United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
EPA-600 2-80-036
January 1980
Research and Development
Direct Reduction:
A Review of
Commercial  Processes

-------
                  RESEARCH REPORTING SERIES


Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional  grouping  was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:

    1. Environmental Health Effects Research

    2. Environmental Protection Technology

    3. Ecological Research

    4. Environmental Monitoring

    5. Socioeconomic Environmental Studies

    6. Scientific  and Technical Assessment Reports (STAR)

    7. Interagency Energy-Environment Research and Development

    8. "Special"  Reports

    9. Miscellaneous Reports

This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This  series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to  repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
                        EPA REVIEW NOTICE
This report has been reviewed by the U.S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the contents necessarily
reflect the views and policy of the Agency, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.

This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.

-------
                             EPA-600/2-80-036

                                   January 1980
   Direct Reduction:
       A Review of
Commercial  Processes
               by

           Larry G. Twidwell



       Program Element No. 1AB604
  U.S. ENVIRONMENTAL PROTECTION AGENCY
     Office of Research and Development
  Industrial Environmental Research Laboratory
      Research Triangle Park, NC 27711
  U.S. ENVIRONMENTAL PROTECTION AGENCY
     Office of Research and Development
         Washington, DC 20460

-------
                                  ABSTRACT

     Direct reduction commercial processes for ironmaking have been reviewed.
The potential for environmental degradation appears to be minimal.  A
detailed environmental assessment does not appear to be warranted.  It is
recommended that samples of scrubber water and sludge material be collected
and characterized from several gas reductant reactor systems.  It is also
recommended that rotary kiln solid reductant reactors be sampled to ensure
that the final gas effluent does not contain harmful concentrations of
organic species, sulfur oxides and trace metal contaminants; and that such
systems be monitored for harmful rates of fugitive emissions.

-------
                                  CONTENTS
Abstract	ii
Figures	iv
Tables   	iv

   1.  Introduction  	   1
          Definition 	   1
          Statistics of Growth 	   1
          Reasons for Growth 	   4
          Terms and Nomenclature	   8
   2.  Direct Reduction Processes  	  10
          Introduction 	  10
          Systems	10
             Gas Reductant Systems 	  13
                Midrex	13
                HyL	18
                Others	22
             Solid Reductant Systems 	  22
                SL/RN	26
                Others	28
             Direct Reduction in United States 	  30
   3.  Potential Pollution Problems  	  31
          Gas Reductant Systems  	  31
          Solid Reductant Systems  	  32
          Recommendations  	  35
   4.  Steelmaking Using Metallized Product  	  37
          DR-EAF Steelmaking 	 	  37
          Other Uses	39

References	40

-------
                                  FIGURES

Number                                                                Page
  1    Growth in production of Direct Reduction sponge iron  	   2
  2    Types of Direct Reduction processes 	  11
  3    Feed material size distribution 	  14
  4    Midrex standard flowsheet 	  17
  5    The HyL process	21
  6    The SL/RN process	29

                                  TABLES
Number                                                                Page
  1    Regional Distribution of Direct Reduction Plants  	   3
  2    Annual Capacity of Direct Reduction Plants,  1977,
       1980, 1985	5
  3    Sponge Iron Production by Process Types (1977)	12
  4    Midrex Process  	  16
  5    Energy Requirements for the Gas Reductant Systems  	  19
  6    HyL Process	20
  7    Gas Reductant Processes Excluding Midrex and HyL (1978)  ....  23
  8    Solid Reductant Systems (1978)  	  24
  9    Rotary Kiln Processes for Treating Steel  Plant  Wastes  	  25
10    Energy Requirements for Solid Reductant Systems 	  27
                                    IV

-------
                                  SECTION 1

                                INTRODUCTION
Definition
     For the purpose of this report "direct reduction" is defined as the
reduction of iron oxides by the use of either solid or gaseous reductants
to produce a solid iron product.

     It would be more appropriate to designate the "direct reduction"
processes as alternative processes, i.e., alternatives to the production
of iron in the conventional blast furnace.  The reactions that occur in the
"direct reduction" processes are no more "direct" in producing iron than
those reactions occurring in the blast furnace.  The difference between the
process types is that the direct reduction processes operate at much lower
temperatures than does the blast furnace and therefore produce a solid iron
product instead of a liquid product.

Statistics of Growth

     Present day large scale commercialization of the direct reduction
processes dates back only about 25 years (however, the first iron produced
was probably via direct reduction processes practiced over 3000 years ago).
Growth and projects of growth are given in two graphs (1-3) presented in
Figure 1.  The most recent projections (2,4,5) suggest that production
may achieve about 50 million tonnes of direct reduced product by 1985
with the possibility that the production could reach 100 million tonnes
(3,5).   This could mean that about 5-10 percent of the world's iron may be
produced by direct reduction processes in 1985.  In 1978 about 3 percent
of the world's iron was produced by direct reduction processes.

     Miller (4) has tabulated the regional distribution of direct reduction
plants by production and by number of plant installations (Table 1).  In
1978 almost three-fourths of the world's sponge iron was produced by the
developing world:  Latin America (30.9%), Middle East (19.7%), Africa (8.4%),
and Asia (13.9%); about one-sixth was produced in North America.  The
projected production distribution in 1985 is about two-thirds for the
developing world, one-eighth for North America.  The number of direct
reduction plants in the world at the end of 1978 was 55.  This number is
anticipated to double by 1985 (4).

     Miller (4) has also tabulated the individual plants presently operating,
those scheduled for completion by 1980, and those planned for operation by

-------
                           I960
                                   1965
1970
                                                    1975
                1980
                                                                    1985
                       Greenwalt,  R.  and J.  Stephensen  (2)
                      (Reproduced  with permission  of AIME.)
                     IIOO
                     ooo
                               I960
                                                     OPEN HEAffTW STEEL
                                       l»70      ISC      I99O
                                         YM,   A'PMEREOUCCD IRON TOR IRON
                                              B*PNEREOUCED IRON FOR STEEL
                                  Miller,  R.   (3)

Figure 1.   Growth in production of direct reduction  sponge  iron,

                                          2

-------
                        TABLE 1.  REGIONAL DISTRIBUTION OF DIRECT REDUCTION PLANTS (4)
                          Distribution by Production
                          production;
co
      Region
 1S78
         % world
kt/year   total
 1979
         % world
kt/year   total
 1980
         % world
kt/year   total
1981-85
         % world
kt/year   total
North America
Latin America
Western Europe
Eastern Europe
Middle East
Africa
Asia
Oceania
Total




Region
North America
Latin America
Western Europe
Eastern Europe
Middle East
Africa
Asia
Oceania
Total
©American Society
2935 16.9
5345 30.9
1650 9.5

3415 19.7
1450 8.4
2409 13.9
120 0.7
17324 100
Distribution by
Installations:
1978
% of
No. total
11 19.6
16 28.6
9 16.1
0
4 7.1
3 5.4
12 21.4
1 1.8
56 100
for Metals and
2935
8645
1650
2500
4415
1450
4134
120
25849
Number of

1979

No.
11
18
9
1
5
3
13
1
61
the Metals
11.3
33.4
6.4
9.7
17.1
5.6
16.0
0.4
2935
9065
3650
5000
5215
1700
4134
120
100 31819
Installations


% of
total
18.1
29.5
14.8
1.6
8.2
4.9
21.3
1.6
100
Society (London)


1980

No.
11
19
11
2
6
4
13
1
67
1977.
9.2
28.5
11.5
15.7
18.4
5.3
13.0
0.4
100



% of
total
16.7
28.8
16.7
3.0
7.6
6.0
19.7
1.5
100

6455
15035
5700
5000
7915
5100
4194
1520
50920


1981-85

No.
16
33
15
2
9
9
14
3
101

12.7
29.5
11.2
9.8
15.6
10.0
8.2
3.0
100



% of
total
15.8
32.7
14.8
2.0
8.9
8.9
13.9
3.0
100


-------
1985.  His results are presented as Table 2.  (It is known (6) that the Steel
Company of Canada has not started up their facility after shutting down in
late 1976.  Also Hecla Mining has closed down its mining operation in Casa
Grande, Arizona.)  Note that the United States' direct reduction capacity is
indeed small; e.g., 11.2% of the world's capacity in 1977.

Reasons for Growth

     The reasons for the optimism in projected growth of direct reduction
produced iron are:

     1.   A high purity product can be produced.  The conversion of iron
          oxide to iron is accomplished in the solid state; i.e., a liquid
          iron product is not produced as in the blast furnace.  This is
          because the reaction temperatures are much lower in direct
          reduction furnaces.  Less reduction of gangue material occurs and
          the solubility of solid iron for impurities is much less in the
          solid state than in the liquid state.  Secondly the product is
          relatively pure iron because the feed materials are chosen to
          have low impurity levels, especially sulfur and phosphorus.

     2.   A wide variety of reducing agents are candidates for use in the
          direct reduction process; e.g., coal, coke, charcoal, reformed
          natural gas, products of coal gasification, oil, coke oven gas,
          and coke breeze.  The blast furnace requires coke as the primary
          reductant.

     3.   Refractory life is longer and less expensive refractories are
          required for the direct reduction furnaces than are required in
          the blast furnace because of the lower furnace operating tempera-
          tures.

     4.   Developing nations can enter the steel business on a moderate
          scale with a smaller investment in a direct reduction-electric
          arc steelmaking furnace (DR-EAF) combination as compared to the
          coke, blast furnace, basic oxygen furnace combination; i.e., the
          capital cost is only about 60 percent of the conventional process
          capital cost (7-9).

     5.   Some developing nations have rich iron ore deposits and natural gas
          or oil  supplies but do not have metallurgical grade coals
          available.

     6.   Developing nations need to develop their basic industries and
          therefore prefer to pretreat their ores before shipping to other
          nations, i.e.;  keep as much industry home as possible.

     7.   Direct  reduction products can be used as feedstock for supplement-
          ing scrap in an electric furnace, as a coolant feed material  for
          the basic oxygen furnace operations, and as an iron value added
          to  increase the productivity of a blast furnace (4).

-------
TABLE 2.  ANNUAL CAPACITY OF DIRECT  REDUCTION  PLANTS, 1977, 1980, 1985  (4)
                  Plants in Operation  on  1 January 1977*
Plant
No.
1-2
3
4-6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
(cont
Year
of Start
1954
1954
1954
1957
1957
1900
1964
1967
1968
1969
1969
1969
1970
1970
1971
1971
1972
1972
1972
1973
1973
1973
1973
1974
1974
1974
1975
1975
1975
1975
1975
1976
1976
•d>1976
Company
Hoganas Grangesberg
Hoganas Corporation
SKF etc.
Tohoku-Satetsu
HYLSA-Monterrey I
HYLSA-Monterrey II
Hitachi Metals
TAMSA
Anglo-American Corp.
HYLSA-Puebla I
Oregon Steel Mills
Kawasaki Steel Co.
New Zealand Steel
Thyssen-Purofer
Georgetown Steel Co.
Nippon Steel Co.
Kawasaki Steel Co.
Hamburger Stahlwerke
Armco Steel Corp.
Acos Finos Piratini
Dunswart Iron & Steel
SIDBEC-DOSCO
MINORCA
USIBA
HYLSA-Monterey III
Nippon Kokan KK
Steel Co. of Canada
Heel a Mining Co.
Sumitomo Metals Co.
Sumitomo Metals Co.
All is Chalmers Co.
Dalmine-Siderca
Sudbury Metals Co.
Fior de Venez.S.A.
Country Process
Sweden Hoganas
USA Hoganas
Sweden Wiberg
Japan Rotary kiln
Mexico HyL
Mexico HyL
Japan Wiberg
Mexico HyL
South Africa High veld kiln
Mexico HyL
USA Midrex
Japan Kawasaki
New Zealand SL-RN
West Germany Purofer
USA Midrex
Japan Koho
Japan Kawasaki
West Germany Midrex
USA Armco
Brazil SL-RN
South Africa Krupp
Canada Midrex
Venezuela HIB
Brazil HyL
Mexico HyL
Japan SL-RN
Canada SL-RN
USA SL-RN
Japan Sumitomo
Japan Kabota
Canada ACCAR
Argentina Midrex
Canada ACCAR
Venezuela FIOR
Reductant
Coke breeze
Coke breeze
Coke breeze
Coal
Natural gas
Natural gas
Coke breeze
Natural gas
Coal
Natural gas
Natural gas
Coke breeze
Coal
Natural gas/CO gas
Natural gas
Coke breeze
Coke breeze
Natural gas
Natural gas
Coal
Coal
Natural gas
Natural gas
Natural gas
Natural gas
Coal
Coal
Coal
Coal
Coal
Coal ,oil ,gas
Natural gas
Natural gas, oil
Natural gas
Annual rated capac-
city of DRI, kt
170
70
90
24
100
270
• 10
280
1000
315
300
72
120
150
400
48
240
400
330
60
150
400
650
250
475
350
360
60
240
210
50
330
240
400

-------
Table 2 (Continued)*
 38
 39
1976
1976
    Total
       Nippon Steel Co.       Japan
       Ferriere di Arvedi     Italy
           NSC           Oil
           Kinglor-Metor Coal
                                      150
                                       40

                                     8804
Plant   Year
 No.  of Start
      Plants Under Contract at Start of 1977 and Scheduled for Completion by 1980
                                                                            Anmia
        Company	
        SIDBEC-DOSCO II
        SIDOR III
        DIDOR III
        COSIQUA
        NISIC
        HYLSA-Puebla II
        Iraq Iron & Steel Co.
        Anglo-American Corp.
        Consol i dated Gol d Fi el ds
        Kawasaki Steel Co.
        Nippon Steel Co.
        NISIC
        SIDERPERU
        Acindar
        BSC-Hunterston
        Qatar Steel Co.
        PT Krakatau
        NISIC
        PT Krakatau
        SIDOR IV
        SIDOR IV
        USSR-Kurak
        North Sea Iron Co.
        Nord. Ferrowerke
        USSR-Kurak
        ISCOTT
        TIKA
        Saudi Arabia-Jubail
                                Country
              Process
                              Annua
                              acity
I  rated cap-
of  DRI. kt
Reductant	
Natural gas             625
Natural gas             360
Natural gas             360
Gasified oil            350
Natural gas             330
Natural gas             625
Natural gas            1485
Coal                    300
Coal                    100
Coke breeze             250
Oi 1                     240
Natural gas            1200
Coal                    100
Natural gas             420
Natural gas             800
Natural gas             400
Natural gas             575
Natural gas            1000
Natural gas            1725
Natural gas            1200
Natural gas            2100
Natural gas            2500
Natural gas             800
Natural gas            1200
Natural gas            2500
Natural gas             420
Gasified naphtha        250
Natural gas             800
                      23015
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
1977
1977
1977
1977
1977
1977
1977
1977
1977
1977
1977
1978
1978
1978
1978
1978
1978
1979
1979
1979
1979
1979
1980
1980
1980
1980
1980
1980
 Canada
 Venezuela
 Venezuela
 Brazil
 Iran
 Mexico
 Iraq
 South Africa
 USA
 Japan
 Japan
 Iran
 Peru
 Argentina
 UK
 Qatar
 Indonesia
 Iran
 Indonesia
 Venezuela
 Venezuela
 USSR
 UK
 West Germany
 USSR
Trinidad-Tobago
 Zambia
 Saudi Arabia
Midrex
Midrex
HyL
Purofer
Purofer
HyL
HyL
Highveld kiln
Hockin kiln
Kawasaki
NSC
Midrex
SL-RN
Midrex
Midrex
Midrex
HyL
HyL
HyL
Midrex
HyL
Midrex
Purofer
Mi drex
Midrex
Midrex
HyL
Midrex
    Total

-------
                      TABLE 2 (Continued).*
Annual Capacity of Direct Reduction Plants 1977, 1980, 1985 (4).
      Projects Planned for Operation Between 1981 and 1985
Plant
No. Country
68
69
70
71
72
73
74
75
76
77
78
79
80
81

82

83
84
85
86
87
88
89
90
91

92

93
94
95
96
97
98
99
100
101
102
103
104
105
106
107


USA
USA
Canada
Argentina
Bolivia
Brazil
Brazil
Brazil
Brazil
Brazil
Mexico
Venezuela
Italy
Spain

Spain

Algeria
Egypt
Tunisia
Turkey
Iran
Iran
South Africa
New Zealand
USA

USA

El Salvador
Argentina
Argentina
Brazil
Brazil
Brazil
Brazil
Colombia
Ecuador
Greece
Abu Dhabi
Libya
Morocco
India
Australia
Total
*<& American Soci
Company and location
Texas Ferro reduction
Gulf Coast Consortium
Interprovincial Iron Co.
Gurmendi
SIDERSA (Santa Cruz)
COSIGUA II
COFAVI
USIBA II
IMBITUBA(Sta.Catarina)
Piritim II
TAMSA II
FIOR de Venezuela
Adriatic Consortium
Sid. Gibraltar

PREPELSA (Huelva)

SNS (Jijel)
Government (He! wan)
Government (Gabes)
EDAS
NISIC (Bandar Abbas)
NISIC (Esfahan)
SCAW & Anglo-American
New Zealand II
Republic Steel
(Massillon)
Republic Steel
(Gadsden)
Government (Acujutla)
HIPASAM (Punto Colorado)
Lucini
DEDINI
IKOSA-Pains
Mendez Junior
Mannesmann
ACENOR (Barranquilla)
EOJASIDOR
SIDERHELLAS
Government
Government (Misurata)
Annual rated
Probable capacity of
Reductant DRI, kt
Natural gas
Natural gas
Coal
Natural gas
Natural gas
Gasified oil
Gasified oil
Natural gas
Gasified coal
Coal
Natural gas
Natural gas
Natural gas
Liquified natural
gas
Liquified natural
gas
Natural gas
Natural gas
Natural gas •

Natural gas
Natural gas
Coal
Coal
Coal

Coal

Coal
Natural gas
Natural gas
Coal
By-product gas
By-product gas
By-product gas
Natural gas

Natural gas
Natural gas
Natural gas
Government Natural gas
Government (AndraPradesh)Coal
Hamersley Iron Co.

Natural gas

900
1500
400
400
200
350
350
300
400
200
300
2000
800
450

500

1200
400
800
400
2800
1200
200
200
360

360

160
400
400
400
160
350
400
200
400
300
400
500
300
60
1200
22600
ety for Metals and the Metals Society (London) 1977.

-------
     8.   There will be a deficit in scrap availability by 1980 (10).  Direct
          reduction sponge iron is viewed as a good substitute feed  material.

     9.   The time required for commissioning a DR/EAF steel  plant is  about
          2 to 3 years compared with 5 to 7 years for a conventional inte-
          grated steel plant  (9).

    10.   Air pollution problems are minimized in a gas direct reduction
          plant because all gases are cleaned and recycled.  One of the main
          air pollution producing stages in the conventional blast furnace
          is cokemaking which is not required in a gas D.R. facility (9).

    11.   Transferability of  technology has been demonstrated  by fast start-
          ups of new plants.   This represents an important economic advantage
    12.   Energy  requirements  are  less  for  the  DR-EAF process than for the
          conventional  coke-blast  furnace-BOF process (9).

    13.   Production  facilities  have  been constructed to produce over a
          million metric  tons  per  year  of D.R.  iron; i.e., two D.R. furnaces
          are presently used to  accomplish  this.   It is anticipated that
          single  furnaces with capacities of a  million tons per year will be
          possible  in the near future (9).

Terms and Nomenclature

     Many articles  have been published  on direct reduction processes.  The
nomenclature used by  the  various authors is sometimes confusing and contra-
dictory.  Therefore,  some of the terms  and  their various meanings will be
presented in this section.

     The charge fed to the  reduction  reactor is an  iron oxide containing
material.   It may be  ore  lumps,  ore fines,  pelletized agglomerates, formed
briquettes, or a  combination of  several of  these materials.  In the litera-
ture, the feed material may be simply characterized as ore, lump ore,
nature ore, mineral ore,  lump  oxide,  pellets, pelletized ore, pelletized
oxide, or pelletized  fines.

     The product  of the direct reduction reaction  is a reduced material that
contains metallic iron, some form  of  iron oxide, some iron carbide, and
gangue that was contained in the original feed.  This product is called
sponge iron, luppen,  metallized  iron, metallized product, reduced iron,
direct reduced iron,  or a trade  name, such  as Midrex iron or HyL iron.

     The reducing reactions remove oxygen from  the iron oxides.   A measure
of the success of accomplishing  this  is described in several  ways.

          Percent metallization  or degree of metallization or metallization
          is defined  as
                                      8

-------
                 Met  =  % Fe (as metal)   inn
                 Met<    % Fe (total)	x 10°
     It is the percent of iron present in the feed material that has been
     converted to metallic iron.

     Equivalent metallization is defined as

               % Eq. Met. = % Met. + 6 (%C in sponge).

     This term is used in presentations on the HyL process but has not been
     widely accepted.

     Degree or Percent of Reduction is defined as:

                            Oxygen removal from feed
                            Original oxygen combined with iron

     The terms percent reduction and percent metallization are not the same,
and percent metallization or simply metallization is the normally reported
value; i.e., metallization is more descriptive of how effective the reduction
was in producing metallic iron.

     It is also important to know how much metallic iron is present in the
final product; i.e., the weight of metallic iron divided by the weight of
product.

                              Wt(Fe)
                     % Fe =   Mire;       x 100
                              Wl( product)

     The metallization and the metallic iron percentages are normally not
the same numerical value; e.g., for a feedstock that contains 65 percent
iron that is reduced to give a 90 percent metallization, the metallic iron
content is about 86 percent.  Bertram (11) points out that the two terms,
metallization and metallic iron, have sometimes been used interchangeably
and, therefore, care should be taken to understand how the author has
performed his calculations.

-------
                                   SECTION  2

                          DIRECT REDUCTION  PROCESSES
 Introduction
      More than 1200 patents have been issued covering  various  aspects  of
 proposed direct reduction processes.   Nearly 100  direct reduction  schemes
 were examined by iron and steel  producers  between 1950 and 1975 (7). Of the
 many proposed and piloted processes  only 12  major designs  have progressed
 to commercial application.   At the end of  1978  there were  55 direct reduction
 plants in the world (1).   Thirty-seven of  these plants use a gas reductant.
 Twenty-five of the 37 plants are based on  either  the HyL or Midrex process.
 Eighteen of the 55 plants use a solid reductant.   Fourteen of  the  solid
 reductant systems are rotary kiln operations and  use coal  or coke  breeze
 as the reductant.

      Kalla and Steffen (1)  present an interesting display  of the distribution
 of process types used in  sponge iron production (July  1977).   See  Figure 2.
 Note that most of the world's sponge iron  is produced  by gas reduction
 processes.

      As already noted, there are two broad classes of  direct reduction
 processes; i.e., gas reductant systems and solid  reduction systems.  A
 further sub-division is normally made in the literature according  to the
 type of furnace; i.e., shaft, rotary kiln, or fluidized bed.   Individual
 direct reduction systems  will be described in the following section.

 Systems

      A summary of sponge  iron production capacity by the type  of process
 is  presented  in Table 3.

      The  gas  reductant systems  that  use a  shaft vessel  as  the  reactor  are
 the Midrex  (Korf Industries), Purofer (Thyssen),  Armco (Armco  Steel),  and
 Wiberg  processes.   These  are all  moving bed  reactors;  i.e., the solid  feed-
 stock moves in  a  direction  countercurrent  to the  flow  of reducing  gas.
 The processes  differ mainly in  the way the reducing gas is produced and
 circulated and  the  shape  of the  shaft.  Also a  major difference in the
 Purofer process  is  that the product  is  discharged hot  into a briquetting
system whereas  all  the  other processes  cool  the product before discharge.

     The HyL  (Hojalata  y  Lamina)  process uses retort reactors.  The process
is a batch operation  and  the charge  is  prereduced,  reduced, cooled, and


                                      10

-------
     Figure 2.  Types of Direct Reduction processes (1)
(Reproduced with permission of Iron and Steel  International.)
                             11

-------
TABLE 3.  SPONGE IRON PRODUCTION CAPACITY BY PROCESS TYPES (1978)
   Gas Reductant
      (80.2%)

   Shaft Furnaces (38.7%)
      Midrex (27.4%)
      Purofer (7.6%)
      Armco (3.1%)
      Wiberg (0.6%)

   Retort Furnaces  (26.7%)
      HyL (26.7%)

   Fluidized Bed Furnaces (10.7%)
      HIB (7.0%)
      FIOR (3.7%)

   Others (4.1%)
Solid Reductant
    (19.8%)

Rotary Kiln (16.0%)
   SL/RN (7.9%)
   Japanese (6.7%)
   Krupp (1.4%)
Retort (2.5%)
Others (1.3%)
                               12

-------
discharged as separate operations.

     There are two commercial  fluid  bed  processes,  the  HIB  (high  iron
briquette - U. S. Steel)  process  and the FIOR  (fluidized  iron  ore reduction  -
Esso Research) process.

     The main process using  solid reductants is  the rotary  kiln.   The  rotary
kiln systems used on an industrial scale are the SL/RN  (Stelco, Lurgi,
Republic Steel, National  Lead) and Krupp processes.  The  iron  containing feed
material is either lump ore  or pellets.   This  differentiates the  two pro-
cesses from the Japanese  rotary kiln reactors  that  use  pelletized steel pi ant
dusts from the feedstock; e.g., Kawasaki  process, Nippon  Steel's  Koho  process,
and Sumitomo Metals SPM and  SDR processes.

     In March 1978 a rotary  kiln  facility was  started up  in Rockwood, TN.
The process (Azcon Corporation) utilized pellets  as  the iron source and coal
as both the energy and reductant  sources.

     Another process that utilizes a rotary kiln  is  the ACCAR  (Allis Chalmers
Controlled Atmosphere Reduction)  process.  Oil and  natural gas are the heat
and reducing sources.

     A few other low production solid reductant systems are in operation;
e.g., the Kinglor-Metor and  the Hoganas  processes.  The Kinglor-Metor
process is a shaft furnace that uses coke, coke breeze, or lignite as the
reductant and external heating as the energy source.  Muffle furnaces are
used to reduce batch charges of ore  and  coke breeze  in the Hoganas process.

Gas Reductant Systems

     More than 80 percent of the world's  sponge iron is provided by gas
reductant systems.  About half of this amount  is produced in shaft-type
furnaces; about a third is produced  in retort  furnaces.  The remainder is
produced in fluid-bed reactors.

     The feed material to gas  reductant systems can be lump ore, pellets, or
ore fines.  The shaft furnaces require lump ore, pellets,  or briquetted
fines.  The fluid bed processes require ore fines.  The ore feed size
requirements are depicted graphically in  Figure 3.

     The feed material to a direct reduction plant is carefully controlled;
not only the feed material particle  size distribution but  also  the feed
chemistry.  Each plant operation has its own chemical specifications  for  the
material it buys.  The specification is influenced by not  only  the type of
direct reduction reactor system but also by the subsequent end  use.  An
illustrative specification is an iron content of at least  65 percent,  a
gangue content (Si02+Al203)/Fe ratio < 5 percent; and a phosphorous and
sulfur content < 0.05 percent  and < 0.01 percent, respectively.

     The Midrex Process

     From published literature it appears that the Midrex  process  is  being

                                     13

-------
            too
             001
                            Groin sin (mm)
       Figures.   Feed  material  size, distribution  (1)
(Reproduced with permission of Iron and Steel International.)
                               14

-------
used at 10 locations.  The companies using the process, their location,
and production capacities are given in Table 4.  About 7 million tonnes
per year of sponge iron is produced by this process.  Midrex (12) predicts,
based on present contracts and agreements, that the production capacity
by 1981 will be 17 million tonnes/year.

     The largest single gas reductant furnace in the world is now on
stream at Contrecoeur, Canada; it is a single reactor with a rated
capacity of 600,000 tonnes/year  (9).  This increases the production
capacity of Sidbec-Dosco's two reactors to over 1 million tonnes per year
of direct reductant iron.

     Detailed process information is available on the two Midrex facilities
in the United States (13-16), the Sidbec-Dosco facility in Canada (17-
19) and the Dalmine Siderca, Sidor, and Acindar plants in Latin America
(19). All of the plants are based on modules that have a 400,000 tonnes/year
capacity except the Sidbec-Dosco II facility (17) which has a 400,000
and a 600,000 tonne/year module.  Only a brief general description of
the process will be presented here.  See the noted references for further
details.

     The Midrex standard flowsheet is presented in Figure 4.   The process
consists of a reduction shaft furnace and a gas reformer plant.   The
shaft furnace is 16 feet (4.88 meters) in diameter and is rated at
400,000 tonnes/year.  The reactor is divided into a reduction zone and
a cooling zone.  In the reducing zone iron oxide is reduced as  it moves
countercurrent to the injected reducing gas.  The reactions of  interest
are the reduction reactions and  the formation of iron carbide reaction:
                    3 + 3(CO+H2) -»-  4Fe + 3(C02+H20)

               (reduction)              (800-1 000°C)

               3Fe + CO + H2  -*•  Fe3C + H20

               (carburizing)

     The product gas exits the furnace near the top.  It is cooled,
cleaned by a scrubber to remove the particulates, mixed with new natural
gas, and passed to a reformer to form more CO and H2-

     The reformer furnace contains long alloy tubes heated externally.
The reactions of interest are:

               CH4 + C02  ->  2CO + 2H2

                             (11 00-11 50°C)

               CH4 + H20  ->  CO + 3H2

The top gas from the reducing furnace is the source of the required oxi-
dants; additional air or steam is not required for the reforming reactions
Some of the top gas which contains about 75 percent CO + H2 is used as


                                     15

-------
                    TABLE 4.   MIDREX PROCESS (1,4,12,20)
   Operator
Oregon Steel
Georgetown Ferreduction
Hamburger Stahlwerke
Sidbec-Dosco I
Sidbec-Dosco II
Dalmine Siderca
Si dor  I
Sidor  II
Acindar
NISIC
Qatar  Steel
British Steel
  Location
Portland, OR
Georgetown, SC
Hamburg, West Germany
Contrecoeur, Canada
Contrecoeur, Canada
Campana, Argentina
Mantazas, Venezuela
Matanzas, Venezuela
Villa, Argentina
Ahwaz, Iran
Doha, Qatar
Hunterston, Gr. Britain

         Total (1978)
Capacity, tonnes/yr
  400,000 (1969)*
  400,000 (1971)
  40Q,000 (1972)
  400,000 (1973)
  600,000 (1977)
  330,000 (1976)
  400,000 (1977)
1,200,000 (1978)
(3-400,000 plants)
  420,000 (1978)
1,200,000 (1978)
(3-400,000 plants)
  400,000 (1978)
  800,000 (1978)
(2-400,000 plants)

6,950,000**
* Start date
** The total capacity is based on published literature figures.  However,
   from private communications with J. Bradley, Midrex Corporation this
   figure should be 6,065,000 tonnes per year (August 29, 1979).
                                     16

-------
Process
Gas Mist   I   *
Eliminator
Oxide
Feed
                                                                               Cooling Gas
                                                                       Mist     Compressor
                                                                       Eliminator
                                                         Process
                                                         Gas
                                                         Compressor
                   Figure  4.  Midrex  standard flowsheet  (18).
                  (Reproduced with permission of Midrex  Corp.)

-------
fuel for firing the reformer chambers.

     Hot metallized sponge  iron descends  into  the cooling zone of the fur-
nace.   It  is cooled by a closed loop of circulating  inert or natural gas.
The cooled iron product is  discharged continuously at about 30-50 C.

     The iron  product may be further treated to  prevent  it from  reoxidizing
during  storage.  Midrex has developed processes  that are used for passivating
the product; e.g., a controlled slow reoxidation of  the  iron surface and a
cold briquetting process that  reduces the total  surface  area so  that reoxi-
dation  is  of minor importance.  However,  in both cases the passivated
product must be stored in areas protected from rainwater and/or  seawater
during  transportation.

      It is of  interest to compare  the energy requirements of gas reductant
systems.   Kalla and Steffen (1) report the results shown in Table 5.  The
Midrex, Purofer, and Armco  processes require about the same energy, with
the Midrex requiring slightly  less than the other two.   The Midrex 600,000
tonne/yr  (Sidbec II) plant  has reported even lower energy use, 2.5 Gcal/
tonne  Fe for gas consumption usage (21).

     The HyL Process

     The Hojalata y Lamina  S.A. (HyL) process  is currently (end  of 1978)
being  used at  nine locations.  The companies using the process,  their loca-
tion,  and  production capacities are given in Table 6.  About 7 million
tonnes  per year of sponge iron is  produced by  this process.  Projections (4)
are that capacity will be increased to about 9 million tonnes/yr by 1981.

     Detailed  process information  is available on the HyL facilities in
Mexico  (22-24).  Only a brief  general description will be presented here.
See the references for further details.

     The HyL process operates  on a fixed  bed principle.  The ore remains
stationary in  a closed retort.  The gas phase  is changed according to the
operation  to be performed in the reactor;  i.e.,  drying,  preheating, reduct-
ion, and cooling.  A flow sheet for the process  is shown in Figure 5 (7). A
process module consists of  four reactors.  Each  reactor  is about 17 feet
(5.18 meters)  in diameter and  49 feet (14.9 meters)  in height.   Three of the
four reactors  are always connected in series (with respect to gas flow) while
the fourth is  loaded or unloaded.   The steps in  the  reduction cycle consist
of first reactor - loading  and unloading,  second reactor - initial reduction,
third reactor  - final reduction, and fourth reactor  - cooling and carburiz-
ing.

     Each  retort is discharged every 9-12  hours.  The sequence of treatment
occurring  in each reactor is:  fresh reformed gas is  used to cool the reduced
product; it then passes on  to another retort where final reduction takes
place;  and gas from this reactor is passed through a third reactor where
preheating of  the charge and preliminary  reduction occurs.  Gas  from this
stage of the operation still contains enough CO  + H« to  be used  as fuel
for firing the reformer combusters.  Note  that the product gas from the last


                                     18

-------
TABLE 5.  ENERGY REQUIREMENTS FOR THE GAS REDUCTANT SYSTEMS (1)
     Reducing Agent
            Gross Energy Consumption      Total
Gas                Electricity         Equivalent
Process
Midrex
HyL
Pruofer
Armco
FIOR
Blast Furnace
H2
53
75
47
68
75
(25)
CO
35
14
45
20
14

C02 H20 (CH4+N2)
2 5
8
2 3
2 9
8

5
3
3
1
3

(Gcal /Tonne Fe)
3
4
3
3
4

.1
.6
.3
.4
.0

(kWh/tonne Fe)
155
15
130
35
150

(Gcal /tonne Fe)
3
4
3
3
4
2
.2
.6
.4
.4
.1
.8

-------
                   TABLE 6.  HyL  PROCESS  (1,2,4,22)
Operator
HYL I
HYL II
HYL III
TAMSA
HYL I
HYL II
USIBA
SIOOR I
SIDOR III
SIDOR IV
NISIC III
Krakatau-Ferrosteel
Tika
Iraq
Location
Monterrey, Mexico
Monterrey, Mexico
Monterrey, Mexico
Vera Cruz, Mexico
Puebla, Mexico
Puebla, Mexico
Bahia, Brazil
Matanzas, Venezuela
Matanzas, Venezuela
Matanzas, Venezuela
Ahwaz, Iran
Kota Baja, Indonesia
Solwezi , Zambia
Khor El Zubeir, Iraq
Capacity, tonne/yi
95,000 (1957)*
260,000 (1960)
475,000 (1974)
280,000 (1967)
315,000 (1969)
625,000 (1977)
250,000 (1974)
400,000 (1977)
350,000 (1977)
2,000,000 (1978)
400,000 (1978)
500,000 (1978)
300,000 (1978)
600,000 (1978)
* Start date
                                     Total  (1978)
6,850,000
                                  20

-------
    Fuel Gas
 o
O
Reactor A
 Loading
  and
Unloading
 o
O
Reactor B
  Initial
Reduction
        Legend

           Dehumidifier
           Gas Heater


    • • k • •  Gas Flow
o
O
Reactor C
  Final
Reduction
                              \
                         Reformer
                                      Dehumidifier
o
O
 Reactor D
 Cooling
  and
Carburizing
                                                            • • • m • *
                    Figure 5.   The  HyL  process  (27).
    (Reproduced with  permission of  Iron and  Steel  Engineer.)
                                       21

-------
reactor is not recycled for reforming as  is the product gas from the Midrex
furnace.

     The HyL product is normally controlled at about 85 percent metallization
and contains up to 2.5 percent C.   It is  claimed that the product is nonpyro-
phoric  (22)  The energy requirements were reported by Kail a (Table 5) to be
4.6 Gcal/tonne Fe.  This is somewhat higher than the value reported by Labee
(22) for the HyL Puebla II plant; i.e., 3.4 Gcal/tonne product (which is
approximately equivalent to 4 Gcal/tonne  Fe).

     Other Processes

     The Midrex and HyL processes accounted for over 88 percent of the
metallized iron produced by gas reduction processes in 1978.  The other
gas reduction processes, their production capacities, and locations are
presented in Table 7.

     The Purofer (26) and Armco (27) processes are both shaft furnace pro-
cesses  that are similar in principle and  design to the Midrex process. Both
processes involve the flow of solids countercurrent to the flow of reducing
gas.  The feeds are similar to the Midrex and the results of metallization
essentially the same, 92-94 percent.  The differences are that the Armco
process does not recycle the reducing gas and the Purofer process discharges
its product hot.  The hot product is briquetted before release and is there-
fore nonpyrophoric.  Another difference is that the Purofer process can use
either  oil (gasified) or natural gas as the reductant.

     Two other gas reduction processes use fluid-bed reactors; i.e., the HIB
(high iron briquette) process (28) and the FIOR (fluidized iron ore reduc-
tion) process (29).  Both are U.S. developed processes but commercialization
has taken place in foreign countries; see Table 7.  The FIOR process produces
a product that is 88-93 percent metallized whereas the HIB process is design-
ed for  the product to be used in ironmaking rather than steelmaking, hence
its metallization is much lower; i.e., 70-75 percent Fe.  The products from
these processes are a fine powder that has been produced at low temperatures
and is, therefore, pyrophoric.  The product is briquetted to prevent
reoxidation during storage.

Solid Reductant Systems

     Approximately 20 percent of the world's sponge iron capacity is provided
by solid reductant systems (1978).  Most  of the capacity is in rotary kiln
reactors (16%) and a small amount is retorts (2.5%) and vertical shaft
furnaces (1.3%).  The solid reductant systems, their production capabilities,
and their location are presented in Table 8. (There are several discrepan-
cies in the production figures given by references 1, 2, and 4.) In addition
to those processes listed in Table 8, a number of rotary kiln furnaces
treat steel  plant dust, see Table 9.

     Approximately 90 percent of the production from solid reductant systems
is by rotary kiln processes, 2,170,000 tonnes/yr from virgin ores and
1,725,000 tonnes/yr from mill wastes.  According to Miller (4), "There has


                                     22

-------
Purofer
Purofer
ARMCO
FIOR
HIB
                   TABLE 7.   GAS REDUCTANT PROCESSES
              EXCLUDING MIDREX AND HYL (1978)  (1,2,4,22)
Process      Operator
COSIGUA
NISIC
ARMCO
Fior de Venez
MONORCA
                     Location
Purofer      Thyssen-Purofer      West Germany
Brazil


Iran


USA


Venezuela


Venezuela
                               Total  (1978)
Capacity, tonnes/yr


  150,000 (1970)*


  350,000 (1977)


  330,000 (1970)


  330,000 (1972)


  400,000 (1976)


  650,000 (1973)





2,210,000
*Start date
                                   23

-------
            TABLE 8.  SOLID REDUCTANT SYSTEMS  (1978)  (1,2,4)
Process
Location
Reactor Type
Capacity, tonnes/yr
Hoganas
Hoganas
SL/RN
SL/RN
SL/RN
SL/RN
Krupp
High veld Kiln
Highveld Kiln
Hockin Kiln
Kinglor-Metor
Sweden
USA
New Zealand
Brazil
Canada
Peru
S. Africa
S. Africa
S. Africa
USA
Italy
Muffle
Muffle
Rotary Kiln
Rotary Kiln
Rotary Kiln
Rotary Kiln
Rotary Kiln
Rotary Kiln
Rotary Kiln
Rotary Kiln
Shaft
170,000
70,000
120,000
60,000
360,000
100,000
150,000
1,000,000
300,000
100,000
40,000
                                      Total  (1978)
                                       2,470,000

-------
TABLE 9.  ROTARY KILN PROCESSES  FOR TREATING STEEL PLANT WASTES  (1,2,4)
      Process

      SL/RN

      Sumi tomo
      Kawasaki
      NSC
  Location

Fukayama, Japan

Wakayama, Japan
Kashima, Japan

Mizushima, Japan
Chiba, Japan
Hirohata, Japan
Capacity, tonnes/yr

     490,000

     340,000
     290,000

     310,000
      85,000

     210,000
                                   Total  (1978)  1,725,000
                                   25

-------
 been no general acceptance of the rotary kiln processes.   Solutions for
 frequent electrical  and mechanical  breakdowns and for complex operating
 difficulties have been neither easy nor consistently successful,  and start-
 up periods have been longer than planned."

      It should be noted that the SL/RN plant in Canada (30)  has been shut
 down since May 1976 (4).  An additional facility, not listed in Table 8,
 uses a mixture of coal and oil;  i.e.,  the Allis Chalmers  Controlled Atmos-
 phere Reduction (ACCAR) rotary kiln (50,000 tonnes/yr).   A second ACCAR kiln
 is located at Sudbury, Canada (capacity 340,000 tonnes/yr) but the reductant
 and fuel are a mixture of oil and natural  gas.   It has been  out of operation
 since October 1976 (4).

      According to Miller (4) there  appears to be reason for  some optimism for
 rotary kiln processes: "There has been an increasing number  of encouraging
 reports since 1975 from the modified SL/RN unit in New Zealand, the Krupp
 plant in Benoni, and the Highveld operation in  South Africa, and  especially
 from several kiln operations with steelplant waste-material  charges in Japan.
 The sponsors of coal-based processes believe, therefore,  that the break-
 through by their designs is very close."

      Another recent development  also lends optimism to the use of coal  based
 rotary kiln processes; i.e., the Western  Titanium, Ltd. coal  fired Hockin
 process (32).   This  process was  developed in Australia and is now being
 demonstrated by the  Azcon Corporation  at  its Rockwood, TN plant.   It is a
 rotary kiln process  that uses coal  for both the reductant and the fuel.   No
 other fuel  is  required.

      A comparison of the energy  requirements for solid reductant  systems is
 presented  in Table 10.   To compare  these  values with gas  reductant systems
 refer to Table 5.  The average energy  requirement per tonne  of iron produced
 by direct  reduction  (all  processes  included)  falls within the range
 3.8 + 1.0 Gcal/tonne  Fe.   It has  been  reported  that the production of  pig
 iron  by the blast furnace uses approximately 2.8 Gcal/tonne  Fe (1973 figure,
 Kono  (25)).  This  energy value is for  Japanese  practice and  is, likely,
 considerably less  than  United States energy consumption.   According to
 Dailey  (33), the best  North American blast furnace practice  (1972)  required
 5  Gcal/tonne Fe.   Depending on the  basis  of comparison, one  can state  that
 direct  reduced  iron can  be (a) produced at a lower energy input than pig iron
 (e.g.,  any  direct  reduction practice compared to "best" North American  blast
 furnace  practice)  or  (b)  produced at essentially the same energy  input
 required for "best" blast furnace produced pig  iron (e.g., Japanese blast
 furnace  practice compared to any direct reduction practice).   At  one time
 or  another  all  of  these  claims have been  reported in the  literature.  Often
 the assumptions  used  in  making the  comparisons  are not clearly stated and
care must be exercised  in using  the results.

     The SL/RN  Process

     Rotary  Kiln  processes are similar in  principle.  Kilns are long cylin-
drical refractory  lined  vessels  that can  be rotated about their longitudinal
axis.  The  vessel  is tilted at about a 3  percent slope.   Feedstocks

                                     26

-------
      TABLE  10.   ENERGY REQUIREMENTS FOR SOLID REDUCTANT SYSTEMS
                     Total Equivalent Energy Consumption
Process	Gcal/tonne Fe	        Reference
SL/RN
Krupp
Hockin
Kawasaki
Accar
Coal /Oil
Oil
Gas
3.7
3.8
3
3.5
2.8
3.0
3.5
- 4.8
- 4.8
.2
- 4.2
- 3.3
- 3.3
- 3.8
1, 34
1, 34
32
34
35
35
35
 Kinglor-Metor                       4.2                          5, 36
                                    27

-------
(usually lump or pelletized ore), fluxes, and coal are fed in the high end of
the vessel.  The ore is preheated and reduced as it is moved by the vessel's
rotation and gravity.  The reduced material is discharged continuously from
the lower end of the kiln into a cooling or briquetting system.

     The rotary kiln processes differ mainly in design of the heating system
and in the selection of operating conditions such as type of ore, type of
coal, feed rates, gas and solid temperatures, and oxygen potential.

     The SL/RN process will be briefly described here as an illustration of
rotary kiln processes.  Further detailed descriptions are readily available
in current literature for all the rotary kiln processes (11,20,28,30,37,38).

     A flow diagram (7) is presented in Figure 6.  The selection of feed
materials is very important.  The reactivity of the ores is carefully tested
before they are accepted for use in the reactor.  The gangue content,
base/acid ratio, sulfur, phosphorus, alkaline metals, and heavy metal conten-
are all specified and tested before an ore is purchased.  The physical
properties are also important; e.g., strength, size.  "Contrary to a
misconception that is more widespread than it should be, direct reduction
will not work with  'any' ore" (7).

     The type of coal is also specified and tested before acceptance for use
in a solid reductant kiln.  It must have a proper reactivity, a specified
low sulfur, phosphorus and volatile content, and a minimum fusion temperature.
A coal that does not have a proper fusion temperature can agglomerate and
stick to the reactor walls and rather quickly shut down the reactor.

     The kiln temperature is controlled by a series of burners along the
length of the kiln.  Ore and coal are charged into the kiln (up to 100 meters
long).  Limestone and dolomite may be included if the sulfur content is to
be maintained at low levels.

     The solid mixture is heated to about 1000°C (this varies depending on
ore and fusion temperature of the coal) as it moves along the kiln.  The
residence time is 3-4 hours and the metallization is from 92-95 percent.  The
discharged product is cooled and screened, and the iron is magnetically
separated from the coal char.

     The waste gases are further burned in an afterburner or are scrubbed
and the waste gas flared.

     Other Sol id Reductant Systems

     There are a few nonrotary kiln processes; e.g., the Hoganas and Kinglor-
Metor (39) processes.  The Hoganas process is based on the use of muffle
furnaces and the Kinglor-Metor process is an externally heated shaft furnace
system.   Neither process is considered suitable for large scale development
and, therefore, will not be discussed here.
                                     28

-------
TO
                           "IRON-ORE CONCENTRATES
                      e=g  _^ PELLET1ZATION
                                      rr
                    EXHAUST GAS
                                    HARDENING AND
                                   OXIDATION OF PELLETS
                      DUST
                                                   EMERGENCY ,
                                                     STACK
                                                                                            AIR
                                                            ....... ..,........  .....-
                                                      \iiJ2J .>-.:_••_...-.. ROTARY KILN. ••.- V '. • "I L
                                                           .^....-.:x:-.:.:..:v..- .. ...
                                                                                          COAL
                                                           WATER
rrOi   IP
                                                                    IT
                                                           SCREEN
                                                  MAGNETIC SEPARATOR
                                   RECOVERED COAL CHAR
                                                          COAL SEPARATOR
                  WATER
          MAGNETIC SEPARATOR
                                                                                   SPONGE IRON
                                                                               ASH AND RESIDUES
                                   Figure 6.  The SL/RN process (7).

-------
Direct Reduction in the United States

     The near future prospects for U. S. adoption of direct reduced iron-
making is at this time questionable.  The present direct reduced iron
capacity in the United States (4) is 1.1 million tonnes per year.  This
production capacity is projected (4) to increase to 4.32 million tonnes per
year by 1985 (Table 2).  About 2.4 million tonnes per year of the projected
increase is based on natural gas reduction units.  The proven reserves in
the lower 48 states amount to 200 trillion cubic feet (Tcf).  Our current
rate of consumption is 19 Tcf per year.  Therefore, unless the quantity of
proven reserves increases dramatically, a supply of only 10 years exists.
Even if an optimistic view is taken that the reserves can be doubled by
acquiring gas from Alaska and Mexico and/or by production of synthetic gas
from coal, only a 20 year supply at current consumption rates could be
possible.  Therefore, coal or coal derived products appear to be the only
fuel type that can be realistically considered for direct reduction use (40,
41).

     Hayes (42) estimates that by 1985 the United States production of coal
will be 818+18 million tonnes per year.  The reserves of coal are estimated
by the Bureau of Mines to be 397 billion tonnes (42).  Of the available coal,
Hayes estimates that 227 billion tonnes is recoverable.  It, therefore,
appears from the projected supply and demand estimates that coal will  be an
energy form available for many years.

     Coal-based direct reduction systems have been discussed previously in
the section, "Solid Reductant Systems."  There are commercial plants present-
ly producing iron.  Recent progress in the technology of solid reductant
processes, particularly rotary kiln reactors, was reviewed at a recent Office
of Technology Assessment (O.T.A.) Seminar on New Techniques in Steelmaking
(41).  The presentations on the SL/RN (43) and ACCAR (44) rotary kiln
processes suggested that the major technical  problems experienced in opera-
ting rotary kilns have been solved.

     The successful demonstration of a new technology is an important step
toward the adoption of that technology by industry.  But the successful
demonstration of new technology does not ensure that it will be adopted even
if it is shown to be superior to the old technology (45).  Consideration of
the factors that impact on the near future adoption of direct reduction
technology in the United States is beyond the scope of the present project;
e.g., impact of foreign steel imports, availability and price of scrap,
expected replacement of old facilities, and need for new capital formation.
However, these factors are being considered in a comprehensive Office of
Technology Assessment study, "Impact of Technology on the International
Competitiveness of the United States Steel Industry" (46).   The results of
this study will be available in the Fall, 1979.  Conclusions of the study
and resultant policy decisions may have an important impact on the near
future development (or nondevelopment) of direct reduction ironmaking in
the United States.

-------
                                   SECTION  3

                         POTENTIAL  POLLUTION  PROBLEMS


     One of the attractive  features  of the direct  reduction route to iron-
making is that the processes  can be  effectively controlled to prevent major
emissions to the environment.  The following discussion outlines several
potential pollution problems  but each  of the potential problems can be
eliminated by good process  design  and  control.

Gas Reductant Systems

     The shaft furnace  gas  reductant systems produce some particulate mater-
ial.  The origin of the particulate  material is feed deterioration brought
about by abrasion and reduction reactions.

     The gas exiting the shaft furnace contains unreacted CO, H2, CH4,
reacted products, C02,  H20, and dust particles.  This gas is normally treated
in wet scrubbers and the cleaned gas is recycled to the gas reforming unit
and subsequently back to the  reactor furnace.  The particulate solids in the
scrubber waters are recovered and  then briquetted, pelletized, or disposed
of in solid waste storage areas.

     It is important to note  that  gas  reductant systems operate under condi-
tions that are unfavorable  for formation of  polynuclear aromatic hydrocarbons
(PAH); e.g.,

     1.  Low order hydrocarbons are  used in  the reforming process (Cfty)  and
         they are not likely  to combine to form the high order PAH's.
     2.  Reforming is conducted by surface catalyst techniques which do  not
         promote formation  of high order molecules.
     3.  The reduction  reactor is  operated at temperatures (=950°C)  below
         which PAH's form.

     Even if PAH's are  formed in the reduction furnace they probably are
effectively removed from the gas streams along with the scrubber sludge.  The
fate of PAH's in the scrubber water  is not known.  It is suspected that  some
of the PAH material will dissolve  in the aqueous phase (the low molecular
weight compounds) and some will remain with  the solids either as condensed
particulate organic solids  (the high molecular weight compounds) or as  ad-
sorbed species on the solids.  That  fraction of the PAH material that is not
removed by scrubbing should not be a problem because the gas stream is
either recycled to the  reduction furnace or  is combusted in a separate
chamber.


                                     31

-------
     Even if PAH's are formed and removed by scrubbing, they would most
likely not be a problem at those facilities that recycle their solids to the
direct reduction furnace and their water to the scrubber system.  Some faci-
lities do not recycle the sludge solids but dispose of them in solid waste
storage areas.  The final fate of any PAH material (if it even exists) in the
solid waste is not known.

     As is true at all facilities that have ore storage, wind blown dust is
a potential pollution problem.  The sponge iron product is not a pollution
problem because it is normally placed in protective enclosures or covered
with canvas to prevent reoxidation by exposure to moisture.

     The literature suggests that gas reductant systems can utilize the pro-
duct from a coal or oil gasification plant (47).  If this occurs, the poten-
tial pollution problem associated with gasification plants must also be
considered (48).

     Another associated source of potential pollution needs to be considered;
i.e., pelletization.  It is estimated (1) that 90-150 million tonnes per year
of ore, that has acceptable chemical composition for use in direct reduction,
is available.  However, about 70 percent of this material is fine-grained.
The fine-grained material is useful for fluid-bed applications but the size
range necessary for use in the fluid-bed reactors is fairly narrow.  There-
fore, a significant fraction of the ore will most likely have to be agglomer-
ated before it can be used.  Emissions from iron ore mining, beneficiation,
and pelletizing have been studied by Midwest Research Institute for the EPA.
Potential emission sources are discussed in reference (49).

     Note that pelletization of ore fines is not unique to direct reduction
feed materials:  it is also used extensively for preparation of blast furnace
feed material.  In fact, only about 3 percent of the pellets produced at
present are used in direct reduction processes (2).

Solid Reductant Systems

     The gas from a rotary kiln coal reductant system could contain CO, C02»
H2, H20, SOX, NOX, hydrocarbons, trace metals, and particulate matter.  Most
systems combust natural gas or oil above a coal/ore mixture to maintain the
desired temperature.  Two systems use coal both as the reductant and as the
combustion fuel; i.e., Azcon Corporation's Hoskin process reactors in
Australia and Rockwood, TN.  The effluent gas phase composition from rotary
kiln reactors is very dependent on the coal chemistry and mineralogy.
Stringent requirements must be imposed on the coal to limit both the sulfur
and trace metal content.  The particulate matter from these reactors is
usually high in flyash and char.  Some systems combust the exit gas in an
afterburner chamber to reduce the concentration of CO, H2, char, and hydro-
carbons in the exit streams.  The afterburner is normally followed by a
wet scrubber device.  The collected scrubber product may be dewatered and
the solids pelletized or briquetted and recycled to the kiln, but usually
they are simply disposed of in solid waste areas.

     The product from the kiln is screened and the iron is magnetically

                                     32

-------
separated from the char  and  other waste solids.   The other waste  solids are
calcium sulfate  (if the  feed contains  some sulfur,  it can  be  controlled by
adding lime to the charge) and  a  slag-like material  from the  coal   The
larger char material  is  recovered and  recycled to the kiln.   The  fine char
material is disposed  of  along with the slag and  gypsum in  solid waste storage
areas.

     The potential pollution problems  are:

     1.  The survival  of PAH's  formed  in the coal-ore bed.  It has been well
documented that  PAH materiel  is created and/or evolved during pyrolysis and
combustion of coal (50-62).   Coal  fired heating  and  power  generating plants
are estimated to contribute  over  30 percent of the BaP (Benzo-alpha-pyrene,
a member of the  Polynuclear  Aromatic Hydrocarbon group of  chemicals that is
known to be carcinogenic) emitted to the United  States air environment each
year (62), and over 45 percent  to the  world's air.   It is,  therefore, reason-
able to suspect  the rotary kiln processes of being potential  Polynuclear
Aromatic Hydrocarbon  emitters.

     The gas phase above the rotary kiln bed is  normally oxidizing and will,
therefore, be expected to combust the  organic material  if  the retention
times are long enough.  The  entrance region of the kiln appears to be an
ideal place for  the formation of  PAH compounds;  i.e.,  the  material is being
preheated and volatile constituents are being evolved;  soot is also being
evolved (in effect the material is smoldering).   The  PAH compounds evolved
into the gas phase will  be oxidized if the  retention  times  in the oxidizing
zone above the bed are sufficiently long.   However, near the feed end of the
kiln the retention times will be  very  short.   Also, it appears likely that
PAH compounds readily adsorb on fine particulate matter, in particular soot,
and that the PAH compounds are  protected and survive  if the retention times
are short.

     2.  The survival  of PAH's  from the combustion of  oil  and coal above the
coal/ore mixture.  The survival of organic  compounds  evolved during the com-
bustion of the fuel is expected to be  very  low.   The  kiln  gas temperature
(32) varies from approximately  1200°C  at the burner end to  500°C at the gas
exit end.  An exposure at greater than 900°C for at least  0.3 sec is neces-
sary for destruction  of  the  PAH material.   The gas flow rates generally allow
an average residence  time in the  reactor of from 5 to  9 sec.  Therefore, the
destruction of most combustion  formed  PAH material is  expected to occur
during steady state operation.  However, emission of  PAH compounds during
start-up and shut-down periods  is expected  to be considerable.

     As stated earlier,  some rotary kiln processes treat the  kiln gases in
an afterburner and then  scrub the gas  in wet scrubbers.  Therefore, the PAH
material that survives the combustion  is most likely  collected in the scrub-
ber sludge.  Its fate in subsequent processing or in  solid disposal sites is
uncertain at this time.

     The Minimum Acute Toxicity Effluent (63,64) value for BaP disposal to
land is 0.006 yg/g.   This is not  a regulation-restricted emission value but
indicates minimum concentrations  suspected  of potential  health effects.  One

                                     33

-------
should be aware that this value  is considered controversial because the natu-
ral background of BaP in soils and sediments is often quoted to be much
greater (65-67).

     3.   Trace element and sulfur oxide emissions from the coal fired react-
ors.  Reactors (only one in U.S. at present) that combust coal above the
surface of a coal/ore mixture emit some of the sulfur and most of the trace
elements from the rotary kiln.  A portion of these elements are removed by
the attached control devices.

     The sulfur content in the feed coal is specified to be low; i.e.,
usually less than 0.5 percent.  Example calculations using the operating data
presented by Cassidy and MacKay  (32) for a coal fired rotary kiln (assuming
0.5 percent sulfur in the feed coal and a heating value of 10,000 Btu/lb
coal) yield a sulfur emission factor of 0.5 Ib sulfur/MM Btu.  In terms of
sulfur concentration, emission is 0.6 grams per cubic meter.  New Source
Performance Standards (NSPS) have not been considered for direct reduction
reactors.  However, if we assume that regulations similar to the proposed
NSPS for electric utility steam generating units (68) apply to a coal fired
direct reduction reactor, then emission levels above 0.2 Ib sulfur/MM Btu
would have to be controlled to meet an 85 percent reduction in input sulfur
content.  About 20 percent of the inlet sulfur will be associated with the
slag phase (69) exiting the reactor and should be effectively removed from
the gas stream by most particulate control devices.  A portion of the sulfur
will be "gettered" from the gas stream while still  in the reactor by lime in
the ore bed.  The fraction extracted by this mechanism cannot be estimated
because the reaction is dependent on such kinetic factors as efficient gas/
solid contact and residence time.  It is anticipated, however, that sulfur
oxide emissions can be controlled without special add-on sulfur removal
equipment.

     Sulfur content in the coal used as a reductant in the coal/ore bed has
been shown to be effectively controlled by the addition of lime or dolomite
to the charge mixture (32).  Gypsum is formed and can be separated from the
sponge iron and disposed of in a solid landfill site.

     The partitioning behavior of trace elements during coal combustion has
been studied (59).  The fate of trace elements falls into three general
classes: those normally partitioned to the slag phase are Al, Ba, Ca, Ce,
Co, Eu, Fe, Hf, K, La, Mg, Mn, Rb, Sc, Si, Sm, Sr,  Ta, Th, and Ti; those
normally partitioned to the fly ash are As, Cd, Cu, Ga, Pb, Sb, Se,  and Zn;
and those normally remaining in the gas phase are Hg, Cl, and Br.  As and Cd
are usually more concentrated in the fly ash particle sizes less than 10
micrometers, so concentration occurs in the size fractions that are least
effectively removed by control devices.  The concentration of trace elements
varies in coals and the release of trace elements during combustion varies
greatly (60,61).   It is certain, however, that great care should be exercised
in designing a control system to include good removal of fly ash particles
which, most likely, will  contain toxic material.   National Emission Stand-
ards for Hazardous Air Pollutants (NESHAP) are being considered for As, Cd,
and Pb (59).   Proposed standards have been generated for Be (63).
                                    34

-------
     For example,  if  one  assumes that the concentrations  of trace metals
given by Thompson  and Harrison (59)  are representative of coal  fuels  and
that values for  combustion  coal  feed rates given by Cassidy and MacKay  (32)
are appropriate, then uncontrolled emissions of As, Cd, Pb, and Be can  be
ca I cu I ated .

     The concentrations of  trace metals in low sulfur  coals are:

              As        9.3 -  13 ppm

              Cd        0.1 -  2.4 ppm

              Pb        8.3-12 ppm

              Be        1.2 -  2.8 ppm

     The process feed rates are:

               Combustion  coal                 0.70 tonnes/hr
               Air (At Standard Conditions    4980 m3/hr
                    0°C,  1 Atm.)

     The calculated trace metal  uncontrolled emission  values are:
              As         1085  -  1517

              Cd           12  -  280

              Pb          968  -  1400

              Be          140  -  326

     Control devices will  decrease these concentration levels, and dispersion
dilution will further decrease  the concentration of trace elements emitted to
the environment.  The NESHAP  proposed standard  for Be is 0.01 yg/m3 as the
outplant concentration maximum;  therefore, the effect of the control device
and dispersion dilution  must  be  to decrease the Be concentration at least
14,000 times.

Recommendations

     The gas reductant systems appear to be environmentally clean processes,
particularly those that  recycle  the reductant gas.  It is recommended that
samples of scrubber effluents be collected and characterized at least through
the E.P.A. Source  Assessment Sampling System (S.A.S.S.) Level 1 analytical
scheme.

     The rotary kiln solid reductant systems are of particular interest
because of the large U.S. coal  reserves.  All systems that involve coal and
oil combustion have been  shown to be potential PAH, sulfur oxide, and trace

                                     35

-------
metal emitters.  It is recommended that an initial sampling be made to ascer-
tain the uncontrolled and controlled emission rates for PAH, SOX, and trace
metals.  As an initial effort, a preliminary sampling program should be
performed to provide mass balances on these effluents from the rotary kiln
and scrubber system.

     The only coal fired rotary kiln process in the United States is the
Azcon reactor at Rockwood, TN.  At present it has neither an afterburner
nor scrubber.  Sampling of this system for kiln emissions would be desirable.
                                     36

-------
                                   SECTION  4

                    STEELMAKING  USING METALLIZED PRODUCT


Direct Reduction - Electric Arc  Furnace  (DR-EAF) Steelmakino

     The major use of sponge  iron  produced by direct reduction processes is
/5A!?"™6 ?f 1>0n in electr1c  arc  steelmaking furnaces.  Two mini mills
114,16) ln the U.S., Georgetown  Steel and  Oregon Steel, use a substantial
fraction of sponge iron  in their charge.   Both use a charge of sponge iron
and scrap, about 50 percent of each.

     If the optimistic projections hold  — that 10 percent (100 million ton-
nes) of the world's iron will be sponge  iron (7) by 1985, and that 30 percent
(300 million tonnes) of  the world's  steel  will be produced by the electric
arc steelmaking process  (70)  —  significant further growth in the DR-EAF
process systems can be anticipated.

     The literature quotes a  number  of advantages for the DR-EAF combination
process.  Several quoted advantages  are  presented below:

     1.   A variety of excellent processes have been developed and proved in
          commercial operation.  These are based upon the moving-bed shaft
          furnace, the fixed  shaft furnace, fluidized beds and the rotary
          kiln.  All represent good  technology for tonnages of 900 to 1800
          tonne/day.  Additional plants  can be added as needed, and they can
          be used with no initial overcapacity as would occur for a modern
          blast furnace-coke  plant-BOF complex (5).

     2.   For socio-economic  reasons, developing nations desire domestic
          steel industries; limited  initial tonnages are the guideline and
          direct reduction provides moderate scale production at moderate
          capital investment.  Moreover, many of these nations having
          reserves of high-grade iron ore  are deficient in metallurgical
          coals, but have abundant supplies of natural  gas and petroleum.
          These latter forms  of  reductant  and energy can be used in -only
          limited quantities  in  blast furnaces but are ideal  for many
          direct reduction processes (5).

     3.   In developed nations,  opportunities occur for mini- and midi-steel
          mills with direct reduction product supplementing local scrap.
          These applications will be hampered in the future by high costs and
          shortages of natural gas or petroleum from which reducing gases can
          be produced, but processes for economically gasifying coals will


                                     37

-------
    doubtlessly be a favorable future factor (5).
4.  The gross consumption of energy per ton of liquid steel is compara-
    ble for the method of combining direct reduction and the electric
    furnace with the blast furnace and the basic oxygen furnace (7).
5.  The time required for commissioning a DR-EAF steel plant is about
    two to three years compared to five to seven years for an integrated
    steelworks (9).
6.  Miller summarized a report by the World Bank comparing DR-EAF to
    the blast furnace-basic oxygen furnace (BF-BOF):
    a.  The quality of common steels produced by DR-EAF and BF-BOF
        practices was approximately equal;
    b.  The estimated capital costs of facilities for pelletizing, iron-
        making, steelmaking, and continuous slab casting of 3-4 million
        tonnes/year at Matanzas favor a DR-EAF plant by nearly 40 per-
        cent over a comparable BF-BOF installation;
    c.  The production cost, excluding fixed charges and income taxes,
        for a tonne of carbon steel slabs is approximately 20% lower
        when produced by a DR-EAF operation than when made by the BF-BOF
        route;
    d.  The average rate of return on investment is from 2.5 to 3 times
        as great for the DR-EAF as for the BF-BOF (4).
7.  Korf extols other attributes of DR-EAF systems with conventional
    BF-BOF steelmaking.  Steel production at an EAF plant becomes
    independent of scrap, a raw material  that is volatile in pricing,
    quality and supply (21).
8.  A number of qualities are quoted for the use of sponge iron (versus
    scrap) as a feed material  to electric arc furnaces (71):
    a.  chemical  composition is known exactly
    b.  chemical  composition is uniform
    c.  contains  no undesirable impurities
    d.  permits dilution with low cost (more available) scrap
    e.  easy to transport and handle
    f.  permits automatic continuous charging
    g.  increases  furnace productivity
    h.  less  noise during melt down
                                38

-------
         i.  more predictable  price  structure.

     9.  Air pollution  problems  are  minimized in  a  closed  cycle gas direct
         reduction plant  because all  gases  are  cleaned  and recycled.  Effect-
         ive control of electric arc furnace  steelmaking is possible but care
         must be exercised  to  assure a  nonpolluting operation.

     The major disadvantage normally quoted for the direct reduction pro-
cesses is that they are limited  to small  production units, whereas blast
furnaces are high production units.   This argument  has  been disputed by
Dancy  (72) and others:  "To put  the  size  of this  (Midrex)  direct reduction
plant  into perspective  in relation to modern  blast  furnaces, it will produce
about  8.0 tonnes/day/m3 of  working volume (bustle to stockline) as compared
with only 2.4 tonnes/day/m3 of a modern 5000  m3 blast furnace.  Furthermore,
its production of over  1800 tonnes/day  is only  a  little less than that of
the 'average1 blast furnace in Japan as recently  as  1967 and about the same
as that of the  'average1  blast furnace  in the USA as recently as 1975."

     A second disadvantage  for direct reduction processes  is that they cannot
treat  a wide variety of ore types, as can the blast furnace.  However, direct
reduction processes should  not be thought of  as a means of eliminating con-
ventional iron and steelmaking processes, but a supplement for special
applications.

      Other  Uses of Metallized  Product

     The major use of metallized product  is as  a  feed material to the elect-
ric arc steelmaking furnace, as  noted in  the  previous section.  Other uses
have been proposed and  tested  on, at least, a pilot  plant scale.   These
suggested uses are:

     a.  To increase production  and  decrease  coke requirements in a blast
         furnace.  Miller (4)  notes  that  the  literature (73) shows that
         blast furnace  productivity  is  increased and coke rate decreased by
         5-6 percent for  each  10 percent  increase in furnace burden metal-
         lization.

     b.  To serve as a  coolant source in  the  BOF process for temperature
         control.

     c  To replace the use of scrap in EAF steelmaking processes, in cast
         iron foundry electric furnaces (73)  and  cupolas (74,75), and in the
         BOF process.

     Miller (4)  proposes  that  every  integrated  steelworks should have a
direct reduction capability.  Its product could be  used in one of the ways
listed above.   Its utility  would be  to  supply prereduced iron wherever it
was required.   Its source of feedstock  would  be steelplant fines, dusts from
pollution control devices,  and supplemental ore.  Its reductant would be
Ske oven gas, blast furnace gas, gasified  coal gas  or, if a rotary kiln
process, coke, coal, or coke breeze.


                                      39

-------
                                 REFERENCES
 1.  Kail a, U. and R. Steffen, "Direct Reduction:   Progress and Plans", Iron
     and Steel Int.. V. 50, No. 5, pp 307-315, October 1977.

 2.  Greenwalt, R. and J. Stephensen, "The Role of Agglomeration in Direct
     Reduction Processes", Agglomeration 1977, AIME, V.  2, pp 765-784, 1977.

 3.  Miller, J. R., "The Inevitable Magnitudes of Metallized Iron Ore", Iron
     and Steel Engr.. V. 49, No.  12, December 1972.

 4.  Miller, J. R., "Use of Direct-Reduced Iron Ore and  Balanced Integrated
     Iron and Steel Operations",  Ironmaking and Steel making, V. 4, No. 5,
     pp 257-264, May 1977.

 5.  Davis, W. L., Jr., "Hicap Direct Reduction Process",  Iron and Steel
     Engr., V. 55, No. 3, pp 42-50, March 1978.

 6.  Personal Communications with W. Griffins, Heel a Mining Company,
     September 1978.

 7.  Miller, J. R., "The Direct Reduction of Iron  Ore",  Scientific Am..
     V. 235, No. 1, pp 68-138, July 1976.

 8.  Miller, J. R., "Global Status of Direct Reduction", Iron and Steel Engr..
     V. 54, No. 9, pp 45-50, September 1977.

 9.  Staff, "Sidbec-Dosco Scales  Up and Captures Top-Tonnage Title in Direct
     Reduction", 33 Metal  Producing, pp 40-43, December  1978.

10.  Thorton, D. S., "Requirements of Scrap for Steelmaking and the Increased
     Use of Scrap Processing", Iron and Steel  Int., V.  50, pp 221-225,
     October 1977.

11.  Bertram, J. M., "What-How-Who-Where - Direct  Reduction", Iron and Steel
     Engr., V. 49, No. 7,  pp 31-40, July 1972.

12.  Midrex Brochure, 1979, Midrex Corporation, Charlotte, NC.

13.  Wiliars, H.  M. and R.  C.  Madden, "The Development of  a New Steelmaking
     Process - Utilising Highly Metallised Sponge  Iron", Iron and Steel Int.,
     V. 48,  No.  4, pp 313-321, August 1975.
                                    40

-------
14.  Schroer, C  A  ,  "Operating  a  Midrex  Direct  Reduction Plant - Current
     State of the Art",  Iron  and Steel  Enq..  V.  53,  No. 8, pp 21-25, August
     I -/ / 0 •

15.  Pietsch, W., "The Midrex Cold Briquetting System:  An Economic Answer
     to Direct Reduction Iron Fines Recovery", Iron  and Steel Int , V  51
     No. 2, pp 119-124,  April  1978.           ~	   '

16.  Jinks, W. 0.,  "Oregon  Steel Mills  -  An  Innovator in the Steel Industry"
     Iron and Steel  Enq., V.  54, No.  7, pp 29-33, July 1978.

17.  Coyne, T. J.,  Jr.,  R.  L.  Hunter, and D.  J.  Werner, "An Update of Sidbec
     II1 , SEAISI Symposium. Manila, Philippines, October 1978.

18.  Caradine, T. R.  and H. J. Klingelhofer,  "Status of Midrex Direct
     Reduction Plants in Latin America",  Latin American Iron and Steel
     Congress on Direct  Reduction, Macuto, Venezuela, July 1977.

19.  Coyne, T. J.,  R. L. Hunter, and D. J. Werner, "An Update of Sidbec  II:
     The First Midrex Series  600 Module", SEAISI Symposium. Manila,
     Philippines, October 9-13,  1978.

20.  Goodman, R. J.,  "Direct  Reduction  Processing in Canada - Its Status  and
     Future", UN Seminar, Bucharest,  Romania, May 24-28, 1976.

21.  Dayton, S., "Direct Reduction",  EMJ. pp  80-84, January 1979.

22.  Labee, C. J.,  "HYLSA's Puebla 2  -  A  Fantastic Success", Iron and Steel
     Eng.. V. 54, No. 10, pp  26-29, October 1977.

23.  Quintero, R. G., "Direct Reduction - The HyL Boom", Iron and Steel  Int..
     V.  47, pp 437-440,  December 1975.

24.  Rodriguez,  F.  A.,  "Concepts Relevant to  Steelmaking with HyL Metallized
     Pellets", Iron and  Steel  Eng.. V.  54, No. 1, pp 57-60, January 1977.

25.  Kono, T., "Consumption and  Conservation  of  Energy In The Japanese Steel
     Industry",  Iron  and Steel Int.,  V. 50, pp 87-92, April 1977.

26.  Staff, "First  Industrial  Application of  Thyssen Purofer Direct Reduction
     Process", Iron and  Steel  Eng., V.  54, No. 8, pp 87-88, August 1977.

27.  Labee, C. J.,  "Armco's Direct Reduction  Facility Operating at Full
     Production", Iron  and  Steel Eng.,  V. 51, No. 11, pp 73-75, November 1974.

28.  McGannon, H. E., Ed.,  "The  Making, Shaping, and Treating of Steel",
     Ninth Edition, United  States  Steel.  1971.

29   Oehlberg, R. J., "FIOR Process for Direct Reduction of Iron Ore", Iron
     and Steel Eng..  V.  51, No.  4, pp 58-60,  April 1974.
                                     41

-------
30.  Murdock, C. H., R. Littlewood, "The SL/RN Direct Reduction Plant at
     Griffith Mine", Second Latin American Seminar on Direct Reduction",
     Porto Alegre, Brazil, May 4-9, 1975.

31.  Bold, D. A. and N. T. Evans, "Direct Reduction Down Under:  The New
     Zealand Story", Iron and Steel Int.. V. 50, No. 3, pp 145-156, July 1977.

32.  Cassidy, P. W. and J. M. MacKay, "Development of the Hockin Process and
     Its Application to the Direct Reduction of Iron Ore", Inst. Min. and
     Met., Eleventh Commonwealth Mining and Metallurgical Congress, Paper 62,
     May 1978.

33.  Dailey, W. H., "Integrated Steel Plants of the Future", Iron and Steel
     Eng:. V. 49, No. 4, pp 87-94, April 1972.

34.  Staff, Direktreduktion von Eisenerz, Eine Bibliographisch Studie im
     Auftrag der Kommission der Europaischen Gemeinschaften, Verlag
     Stahleisen, M.B.H., Dusseldorf, Germany, 1976.

35.  Rierson, D. W. and A. A. Albert, Jr., "Development of the Accar Process
     at Allis-Chalmers, Company Report, All is Chalmers, Milwaukee, Wisconsin.

36.  Ferrari, R. and F. Colautti, "The Kinglor Metor Process - Direct
     Reduction  Using a Solid Reducing Agent", Iron and Steel Eng., V. 52,
     No. 5, pp  57-60, May 1975.

37.  Meyer, G.  and U. Bongers, "Reduction by Solid Fuels with the Krupp
     Sponge Iron Process", 2nd Latin American Seminar on Direct Reduction,
     Porto Alegre, Brazil, May 4-9, 1975.

38.  Staff, "Krupp Waste Recovery Process", Metal  Bulletin Monthly, p 31,
     February 1977.

39.  Ferrari, R. and F. Colantii, "The Kinglor Metor Process - Direct
     Reduction  Using a Solid Reducing Agent", Iron and Steel Eng., V. 51,
     No. 5, pp  57-60, May 1975.

40.  Field, L.  I., "The Impact of Energy Conservation", Iron and Steelmaker.
     V. 6, pp 8-16, January 1979.

41.  Office Technology Assessment, O.T.A. Seminar on New Techniques in
     Steelmaking, Washington, D. C., May 2-3, 1979.

42.  Hayes, E.  T., "Energy Resources Available to the United States, 1985 to
     2000", Science, V. 203, January 1979.

43.  Reuter, G., "Rotary Kiln Direct Reduction Plants Built by Lurgi", O.T.A.
     Seminar on New Techniques in Steelmaking, Paper No. 3, Washington D.C.,
     May 2, 1979.
                                     42

-------
44.  Lepinski, J. A.,  "Raw Material  and Product Flexibility  of the ACCAR
     System for Direct Reduction of  Iron Ores", O.T.A.  Seminar on New
     Techniques in  Steelmaking.  Paper No.  1,  Washington D.C., May 2, 1979.

45.  Gold, B., "Tracing Gaps  Between Expectations  and Results of Technologi-
     cal Innovations:   The Case  of Iron and Steel",  J.  Ind.  Econ., V  25
     No. 1, pp 1-28, September 1976.                	

46.  Hirshhorn, J., et al., "Impact  of Technology  on the  International
     Competitiveness of the United States  Steel  Industry", O.T.A., Washington,
     D.C., September 1979.

47.  Johnston, H.,  "Oil Gasification is Teamed  with  Iron-Ore Reduction," Chem.
     Engr., pp 124-125, March 26, 1979.                                	

48.  Ayer, F. A., Symposium Proceedings:  Environmental Aspects of Fuel
     Conversion Technology, II.  EPA-600/2-76-149,  NTIS  No. PB 257-182,
     USEPA, RTP,  NC, June  1976.

49.  Midwest  Research  Institute, Emissions from Iron Ore Mining. Beneficia-
     tion. and Pelletizing. U.S.E.P.A., Industrial Environmental Research
     Laboratory,  RTP,  NC,  Contract 68-02-2113,  Draft Report, March 1979.

50.  Crittenden,  B. D. and R. Long,  "The Mechanism of Formation of Polynuclear
     Aromatic Compounds in Combustion Systems",  Carcinogenesis, V.  1, Edited
     by  R. I. Freudenthal  and C. W.  Jones, Raven Press, New York, N.Y.,
     pp  209-223,  1976.

51.  Herlan,  A.,  "On the Formation of Polycyclic Aromatics:  Investigation of
     Fuel Oil and Emissions by High  Resolution  Mass  Spec.", Combustion  and
     Flame. V. 31,  No. 3,  pp 297-308, 1978.

52.  Blumer,  M.,  "Polycyclic Aromatic Compounds  in Nature", Sci. Am.. V.  234,
     pp  34-45, 1976.

53.  Bridbord, K.,  J.  Finklea, J. Wagoner, J. Moran, and P. Caplan, "Human
     Exposure to  Polynuclear Aromatic Hydrocarbons", Carcinogenesis. V.  1,
     pp  319-324.

54.  PEDCO Environmental,  "Assessment of Arsenic and Other Trace Contaminant
     Emissions from Emerging Energy  Technologies", Draft Report, U.S.E.P.A.,
     Strategies and Air Standards Division, Research Triangle Park, NC,
     Contract No. 68-02-2515.

55   Committee on Medical  and Biologic Effects  of  Environmental Pollutants,
     Vapor-Phase  Organic Pollutants, Volatile Hydrocarbons and Oxidation
     Products, National Academy of Sciences,  Washington, D.C., 1976.

56.  Salvesen, K. G.,  K. J. Wolfe, E. Chu, and  M.  A  Herther, Emission
     Characterization  of Stationary  NOX Sources:  Volume  I.	Results, EPA-
     600/7-78-120a, NTIS No.  PB284-b20, USEPA,  RTP,  NC, June 19/8.
                                     43

-------
57 •   Ibid., Volume 2. EPA-600/7-78-120b, NTIS No. PB285-429, June 1978.

58.   PEDCo Environmental, Assessment of Arsenic and Other Trace Contaminant
     Emissions from Emerging Energy Technologies, EPA Contract 68-02-2515,
     Task Assignment No. 19, Draft Report, September 1978.

59.   Thompson, W. E. and J. W. Harrison, Survey of Projects Concerning
     Conventional Combustion Environmental Assessments, EPA-600/7-78-139,
     NTIS No. PB285-188, USEPA, RTP, NC, July 1978.

60.   Maloney, K. L., 6. L. Moilanen and P. L, Langsjoen, Low-Sulfur Western
     Coal Use in Existing Small and Intermediate Size BoiTers. EPA-600/7-78-
     153a, NTIS No. PB287-937, USEPA, RTP, NC, July 1978.

61.   Colley, J. D., C. A. Muela, M. L. Owen, N.  P.  Meserole, J. B.  Riggs, and
     J. C. Terry, Assessment of Technology for Control  of Toxic Effluents
     From the Electric Utility Industry, EPA-600/7-78-090, NTIS No.  PB283-716,
     USEPA, RTP, NC, June 1978.

62.   Suess, M. J., "Th6 Environmental Load and Cycle of Polycyclic  Aromatic
     Hydrocarbons", The Science of the Total Environment.  V. 6, pp  239-250,
     1976.

63.   Cleland, J. 6. and G. L. Kingsbury, Multimedia  Environmental Goals for
     Environmental Assessment, Volume 1, EPA-600/7-77-136a, NTIS No.  PB276-
     919, USEPA, RTP, NC, November 1977.

64.   Ibid.. Volume II, MEG Charts. EPA-600/7-77-136b, NTIS No.  PB276-920.
     USEPA, RTP, NC, November 1977.

65.   Hites, R. A. and R. E. LaFlamme, "Sedimentary  Polycyclic Aromatic Hydro-
     carbons:  The Historical Record", Science,  V.  198, pp 829-831,  November
     1977.

66.   Blumer, M., W. Blumer and T.  Reich, "Polycyclic Aromatic Hydrocarbons
     in Soils of a Mountain Valley:  Correlation with Highway Traffic and
     Cancer Incidence", Env. Sci.  and Tech., V.  11,  No. 12, pp 1082-1084,
     November 1977.

67.   Hites, R. A. and J. B. Howard, Combustion Research on Characterization
     of Particulate Organic Matter from Flames,  EPA-600/7-78-167, NTIS No.
     PB291-314, USEPA, RTP, NC, August 1978.

68.   Dunlap, R. W. and B. J. Goldsmith, "NSPS:  Critique of Proposed Rule-
     making", Env. Sci. and Tech., V. 13, No. 2, pp  172-178, February 1979.

69.   Davis, J. A., Screening Study on Cupolas and Electric Furnaces  in Gray
     Iron Foundries, EPA Contract  68-01-0611, Task  No.  8,  Draft Report,
     August 1975.
                                    44

-------
70.  Secretariat,  "An Appraisal  of Some  of the  Direct Reduction Processes
     for the Production  of Sponge  Iron", Third  Interregional Symposium on
     the Iron and  Steel  Industry,  Brasilia, Brazil, October 14-21, 1973.

71.  Brown, J. W., "Direct Reduction - What Does  It Mean to the Steelmaker",
     Iron and Steel  Eng.,  V.  54, No. 6,  pp 37-46,  June 1976.

72.  Dancy, T.,  "The Evolution of Iron Making", Met. Trans., AIME, June 1977.

73.  Maschlanka, W., G.  Post and E. Eisner, "Utilization of Direct Reduced
     Iron  In  Different  Iron and Steel  Production  Processes", Int. Iron and
     Steel  Congress, Chicago, IL,  April  1978.

74.  Geek,  H.  G. and W.  Maschlanka, "Use of Midrex Sponge  Iron in the Induc-
     tion  and Cupolas Furnaces of Foundries",  Sem. on the  Utilization of
     Pre-Reduced Materials in Iron and Steel Making, Bucharest, Romania,
     May 24-28, 1976.

 75.   Pietsch, W. G. and R. P. Kreimendahl, "Use of Direct  Reduced Iron  in
      Ironmaking", 2nd ILAFA Direct Reduction Conference, Macuto, Venezuela,
      July 1977.

 76.   Editor, "First Industrial Application of Thyssen Purofer  Direct Reduction
      Process",  Iron and Steel Eng., V. 54, No.  8, pp 87-88, August 1977.
                                       45

-------
                                TECHNICAL REPORT DATA
                          (Please read Instructions on the reverse before completing)
 1. REPORT NO.
  EPA-600/2-80-036
2.
                           3. RECIPIENT'S ACCESSION NO.
 4. TITLE AND SUBTITLE
 Direct Reduction: A Review of Commercial Processes
                           5. REPORT DATE
                            January 1980
                                                       6. PERFORMING ORGANIZATION CODE
 7. AUTHOR(S)

 Larry G.  Twidwell
                                                       8. PERFORMING ORGANIZATION REPORT NO
 9. PERFORMING ORGANIZATION NAME AND ADDRESS
                                                       10. PROGRAM ELEMENT NO.
 See Block 12.
                                                       1AB604
                           11. CONTRACT/GRANT NO.
                                                       NA
 12. SPONSORING AGENCY NAME AND ADDRESS
 EPA, Office of Research and Development
 Industrial Environmental Research Laboratory
 Research Triangle Park, NC  27711
                           13. TYPE OF REPORT AND PEfllOD COVERED
                           Tnhouse Final; 1-7/79	
                           14. SPONSORING AGENCY CODE
                             EPA/600/13
 15. SUPPLEMENTARY NOTES Author Twidwell is no longer with the EPA.  For report details,
 contact Korman Plaks, Mail Drop 62, 919/541-2733.
  . ABSTRACT
               repOr^ gjves results of a review of direct reduction commercial pro-
 cesses for ironmaking. The potential for environmental degradation appears to be
 minimal.  A detailed environmental assessment does not appear to be warranted.  It is
 recommended that samples of scrubber water and sludge material be collected  and
 characterized from several gas reductant reactor systems. It is  also recommended
 that rotary-kiln solid-reductant reactors be sampled to ensure that the final gas
 effluent does not contain harmful concentrations of organic species, sulfur oxides,
 and trace  element contaminants; and that such systems be observed to ascertain that
 fugitive emissions are not released from the kiln at harmful rates.
 7.
                             KEY WORDS AND DOCUMENT ANALYSIS
                 DESCRIPTORS
               b.lDENTIFIERS/OPEN ENDED TERMS
                                                                   c. cos AT I Field/Group
 Pollution
 Iron and Steel Industry
 Reduction
 Scrubbers
 Sludge
 Kilns
               Pollution Control
               Stationary Sources
               Ironmaking
               Direct Reduction Pro-
                cesses
13B
11F
07B,07C
07A,13I

13A
 8. DISTRIBUTION STATEMENT
 Release to Public
                                           19. SECURITY CLASS (This Report)
                                           Unclassified
                                        21. NO. OF PAGES
                                            50
               20. SECURITY CLASS (Thispage)
               Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
             -46-

-------