on the other hand, is almost totally dependent
on imports from other countries, such as Venezuela,
Brazil, Ukraine and Russia. Thus, US EAF makers
are subject to supply instability and large price
variations. These issues in the U.S. can be resolved
by the prevalence of granular ironmaking plants.
In the U.S., most ore mines are found in the
Mesabi Iron Range, to the north of the Great Lakes.
Therefore, many steel makers are located in the
area around the Great Lakes. They also import about
ten million tonnes of clean iron source every year.
The iron source is landed at New Orleans in the
south and transferred upstream on the Mississippi
River at a high transportation cost. The prevalence
of ITmk3 plants around the Mesabi Iron Range, in
the area around the Great Lakes, is expected to
replace imported iron source, including scrap, with
granular iron.
The following are the advantages to each sector
of the industry that have been brought about by
granular iron and granular ironmaking plants.
1) Mining sector
- Mining companies can make granular iron, which
is much more value additive than conventional
iron ore and pellets.
- Mining companies can expand their customer
bases to EAF based steel mills in addition to blast
furnace based integrated steel mills.
- Allowing the use of low grade ore will extend the
lives of mines.
- The new process is feasible even for small scale
mines.

2) EAF based steel manufacturers

- Use of granular iron improves the productivity
and energy efficiency of the steelmaking process.
- The manufacturers can secure the source for clean
iron.
- Granular iron allows conversion to higher end
steel products.
- Installing granular ironmaking plants in steel mills
allows the use of hot granular iron, which further
increases energy efficiency.
3) Blast furnace based steel manufacturers
- The new process does not require equipment with
a high environmental load, such as coke ovens and
sinter plants.
- The new process allows the use of low grade
materials.
- Ironmaking by the new process requires a small
capital investment.
- The new process facilitates production adjustment.
- Granular iron production overseas enables the
offsetting of CO2.
- Granular iron decreases the cost and CO2
emissions associated with transportation, compared
with iron ore and pellets.
- Granular iron generates less slag than do iron ore
and pellets.
Conclusions
The application of ITmk3 has just begun. In the
same manner as blast furnaces have evolved for
over a hundred years, ITmk3 is expected to expand
its applicability from granular ironmaking to, for
example, molten ironmaking and low temperature
ironmaking. Kobe Steel will continue its efforts in the
research and development of these new processes.
References
1) Ko-ichiro Ohno et al. : "Steel Research," 74 (2003), pp. 5-8.
2) Taichi Murakami, et al. : CAMP-ISIJ Vol.12 (1999), p. 247.
3) S. Meissner et al. : The 11th Japan-Germany Seminar on
Fundamentals of Iron and Steelmaking, Reduction and Meltdown
of Pellets in the ITmk3 Process, (2002), p. 6.
4) Larry Lehtinen : Mesabi Nugget Research Project, United States
Department of Energy (2005).

KOBELCO TECHNOLOGY REVIEW NO. 29 DEC. 2010

84