-------
99.9
-
c:
Q)
()
:u 99
c.
>.
()
c:
Q)
()
;;: 90
-
W
FIGURE VI-4.
VI-6
00
2 3 4
A (Collecting Surface Area), ft2
V (Gas Volume Flow), feet per second
5
6
PRECIPITA TOR EFFICIENCY AS RELA TED TO
COLLECTING-SURFACE AREA, GAS-FLOW
RA TE, AND DRIFT VELOCITY
The effect of variations in gas -volume flow on precipitator efficiency is illustrated
in Figure VI-5. The decrease in efficiency as the rated volume of the precipitators is
exceeded is quite evident. .
+-
c:
Q)
o
....
Q)
a. 90
>-
o
c:
Q)
o
'+-
'+-
W 80
FIGURE VI-5.
100
Designed efficiency
98 percent
Designed eff ic iency
95 percent
o
0.50
1.0
Ra ted Vol ume
1.5
2.0
EFFECT OF GAS VOLUME ON PRECIPITATOR EFFICIENCY
Electrostatic precipitators are used for a variety of processes in the iron and
steel industry, ranging from their use in sinter plants to their application in scarfing
operations.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-7
Sinter -Plant Applications of
Electrostatic Precipitators
The very nature of a sinter plant (with its multitude of transfer points for materials,
and discharge points for receiving, cooling, and screening the sinter) creates a severe
emission problem. Various types of pollution-control equipment are used for a single
plant; or if the operation can be sufficiently enclosed, a central pollution-control instal-
lation may suffice. Electrostatic precipitators are used as secondary air-cleaning units
in sinter -plant operations for the treatment of dust-laden gases coming from sintering-
strand windboxes. The only reported information located on dust loadings for electro-
static precipitators on sintering machines is that for the Inland Steel Company sintering
machine in East Chicago, Indiana. This installation was reported to handle an input dust
loading of 2.5 grains/ scf of gas at 457,000 cfm, and yield an output dl.lst loading of
0.038 grain/ scf gas; (16) an efficiency of 98.5 percent. However, since the pollution-
control system was installed, the materials charged to the sintering machine have
changed from straight ore fines to ore, flue dust, and lime. The characteristics of the
ore used has also changed. These changes in materials have resulted in an increase of
output dust loading to 0.25 grains per cubic foot, and a decrease in collection efficiency
to 90 percent.
Installation of most sintering machines in the United States was done at a time
when the advantages of self-fluxing sinter as blast-furnace burden had not been well
established. However, with the advancement of sinter technology, the use and produc-
tion of self-fluxing sinters became the rule rather than the exception. Lime additions
required in the production of self-fluxing sinters created increased dust problems for
the dust-collecting systems, with the result that additional electrostatic precipitator
capacity was required( 17), or use was made of other types of equipm~nt that were not
as vulnerable to such changes in operating procedures.
Blast-Furnace Applications of
Electrostatic Precipitators
The use of electrostatic precipitators for cleaning blast-furnace gas has come
about because of the requirements for cleaner gas for the hot-blast stoves. Trends to-
ward the use of higher blast temperatures required the use of checker bricks with
smaller holes, which in turn dictated the requirement for cleaner blast-furnace gas to
prevent plugging of the holes. Plugging would present major problems in efficient oper-
ation of the blast stoves. In all recorded installations, electrostatic precipitators have
been added to the existing emission-control systems on blast furnaces.
Operating problems have not been a point of major concern for application of
electrostatic precipitators to blast-furnace emissions. There are two possible reasons
for this trouble -free operation. First, the blast furnace is an almost continuous pro-
ducer of gas, except for the comparatively short intervals when the blowing rate of the
blast furnace is lowered during the time slag is flushed or iron is cast. Second, a
high percentage of the particulate emissions are removed by the wet-scrubbing systems
that had previously been used to clean the gases. In addition to performing a significant
cleaning operation, the wet scrubbers serve to condition the gases prior to their entry
into the electrostatic precipitator.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-8
Open-Hearth Applications of
Electrostatic Precipitator s
Use of oxygen lancing in the refining of steel in open-hearth furnaces made the use
of some type of pollution-control equipment mandatory. Electrostatic precipitators were
used first in the treatment of this emission problem. They were a rather logical choice
because some existing open hearths made use of waste -heat boilers to recover heat and
to generate steam from the exhaust gases, so no additional conditioning equipment was
required.
Electrostatic precipitators can be applied to open hearths in various ways. The
first installations in 1953 were essentially a precipitator for each open hearth. Kaiser
Steel Corporation, Fontana, California, installed one precipitator for each of nine open
hearths (l8), and in the same year the U. S. Steel Corporation (at its Fairless Plant)
installed twin, parallel precipitators for each of nine open hearths(l9). In 1959, U. S.
Steel installed four electrostatic precipitators to treat emissions from 11 open
hearths(20), followed by Bethlehem Steel's (Sparrows Point) installation of six pre-
cipitators for seven furnaces(21), and Weirton's 1965 installation of one precipitator for
two open hearths (22). This sampling of installations appears to suggest indecision on
the part of the various companies as to what type of precipitator installation best suited
their particular open-hearth shops. This however, was not necessarily the case.
Availability of space was a major factor in many cases in determining whether to use
one precipitator per furnace or not. Most of the earlier installations did have sufficient
space available, and there were technological uncertainties about the suitability of the
precipitators. These factors encouraged the use of one precipitator per furnace. U. S.
Steel's installation of four precipitators to serve 11 furnaces was the first attempt at
manifolding the gas off takes from the furnaces into a common collecting main that in
turn channeled the gases into four precipitators.
Manifolding of the exhaust gases from 11 open hearths served two purposes:
(l) it provided a mixing of the waste gases so that the temperature would not exceed
600 F, even though two waste-heat boilers could be by-passed with the discharge of
waste gases at 1200 F into the collecting main, and (2) a diluting of the fume and dust
took place with the result that the gas entering the precipitator had a more uniform dust
loading. (20) Diluting and mixing of the dust was a particular advantage because the
open-hearth furnaces in the shop were at different stages of processing the heats at any
given time, and at any instant the generation of emissions was different from each fur-
nace. One of the most-recent precipitator installations (1968) for open hearths uses
manifolded gas collection. Inland Steel Company's seven-furnace open hearth shop at
Indiana Harbor, East Chicago, Indiana, has one large precipitator. (23) The new instal-
lation at Youngstown Sheet and Tube Company's, Campbell Works will make use of
essentially a one-to-one installation with six precipitators for seven furnaces. (24) The
first unit was placed into operation in October, 1968.
Two major problems that have faced steel companies and equipment manufacturers
in the installation of electrostatic precipitators for open hearths have been (1) the design
of the ducts used to carry the gases from the open hearths to the precipitators, and
(2) the design of the gas -distribution systems at the entrance to the precipitators. Even
though a great deal of theoretical knowledge is available on the design of ducts, the use
of transparent models is considered to be almost a necessity in the practical design of
ducting. This approach was used in the design of the precipitator system at Kaiser in
1953(18), at Bethlehem - Sparrows Point in 1961(21), and at Weirton in 1965(22).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-9
The major problem with re spect to actual efficiency of electrostatic precipitators
on open hearths is the open-hearth process itself. The problem sterns from the varia-
tions in fuel used during the open-hearth refining of steel, which in turn affects the
moisture content of the gases. (25,26) A dry-gas condition occurs shortly after the hot-
metal addition, and lasts for about 15 to 20 minutes. The low moisture content is caused
by a low fuel-firing rate, low use of atomizing stearn, and a low initial oxygen-lancing
rate. (26) The moisture content drops as low as 2 percent, which is a basic cause of
poor efficiency, because low moisture levels result in higher resistivities with accom-
panying higher power requirements to achieve collection of the fume. In some case s,
the situation may be corrected in two ways: (I) power input to the precipitator can be
increased, or (2) a stearn-injection system can be installed to supply the desired mois-
~u.re. Increased power may be ineffective, as in the case of lime which causes back
ionization.
Basic -Oxygen-Furnace Applications of
Electrostatic Precipitators
Electrostatic precipitators were not the first type of pollution-control equipment
installed on basic oxygen furnaces. The first system was a wet system that was com-
bined with a distintegrator at McLouth Steel Corporation in 1955. (25) The first electro-
static precipitators were placed into service in 1957 by Jones and Laughlin at their
Aliquippa plant, followed in 1958 by the Kaiser Steel Corporation at Fontana,
California. (26) A list of the basic oxygen steelmaking shops in the United States, with
their respective pollution-control equipment is given in Table VI-2. The compilation is
based on the latest available information pertaining to current installations, plus re-
ported plans for new installations.
Problems associated with applications of electrostatic precipitators to basic oxy-
gen furnaces are basically those of variability in gas flow and the moisture content and
temperature of the entering gases, (which are functions of the process) and of mainten-
ance. Gas-flow rates for the process must be determined on the basis of theoretical
calculations(27), or on data obtained from similar operations. Calculation of theoretical
gas volumes is quite straight-forward and can even be developed as a nomograph, as
shown in Figure VI-6, that was developed as part of this study. As illustrated by the
dashed line in Figure VI-6, the off-gas volume for a 220-ton BOF heat using 70 percent
hot metal at 4. 0 percent carbon, an exce s s air factor of 100 pe rcent, and a blowing time
of 20 minutes would be 53,000 dm. However, elimination of carbon from the hot metal
is not the only source of carbon monoxide, and it appears that reactions in the hot metal
and slag contribute additional gases. This is illustrated in Figure VI-7, which shows
gas evolution from three different plants as they compare to the theoretical maximum
values. (28)
A significant design problem that is encountered in the design of electrostatic pre-
cipitators for new basic oxygen furnace installations is the potential production rate of
the BOF. The existing state of technology may predict a certain rate of production, and
the electrostatic pollution-control system may be designed for a nominal increase in
capacity; but should the BOF technology develop (as is quite likely) so as to result in a
larger increase in production, the electrostatic pollution-control equipment may soon be
inadequate. Production increases as high as 20 percent can be realized, as shown by
a 150 -ton BOF plant in the Chicago area that is now producing 205 net tons per heat. (29)
Some of the increased productivity is undoubtedly due to increased oxygen-blowing rates.
What effect such increases may have on the amounts of particulates is not known, but it
can be assumed that there will be at least a proportionate increase in iron-oxide fume.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
TABLE VI-2. BASIC OXYGEN FURNACE INSTALLATIONS AND ASSOCIATED AIR-POLLUTION CONTROL EQillPMENT
Annual Capacity,
Net Tons net tons Electrostatic
Number Per Heat March 1969 Future Startup Date Precipitator High-Energy Wet
Alan Wood Steel Co. Conshohocken, Pa. 2 140 1, 250, 000 1968 X
Allegheny-Ludlum Steel Corp. Natrona, Pa. 2 80 500,000 1966 X
m Armco Steel Corp. Ashland, Ky. 2 160 1,400,000 1963 X
»
-i Middletown, Ohio 2 200 2,000,000 1969 X
-i Bethlehem Steel Corp. Bethlehem, Pa. 2 250 2,500,000 1968 X
l'll Burns Harbor, Ind. 2 250 1,800,000 1970 X
r
r Lackawanna, N. Y. 3 290 4,700,000 1964 -66 X
l'll Sparrows Point, Md. 2 200 2,500,000 1966 X
3: CF & I Steel Corp. Pueblo, Colo. 2 120 1,100,000 1961 X
l'll Crucible Steel Corp. Midland, Pa. 2 90 1,250,000 1968 X
3:
0 Ford MotOr Co. Dearborn, Mich. 2 250 2,500,000 1964 X
a! Granite City Steel Co. Granite City, Ill. 2 225 2,200,000 1967 X
» Inland Steel Co. East Chicago, Ind. 2 255 3,000,000 1966 X
r 2 210 2.000,000 1973 X
Z Interlake Steel Corp. Chicago, Ill. 2 75 730,000 1959 X
(I) Jones & Laughlin Steel Corp. Aliquippa, Pa. 2 80 1,000,000 1957 X
::! 3 200 3,000,000 1968 X
-i Cleveland, Ohio 2 225 2,250,000 1961 X <:
C H
-i Kaiser Steel Corp. Fontana, Calif. 3 110 1,440,000 1958 X I
I'll McLouth Steel Corp. Trenton, Mich. 2 110 1958 X ......
o
I 1 110 2,800,000 1960 X
n 2 110 1969 X
0 National Steel Corp.
r
C Great Lakes Steel Div. Ecourse, Mich. 2 300 3.500,000 1962 X
3: 2 200 2,000,000 1970 X
m
C Weitton Steel Div. Weirton, W. Va. 2 325 3,400,000 1967 X
(I) Republic Steel Corp. Buffalo, N. Y. 2 100 1,000,000 1970 X
r Cleveland, Ohio 2 240 2,400,000 1966 X
» Gadsden, Alabama 2 190 1,500,000 1965 X
m Warren, Ohio 2 180 1,600,000 1965 X
0
:u United States Steel Corp. Braddock, Pa. 2 220 2,250,000 1972 X
» Duquesne, Pa. 2 215 2,400,000 1963 X
-i
0 Gary, lnd iana 3 200 3,700.000 1965 X
a! Lorain, Ohio 2 220 2,250,000 1970 X
l'll South Chicago, Ill. 3 150 3,000,000 1969 X
(I)
Wheeling -Pittsburgh Steel Corp. Monessen, Pa. 2 200 1,500,000 1964 X
Steubenville, Ohio 2 250 2,000,000 1965 X
Wisc.onsin Steel Div.
International Harvester Co. South Chicago, Ill. 2 140 1,200,000 1964 X
youngstOwn Sheet & Tube Co. East Chicago, Ill. 2 265 2,400,000 1969 X
Total 57,320,000 18,700,000 23 15
-------
VI-II
100
20,000 30,000 40,000 50,000 70,000
Theoretical Total Off-Gas to the Atmosphere, scfm
(One percent carbon monoxide)
100,000
10,000
FIGURE VI-b.
THEORETICAL TOTAL OFF-GAS VOLUME FROM BOF FURNACES
AS INFLUENCED BY HEA T SIZE, PERCENT HOT METAL, AND
EXCESS COMBUSTION AIR FOR A 4.0 PERCENT CARBON HOT
METAL AND 20-MINUTE BLOWING TIME
Notes:
(a) For other carbon contents in hot metal, multiply off-
gas volume by the ratio: new carbon content/4. O.
(b) For other blowing times, multiply off-gas volume by
the ratio: 20 minutes /new blowing time.
,
OJ
E
'"
g
Shop B
COz- maximum
thearetical fram
0z blawing rate --...... .
COz- "flow
(as measured)
Shop B
~
LL
U
(/)
Shop A
-<;o:COz - flow(as measured)
('"COz-maximum
- - .~ theoretical from ~
0z blowin grate u
(/)
Blowing Time
Blowing Time
Blowing Time
FIGURE VI-7.
THEORETICAL AND ACTUAL GAS RATES DURING BLOWING
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-12
Maintenance problems with electrostatic precipitators are generally associated
with the precipitator proper which includes rappers, vibrators, and insulators. The
hoods over the BOF are a necessary part of the collection system, and can result in
operating problems. The gap between the BOF and the hood is usually dictated by the
anticipated operating conditions and the anticipated buildup of a skull on the mouth of the
furnace. Excess buildup can restrict the flow of air required for combustion of the car-
bon monoxide, with the re sult that a significant amount of carbon monoxide may reach
the electrostatic precipitator with possible disasterous results. The explosion hazard
with an electrostatic precipitator is a reality, and not just an anticipated possibility, as
attested by an explosion in 1968 at the Monessen, Pa., plant of the Wheeling-Pittsburgh
Steel Corp. (30) All iron and steel production was stopped for 1 week, and only partially
re sumed for the second week while repairs were completed.
Electric-Furnace Applications of
Electrostatic Precipitators
Only one known installation of an electrostatic percipitator with an e1ectric-
furnace plant is in operation. This installation is the electric-furnace shop of the Jones
and Laughlin Steel Corporation, Cleveland, Ohio. The electrostatic precipitators are
considered to be operating satisfactorily. Bethlehem Steel Corporation installed elec-
trostatic precipitators on the electric furnace plant at Los Angeles, California in 1955.
They were replaced by bag houses in 1967.
Wet Scrubbers
Wet scrubbers of various types have been used in the integrated iron and steel
industry for many years. The first wet scrubbers were simple spray towers used to
clean blast-furnace gas. However, as cleaner gas became a requirement for firing
blast-furnace stoves to higher temperatures, other types of wet scrubbing were used.
The introduction of fixed-orifice, then variable-orifice, and finally venturi scrubbers
was a natural for blast-furnace operation. Advances in blast-furnace technology and
improvements in burden materials started to reach a point where improvements in the
flow of reducing gases up through the burden became a necessity. One method of ob-
taining such improvement was the use of blast-furnace top pressures that were above
atmospheric pressure. In essence, the blast furnace system became a pressurized
system. The installation of orifice, variable-orifice, or venturi scrubbers in the gas
system was a practical way of obtaining cleaner gas. The pressure required to achieve
the necessary cleaning action was already in the blast furnace, and little additional
auxiliary equipment was required.
High-energy scrubbers (i. e., those capable of pressure. drops of 30 inches of
water or higher) are used in steel-plant applications. High-energy scrubbers are used
for controlling emissions from sinter plants, open hearths, BOF's, as well as from
blast furnaces. In some operations where particulates in blast-furnace gas must be
lowered to 0.005 grain per cubic foot, electrostatic precipitators have been installed
in series with wet scrubbers to obtain final cleaning.
One of the principal advantages of high-energy wet scrubbers is their ability to
handle variations in gas volumes, while still maintaining the required operating effi-
ciency. This characteristic of venturi scrubbers is illustrated in Figure VI-8(31),
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
FIGURE VI-B.
t
+-
c:
~
Q)
a.
>-
o
c:
.~
o
-
-
w
c:
o
+-
o
Q)
o
U
VI-l3
Water /Gas Ratio,gal./IOOOff-
OPERATING.CONDITIONS FOR A VENTURI SCRUBBER
The effect of water rate at a constant throat velocity on the output dust loading of
a venturi scrubber handling blast-furnace gas is shown in Figure VI-9. (32)
2 4 6 8 10
Water Level, gallons/ 1000
EFFECT OF WATER RATE ON OUTPUT DUST LOADING
FOR A VENTURI SCRUBBER
FIGURE VI-9.
0.12
- 0.10
o
(/)
(/)
c:
'5 0.08
~
0'1
g' 0.06
"C
c
..3 0.04
+-
(/)
::J
o 0.02
o
o
14
Theories that can explain the various mechanisms of collection involved in the per-
formance of wet scrubbers and can serve as a basis for comparison are not completely
developed, as indicated from the following statement from a recently published manual
on the subject. (33)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-14
where it can be seen that for a given throat velocity of the gases, the efficiency is in-
creased by increasing the water/ gas ratio, or simply by pumping water at an increased
rate. However, the maximum and minimum values of water and/or gas rates for high-
energy scrubbers used in the iron and steel industry are unknown.
"The theories of the various mechanisms involved in wet collection have
not been completely developed; some, such as the electrostatic effect and
humidification are more presumed than understood, and the air cleaning
tasks to which wet collectors are applied rarely involve a simple, uniform}
nonreactive particulate dispersed in a simple carrier gas. The wet col-
lectors themselves are typically not single mechanism units but usually
function on the basis of several collection mechanisms. : This makes
clear-cut classification of equipment impossible and imposes difficulties
in selecting a collector for a given task without knowledge of their pre-
vious application. "
An empirical method has been developed to correlate scrubber efficiencies.
method is called "The Contacting-Power Concept", and is defined as follows (34):
This
"In the gas -liquid contacting process, power is dissipated in fluid turbulence
(in gas and liquid phases) and, ultimately, as heat; it is this power, expressed
as power per unit of volumetric gas flow rate, that is the criterion of scrubber
efficiency, and it has been designated 'contacting power'. "
It should be noted that the power referred to excludes power consumed by motors, fric-
tion, and mechanical losses. A conclusion of the initial investigators which led to the
development of the contacting-power concept is as follows(35):
"When compared at the same gas power consumption, all scrubbers give
substantially the same degree of collection of a given dispersed dust,
regardless of the mechanism involved and regardless of whether the
pressure drop is obtained by high gas flow rates or high water flow
rates. The collection efficiency increases as the pressure drop in-
creases, the increase being especially rapid for pressure drops over
lO-in. water. II
Mathematically the contacting-power concept can be expressed as follows:
Nt = exP;
where
Nt = the number of transfer un:its
1
Nt = 2. 3 log 1 -E
E = collection efficiency
PT = total contacting power.
\
While the calculation of Nt is quite simple, similar calculations for PT re9,uire data on
the gas-flow rate, water-flow rate, and water feed pressure. . When the required data
is plotted on log-log coordinates, a straight line correlation is evident as shown in
Figures VI-lO and VI-ll(34). The coefficient ex is the value of the intercept where PT is
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
4
+-
Z 3
VI
:=
c 2
:::>
98
95
90 "E
a>
o
L-
80 ~
70 :>.
o
c
60 .~
o
....
50 W
L-
a>
....
VI
C
o
L-
t- 1
.... 0.9
00.8
a; 0.7
E 0.6
~ 05
c::. .
0.4, 2 3 4 5 6 78910
Contacting Power, hp/(IOOO cuft min}
40
.r IGURE VI-IO. PERFORMANCE OF A
VENTURI SCRUBBER ON A METAL-
T TJRGICAL FUME
VI-IS
7
6
99.9
99.5
99 +-
c
98 ~
L-
a>
a.
95
>.
o
c
90 .~
;0=
....
w
80
70
FIGURE VI-II. PERFORMANCE OF A
VENTURI SCRUBBER ON OPEN -HEAR TH
FUME
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
z+- 5
~ 4
c
:::>
L-
.&
VI
c
o
.=
....
o
3
2
Oxyge n.
in use
L-
a>
.Q
E
:J
Z
No oxygen used
one, and 'Y is the slope of the line. The coefficients are essentially functions of the dust
and/ or fume, and are not influenced by the way the contacting power is applied to the
scrubber. Some industrial data were used in analyzing the contacting power concept;
however, full-scale industrial tests in steel plants were not made.
I .
2 3 4 5 6 7 8 9 10
Contacting Power, hp/(IOOO cu ft min}
A characteristic of wet scrubbers over other types of emission-control equipment
is that gases such as carbon dioxide and sulfur dioxide will dissolve in the water. (36)
This can be particularly advantageous where sulfur dioxide is in concentrations that
would exceed the concentrations permitted by regulations. However, this plus in per-
formance is offset by the severe corrosion problems resulting from the formation of the
respective acids, and by the effects the acidified water has on the operation of the water-
treatment facilities. If the maintenance problems involved are disregarded, the wet
scrubber may be an effective means of reducing the emission of sulfur dioxide to the
atmosphere. It has been stated that there is a significant removal of sulfur dioxide by
wet scrubbers(37), but also that there are no data on the input and output of sulfur dioxide
for wet scrubbers(38).
Application of Wet Scrubbers in
Sinter Plants
Of about 40 sinter plants at various steel plants, only two are known to have wet-
scrubber installations. One uses a venturi scrubber to treat emissions from the wind-
box of the machine(39), and the other uses flooded-disk scrubbers at the discharge end
of the sintering machine(40). The output loading of the flooded-disk scrubbers is report-
edly O. a I grain per cubic foot.
Early application of wet scrubbers to sinter plants resulted in operating problems
which were traced to erosion and imbalance of the fan blades on the exhaust-system
blowers. These are the blowers that provide the draft through the sinter bed required
-------
VI -16
to ignite the fuel. Erosion of the blade s has been a problem even with dry pollution-
control systems. However, the imbalance occurring in the fan blades is aggravated in
sinter plants having wet pollution-control systems because the dust that is carried over
to the fan is moist, and has a greater tendency to accumulate on the blades. The im-
balance problem has been reported to be caused by uneven buildup of the dust and also by
the breaking off of large amounts of built-up dust which places severe unbalanced loads
on the blades. Both situations have caused severe vibrations and sometimes major
breakdowns in the blowers. This situation is minimized by constant preventive main-
tenance to remove the dust buildup.
Application of Wet Scrubbers to
Blast Furnaces
During 1967, about 170 blast furnaces were producing hot metal, and of this num-
ber about 51 were using wet scrubbers as the principal method of cleaning blast-furnace
gas. Of these, about 33 were high-energy scrubbers. In addition, about 35 blast fur-
naces were equipped with high-energy scrubbers as cleaning units preceeding electro-
static precipitators which serve as final cleaning units.
The first high-energy scrubbers were installed in 1955. They were simple, fixed-
orifice plates installed in the gas lines, with water introduced into the main at some dis-
tance upstream from the orifice plate. These scrubbers operated at pressure drops of
30 to 50 inches of water3 with resulting output loadings varying from 0.01 to 0.03 grain
per cubic foot. (41,42,4 ) The orifice scrubber, however, had the major disadvantage
that it could not handle the variations in gas flow, and consequently could not meet the
required emission limits during certain phases of blast-furnace operation when the
velocity of the gases coming from the blast furnace was lowered.
The need for high-energy wet scrubbers that could handle variations in gas flow
from a blast furnace led to the development of variable-orifice scrubbers that attempted
to cope with the variability of gas flow by adjusting the size of the orifice opening. The
performances of a variable-orifice scrubber and a fixed-orifice scrubber are illustrated
in Figure VI-12. (44) The effect of a reduction in wind rate on the two types of equip-
ment is quite evident.
During the same period that the orifice and variable-orifice scrubbers were re-
ceiving attention from blast-furnace technologists, the venturi scrubber was also under
investigation as a possible means of cleaning blast-furnace gases. The first application
of venturi scrubbers to blast-furnace gas was reported in 1955(45), with other installa-
tions reported in 1956(46) and 1960(47). Technical data relating the output dust loading
to the water rate and pressure drop of a blast-furnace venturi scrubber are shown in
Figure VI-13(45). The relationship between clean-gas dust loading and pressure drop
is shown in Figure VI-14(47), that illustrates the low output dust loadings that can be
achieved with this equipment. This high performance, however, can be achieved only
if the blast furnace is operating at a high-enough top pressure to provide the required
pressure drop. This level of high top pressure has not been achieved, as evidenced
by the large number of blast furnaces operating with electrostatic precipitators as the
final gas -cleaning unit. (See Table VI-l that shows 108 blast furnaces operating with
electrostatic precipitators. )
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-I?
0.12
't 0.10
If)
"'-
~
~ 0.08
0\
c
-g 0.06
.3
1;; 0.04
:J
o
0.02
o
o
20 40 60
Reduction in Wind Rate, percent
FIGURE VI-12. EFFECTIVENESS OF GAS CLEANING BY A FIXED-
ORIFICE SCRUBBER AND A VARIABLE-ORIFICE
SCRUBBER WHEN GAS -FLOW RATE IS VARIED
.~ ..
0.12
0\1f)-
CO,+-
15 (.!):J 0.08
ouu
o """'-
...J ~ If) 0 04
1;; o.S .
:JQ)o
OU ~ 0
+-=' ~ 8
80\
1: '0 6
f--
o'+-
-:J 4
OU
.~o
0-
Q: ';;; 2
~ C
Q)o
0= 0
~ g 0
1969
operation
10 20 30 40 . 50
Pressure Drop Across Venturi Section, inches water
FIGURE VI-B. OPERATING CHARACTERISTICS OF A BLAST-
FURNACE VENTURI SCRUBBER
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-I8
100
0(\1
:J:
.E
a.
o
...
CI
ell
...
::J
VI
VI
ell
...
a..
10
...
::J
-
c:
ell
>
1
0.001 0.01 0.1 1.0
Clean Gas Dust Loading, grains per standard cubic foot
FIGURE VI-14.
CALIBRATION CURVE FOR A BLAST-FURNACE
VENTURI SCRUBBER
Application of Wet Scrubbers to
Open-HeariliFurnaces
Wet washing of open-hearili gases was first considered to be economically ex-
pedient for shops that were to be operated only during high peak demands for steel,
under which conditions the low capital cost for ilie wet system was considered an ad-
vantage. (48) However, some open-hearili shops iliat were considered to be fairly new
found that wet scrubbers were economically attractive when ilie shop either had no waste
heat boilers or ilie existing boilers could not lower the gas temperatures enough to war-
rant the installation of electrostatic precipitators or bag houses. (49, 50) The first open
hearili installation was made in 1959(48) at U. S. Steel's Edgar Thomson Works, and
others subsequently followed. Output gas loadings of 0.01 to 0.05 grain per cubic foot
have been reported for ilie installations, again wiili the cleaning efficiency relating di-
rectly to the pressure drop of the scrubber. The relationship between clean-gas dust
loadings and pre ssure drop for an operating open hearth installation is illustrated in
Figure VI-15(48J. Oxygen lancing was used during the refining period. The figure is
representative of open-hearth practice with low oxygen-blowing rates and is not neces-
sarily representative of present-day practice using higher oxygen-blowing rates.
Application of Wet Scrubbers to
Basic Oxygen Furnaces
Wet scrubbers were first installed in 1954 on a basic oxygen furnace at the
Hamilton Plant, Dominion Foundries and Steel Ltd., Ontario, Canada. The first in-
stallations in the United States was made at the Duquesne Plant of the U. S. Steel Cor-
poration in 1963. The number of high-energy wet scrubbers installed over the years as
compared to the number of electrostatic precipitators is shown in Figure VI-16. One of
the principal reasons for selecting electrostatic precipitators over high-energy wet
scrubbers is the existence of a water-treatment problem in a plant. Problems can in-
clude inadequate water-treatment facilities, or the lack of sufficient water. Even though
there are 15 BOF plants in the United States with high-energy scrubbers as the primary
pollution-control equipment, there is a notable lack of published information concerning
their operation. It appears that problems associated with wet scrubbers in other appli-
cations apply al so to BOF installations.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-19
........ 0,10
en
c 0,08
C+-
~ 0
010 0,06
-
010 Ore and lime boil
c.- 0.04
.- .D and working period
'0 :J
C 0
0
..J'O
~
+- C 0.02
en '0
:J C
0 C Charging, melt down
+-
en and hot meta I
+- ~
:J Q) 0.01
,2-a.
:J 0,008
0
26 28 30 32 34 36 38 40
Pressure Drop, inches of water
FIGURE VI-IS.
RELATIONSHIP BETWEEN CLEAN-GAS DUST LOADING AND
PRESSURE DROP FOR A WET SCRUBBER ON AN OPEN-
HEAR TH FURNACE (OXYGEN LANCING USED DURING THE
REFINING PERIOD)
30
25
5
Electrostatic
prec i pi tators
en
(; 20
+-
o
~
o
U
15
o
1956
,,"
--
/-
/
/
/
,,"
,,"
1""- High energy wet
1 scrubbers
/
"
,,"
,.
-
o
~
Q)
.D
E 10
:J
Z
1958
1960
1962
1964 1966
Year
1968
1970
1972
1974
FIGURE VI-16. INSTALLATION OF ELECTROSTATIC PRECIPITATORS AND
HIGH-ENERGY SCRUBBERS FOR AIR-POLLUTION
CONTROL AT BOF STEELMAKING PLANTS
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-20
Problems associated with improper performance of wet scrubbers can be attrib-
uted to a lack of materials to withstand the abrasive and corrosive nature of the dust-
laden water, or a misapplication of construction materials. Design of wet scrubbers
does not appear to be a contributing factor to deficiencies in scrubber performance.
Application of Wet Scrubbers to
Electric Furnaces
High-energy scrubbers for electric steelmaking furnaces are known to be used in
only two shops; both owned by the Armco Steel Corporation. The first installation was
made at the Butler, Pa., plant on a 70-ton furnace in 19S8(S1), and the second the
Armco plant in Houston, Texas. The Butler plant (currently in the process of expan-
sion) will soon include three ISO-ton furnaces(S2) that probably will be serviced with
high-energy wet scrubbers. When expansion plans are completed in 1969, the Houston
plant will be equipped with high-energy wet scrubbers operating at a pressure drop of
60 inches. (S3) No other electric furnace shops are known to be using high-energy wet
scrubbers.
Fabric Filters
Fabric-filter installations (or baghouses as they are more commonly called) have
their biggest steel-industry application in the control of emissions from electric
furnaces -- a total of 29 installations. There also are three known applications at sinter
plants, and two applications on scarfing machines. Fabric filters are used on BOF's
in Europe(S4), but no installations of this type have been made in the United States.
The performance of fabric filters has been well developed on the basis of the-
0retical principles, and numerous descriptions are available in the published literature.
A reduction of the theoretical concepts to a simplified mathematical form results in the
following equation(SS):
r Lt I v2
H' =
7000
where
H' = filter resistance increase in inches of water
r = specific resistance of the dust (determined experimentally) in inches of
water gage per pound of dust per square foot of filter cloth area per foot
per minute of filtering velocity.
L = input dust loading to the filter in grains per cubic foot of air
t' = operational time in minutes
v = filtration velocity in feet per minute.
This equation leads to the conclusion that resistance of the filter to flow is dire ctly pro-
portional to (1) the square of the superficial face velocity, (2) the weight of the dust
collected on the fabric, and (3) the time of operation since the last cleaning. (S4)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-2l
A more recent method of evaluating fabric-filter performance is based on '~filter-
drag". The development of the method is based on similarity to electrical circuitry and
to heat transfer. The equation for filter-drag evaluation is as follows:
v = ~ ' or in electrical analogy I = ~
where
v = velocity in feet per minute
6P = pressure drop in inches of water
S = filter drag in inches of water per foot per minute.
.Lne factor V is similar to the air-to-cloth ratio of more common usage. Filter drag is
independent of the size of the unit, the filter ratio, the type of dust, the style of fabric,
and all other specifics of application. (56) The method has been used to evaluate the per-
formance of fabric materials prior to full-scale installation in a bag house.
Application of Fabric Filters
to Sinter Plants
Typical of the use of fabric filters in sintering plants is the collection of dust
generated at the discharge and screening locations at the Bethlehem Steel Corporation's
plant at Bethlehem, Pennsylvania(57), and the U. S, Steel Corporation's plant at Gary,
Indiana(l7). Both units are sectionalized, with sufficient capacity to permit shutting
down sections for maintenance without affecting the cleaning efficiency of the units.
Pertinent statistics for the two units are given in Table VI-3.
Application of Fabric Filters to
Open-Hearth Furnaces
An experimental baghouse was installed at the Lackawanna plant of the Bethlehem
Steel Corporation in February, 1960. (58) The development work was done with an
oxygen-lanced open-hearth furnace, and results were satisfactory. A production bag-
house was placed into operation by Bethlehem Steel at the No.2 Open-Hearth Shop at
Sparrows Point in 1963. (59) The only change that had to be made in melting practice
was the elimination of fluorspar as a flux. The baghouse had 10 hoppers of 80 bags
each. The bags were 11.5 inches in diameter and 34 feet long. Design capacity was
145,000 cfm at 500 F. With an input dust loading of 1.4 grains per cubic foot, the out-
put loading was 0.0007 grain per cubic foot, for an efficiency of 99.95 percent. The
open-hearth furnace and baghouse were operational in January, 1966, but the results of
subsequent operations are unknown.
Application of Fabric Filters to
Electric Furnaces
Fabric filters have been successfully applied to the control of emissions from
electric furnaces ranging up to 100 and l50-net-tons capacity, and for multiple-furnace
shops as well as one -furnace shops. (60, 61, 62, 63, 64) However, because electric
BATTELLE MEMORIAL INSTfTUTE - COLUMBUS LABORATORIES
-------
VI-22
TABLE VI-3.
DESIGN AND OPERATING DATA FOR SINTER-PLANT
FABRIC FILTERS ON SINTER STRAND DISCHARGE
Design or Operating Variable
U. S. Steel Corp.,
Gary, Indiana
Bethlehem Steel Corp. ,
"Bethlehem, Pa.
Volume of Air, cubic feet per minute:
172,000 at 255 F
240,000 at 350 F
Suction, inches of water:
12
n. a.
Pressure Drop Across Bags,
inches of water'
Total Bag s, numbe r
4 n. a.
10 16
88 72
880 1152
Hoppers in Unit, number
Bags per Hopper, number
Bag size:
Diamete r, inche s .
Length, feet'
11. 5 12
32.2 28
Fiberglas s Fiberglass
17 36
12 -20 n. a.
2. 17 2.29
Bag Type
Bag Life, months
Bag Permeability, cfm per foot2 of cloth
Air-to-Cloth Ratio (Normal), cfm per
ft2 of cloth
Air-to-Cloth Ratio
(One Compartment Cleaning),
cfm per ft2 of cloth
Theoretical Design Efficiency, percent
2.41 2.44
1 75 to 3 0 0 200 to 500
99+ 99+
Air Temperature, F
n. a. - not available.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS. LABORATORIES
-------
VI-23
furnaces have design and operating characteristics that can vary from furnace to
furnace and shop to shop, the development of pollution-control facilities is not as
straight-forward as for other steelmaking operations. One of the major problems in
the design of electric-furnace pollution-control systems is the design of methods that
will completely capture the fumes. The majority of electric-arc furnaces are top
charged, which means the roof is removed during charging. As a re suIt, the emissions
generated during the charging operation are difficult to capture and contain. Capture of
fumes by hoods and by direct-extraction techniques during melting and refining have not
completely solved the problems of collection and containment, and extraction of fumes
through the plant roof has been developed to control emissions through the entire
plant. (64, 64)
Cyclone Dust Collectors
One other type of dust cleaning equipment that has found extensive use in the con-
trol of emissions from ironmaking and steelmaking operations is the cyclone dust
collector. The cyclone separates particles from the gas by means of a centrifugal
force exerted on the particles in a vortex flow that drives the particles toward the wall
of the body of the collector. The particles at the wall move toward the discharge open-
ing of the cyclone as a result of an axial component of the vortex flow, aided in the
case of cyclones used in the iron and steel industry by the force of gravity. The magni-
tude of the radial forces acting on the particles depends on the nature of the vortex flow
in the different sections of the cyclone. Radial gas velocities tend to act as counter-
acting forces, and tend to offset the separating forces generated in the cyclone.
Cyclones are suitable for collecting medium and coarse dusts, but are not suited
for very fine dusts or metallurgical fumes. Their advantages are that there are no
moving parts, there is a wide choice of construction materials, and maintenance costs
are low. Power costs can be quite high because a high degree of efficiency is required.
Cyclones find their principal application as pre cleaners for other types of
emission-control equipment. Some of the applications apparently are deliberate,
others occurred in a transition from lower to higher colle ction efficiencies.
while
Cyclone s find application in pelletizing plants and in lime stone plants for the col-
lection of the large-size dust generated in certain of the operations. As pre cleaners,
they are used in series with dust catchers, wet scrubbers, and electrostatic pre-
cipitators for cleaning blast-furnace gas; as precleaners for electrostatic precipitators
handling the dust and gas from a sinter -plant wind box; and as part of the series of
equipment used in open-hearth-furnace -emission control. No information has been
located in the literature or during this study concerning the efficiencies of cyclones in
the various iron and steel plant operations, or concerning any particular problems that
have arisen with their use.
Cost of Applied Control Equipment
As part of the companion study for the Division of Economic Effects Research of
NAPCA, estimates were made of the capital costs and operating costs of the principal
types of emission-control equipment used in the integrated iron and steel industry. The
types of equipment considered were electrostatic precipitators, high~energy wet
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-24
scrubbers, and fabric filters. Cost estimates were based on established industrial
estimating techniques, with cost information supplied by certain steel companies and
equipment manufacturers.
For the purposes of the estimates made, a control system was considered to be
made up of all the items of equipment and their auxiliaries which are used solely for
the general abatement of atmospheric pollution in the neighborhood of the steel works.
Typically this will include a collecting hood or gas -collecting pipe at the furnace,
ductwork, spray cooler, dust collector, fan and motor, and stack. Included also will
be structural steel, foundations, control instruments, insulation, piping, water treat-
ment, and electric power supply facilities for the entire gas-cleaning system. Excluded
are those equipment items which, while they may contribute to the functioning of
pollution-control equipment, would be used for process or economic reasons even if
there were no pollution-control requirements. For example, an open hearth furnace is
usually connected to a stack whose primary task is to supply draft for causing gas flow
through the heat regenerative stoves. In this study, the cost of the original stack is at-
tributed to the steelmaking proce ss because the furnace cannot operate without it. How-
ever, any increase in stack height, or other modifications nece ssary when air pollution
equipment is installed, is charged as a cost of pollution equipment.
The cost of land occupied by pollution-control equipment has not been included. It
is recognized that such land has a real value, but a satisfactory method for estimating
it has not been established. Costs associated with preparation of the site, start-up
operations, and working capital also are not included. Certain portions of a control sys-
tem occupy or utilize parts of steel plant buildings and, therefore, might be charged
with a share of general building costs. This item has not been estimated here. In cal-
culating operating costs, no attempt was made to allocate a portion of general overhead
to control systems.
Capital and operating costs in the estimates are based upon collectors whose
efficiency can be relied upon to produce an outlet dust loading of O. 05 grain/scf of gas.
The estimates of capital costs include facilities for loading the collected dust or
sludge into trucks for transportation elsewhere. No by-product values have been
assigned. Central engineering costs, overheads, and fees were based upon a standard
sliding scale generally used by contract engineers. Labor cost was calculated at the
nominal value of $5. OO/man hour, including all welfare and fringe costs.
It is believed that the general precision of the capital cost estimates is such that
most specific plant situations will fall within :1:15 percent of the estimated values. In
more statistical terminology, it might be suggested that the standard deviation is about
:1:10 to 12 percent. It is to be expected that any specific plant location which presents
unusual cost problems associated with layout, structure, power supply, etc. might fall
outside these limits. In such cases, a detailed plant design and estimate must be pre-
pared if accurate capital cost data are required. The accuracy of operating cost
values is influenced by many factors which may vary considerably from one ,company
to another. The selection of control equipment should not be based upon small differ-
ences in estimates of operating cost.
Operating cost estimates include costs for the following items:
(1) Electric Power
(2) Maintenance
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-25
(3) Depreciation
(4) Capital Charges
(5) Operating Labor.
Electric energy is calculated at a standardized rate of l/:. per kilowatt hour.
Maintenance is taken at a nominal cost of 4 percent of the total investment. Deprecia-
tion is calculated on a straight-line method using total investment with an expected life
of ten years. Other studies of depreciation have suggested longer service life times,
but these are considered to be greater than average plant experience will confirm.
Advancing technology and rising standards give importance to the factor of technical
obsole s cence.
Capital charges are taken at 10 percent per annum. It is believed that this will
be reasonable in the light of rising interest rates and local taxes.
Annual calculations are based on 330 operating days per year, 24 hours per day.
This gives a total of 7,920 operating hours per year.
The estimated capital and operating costs are summarized in the form of graphs
in Figures VI-17 through VI-24. For further details on these estimates, reference
should be made to the companion report (A Cost Analyse s of Air Pollution Controls in
the Integrated Iron and Steel Industry); especially to Appendix C of that report.
Cost-Effectiveness of Applied Systems
The three principal types of equipment used to control iron oxide fumes from the
various process segments in the integrated iron and steel industry can meet current
air-quality requirements. This holds if the equipment has been correctly designed and
constructed and is properly maintained. The particular equipment selected must per-
form its designed function at a cost that will not create an excessive increase in the
cost of the finished steel products. Many factors are involved in determining the
actual cost of air-pollution control to any given steel plant. Factors such as the local
level of allowable emissions, power costs, availability of water, ease of dust disposal,
and possible reclamation or marketing of the recovered iron oxide dust enter into the
final determination of costs. Estimated operating costs for air -pollution-control
equipment on open-hearth furnaces, basic oxygen furnaces, and electric furnaces are
illustrated in Figure VI-25. The wide range possible in these costs is quite evident.
As regulations governing allowable emis sions become stricter, the costs for con-
trol increase, but not as a simple direct relationship. This can be illustrated in the
case of electrostatic precipitators as shown in Figure VI-26. (66) As collection ef-
ficiencies increase beyond 95 percent, the cost increases at a rapid rate. The effect of
collection efficiency on the cost of electrostatic precipitators only, as well as installed
cost, is shown in Figure VI-27 for installations in the iron and steel industry. Costs are
1968 costs. The one type of equipment whose costs are not drastically affected by in-
creased requirements for efficiency is the fabric filter or baghouse.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
6.0
5.0
(/)
L-
o
o
'U
-
o 4.0
(/)
c:
o
E
~
+-
(/)
o
U
E 3.0
a.
o
U
'U
Q)
o
+-
(/)
c:
H
'U 2.0
Q)
+-
o
E
+-
(/)
w
1.0
0.0
20
VI-26
Legend
Electrostatic Precipitator
- - High- Energy Wet Scrubber
-- --- Fabric Filter
/
/
/
BOF /
./
I "
,
II I
/ II
/
Open hearth / Open h~arth
/ Electric~ // Open hearth'
/ Furnace#,/ /
// .h II' '"
~ ///"///
~ ,a "';//
----:: ~,...~./ Electric furnace
~ ,..."" '"
....... ,...
- - ---.....
---
.i
30
40 50
100 150 200 300 400 500
Design Capacity, ACFM
1,000
ESTIMATED INSTALLED CAPITAL COSTS (1968) OF AIR-
POLLUTION-CONTROL EQUIPMENT AS RELATED TO
DIFFERENT STEEL-MAKING PROCESSES, ON THE BASIS
OF DESIGNED ACTUAL CUBIC FEET PER MINUTE (AT
TEMPERATURE) OF GAS FLOW RATE
FIGURE VI-I?
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
2.6
2.4
01 2.2
»
'-i :2:
-i
m LL 2.0
u
r
C
en .~
:! c
~ 1.2
-i Q)
C Q.
-i O
m '0 1.0
I ~
C
o 'C
.0
-------
6.0
~
c
0
+-
0
U
Q) 5.0
....
a.
m Q)
» 0
-i "0
-i c
111 o
r I/)
r Q)
111 2'
~ 0 4.0
111 ~~
~ ulJ...
0 OU
;u .4.:: 0 JLU" ON-CON" ~O.. ~QUIPMENT ;'0 t S"!: !:LMAKIl' G :> tOC !:SS!:S
-------
m
»
....
....
III
r
r
III
~
III
~
o
:u
»
r
en
~
o
"0
"0
-
o
en
.~ 1.5
.-
E
+-
en
8
o
+-
'0. 1.0
8.
z
(II
:f
....
c
....
III
I
o
o
r
c
~.
m
c
(II
r
»
m
o
:u
»
....
o
!
III
(II
"0
.!!
o
+-
en
c::
-
"0 0.5
Q)
+-
o.
E
+-
en
W
2.0
/
,
/
/
.. l ~
, l
/ - ,'/
2:/'
, ~.
/h
, 0
//
..~
~
Number of
furnaces
\3
Fabric filter, 275 F
I
I
I
I
I
I
I
I
I 1
1 1
1 II
1 II
I II
1
31 //
1 I'
I //
/ //
// 2 //
/ ///
/ /r/
/ ",/
'" '" ",'" I
",'" , "
" ...."""
-" ....' ....
- ,""" "",,'"
-- -.--::,..-"""
---
<
......
I
N
~
High-enery
wet scrubber, 180 F
Electrostatic
precipitator, 500 F,
100,000
4OOpoo
00 '
4O/XX)
100,000
200/XX) ?l:)Qpoo
- 40,000
4OpOO
100,000 200.000 '&YJ,OOO .
Designed Capacity. ACFM
2OOpoo
FIGURE VI-20. ESTIMATED INSTALLED CAPITAL COSTS OF AIR-POLLUTION-CONTROL EQUIPMENT
INSTALLED ON ELECTRIC-ARC STEELMAKING FURNACES. CONTROL EQUIPMENT
DESIGNED TO HANDLE EMISSIONS FROM ANY ONE FURNACE ATONE TIME
-------
5.0
m
>
-i
-i
III
r
r
III
3:
III
3:
o
!
>
r
If)
~
.Q
(5
"0
b 4.0
If)
c
.Q
E
00-
:3 3.0
u
C
00-
"5-
c
u
"0
Q) 2.0
c
00-
If)
C
......
Z
III
j
-i
C
-i
III
I
o
o
r
c
3:
m
c
III
r
>
m
o
;n
>
-i
o
;n
III
III
"0
~
c
E
+: 1.0
If)
w
0.0
3OpOO
I
,
/3
.
/
2#
D'
,~
If
yJ'
/'
'Ii'
<:
......
I
VJ
o
High - energy
wet scrubber, 180 F
Electrostatic
precipitator, 500 F
Fabric filter, 275 F
~umber of
furnaces ~
3
./
I
31
I
I
I'
I.~
2/.7
~'/
/~'/
/J':"/
~~~
,,,7
I.".~"
."~.,,
~."
,'"
--
100,000
500,000
500pOO
20,000
Designed Capacity. ACFM
FIGURE VI-21. ESTIMATED INSTALLED CAPITAL COSTS OF AIR-POLLUTiON-CONTROL EQUIPMENT
INSTALLED ON OPEN-HEARTH FURNACES. CONTROL EQUIPMENT DESIGNED TO
HANDLE EMISSIONS FROM FURNACES OPERATING AT THE SAME TIME
-------
VI
L-
.2
o
"0
'0 1.6
VI
c
o
- 1.4
E
~ 1.2
u
o
~ 1.0
o
U
"0
Q) 0.8
o
.-
VI
c
H 0.6
"0
Q)
.-
o
E 0.4
.-
VI
W
0.0
. 20,000
2.0
1.8
0.2
VI-31
Legend
Electrostatic Precipitator
--- Low- Energy Wet Scrubber
----.;.;;. Fabric Filter .
--- - Cyc lone
~
Sinter-plant
wind box
at 325 F
Sinter-plant
material handling
at 135 F ~
~...
--...--... ---. ,/
.--- - -- --... "
-------- _Pellet ~
-- plant ... ~ ...'" Pellet plant
----...
.
50,000
100,000 . 200,000
Designed Capacity, ACFM
400,000
800,000
FIGURE VI-22. ESTIMATED INSTALLED CAPITAL CQSTS OF AIR-POLLUTION-
CONTROL EQUIPMENT USED IN SINTER AND PELLET PLANTS
BATTELLE MEMORIAL .INSTITUTE - COLUMBUS LABORATORIES'
-------
1_.. -.---------
2.0
~
lL..
U
cd:
'-
Q)
a.
~ 1.5
o
<5
'U
~
-
If)
o
U
01
C
VI-32
\
Sinter plant
material handling
at 135 F
Legend
Electrostatic Precipitater
-- - Low- Energy Wet Scrubber
------ Fabric Filter
--- - Cyclone
..
\
,
\
\
"
''"'
,
"-
',,-
~
o
:J
C
c
-------
VI-33
450
-
I/)
o
U
o
-
'0.
o
u
'0
~
Capital cost
350
Electrostatic
precipitator,
100F
400
o
-
~ I/) 300
......""
o
'0-
Q)-
_0
0'0 250
.~o
-0
1/)0
w-
High - energy
wet scrubber,
100 F
~
-
I/)
o
U
01
.S
-
o
....
~ 0.80
0:E
°lL
::Ju
~
-------
VI-34
1.80
c
S2 - 1.60
+- c
Q) .2
c+-
L- .2
~ ~ 1.40
L-
VI Q.
L- Q)
.S!'O
o '0 1.20
"t:! c
~ 0
+- VI
VI Q)
8 ~ 1.00
010
c.c.
:;:: 0
0-
~.E 0.80
g'~
"t:! 0
~ ~ 0.60
0'0
E :J
:;::0
VI c
W ~ 0.40
0.20
o
1.0 1.5
Annual Production of Raw Steel.
2.0
2.5
3.0
millions of net tons
FIGURE VI-25. RANGE OF ESTIMATED OPERATING COSTS FOR AIR-POLLUTION-CONTROL EQUIPMENT PER NET TON OF RAW
STEEL - OPEN-HEARTH FURNACES, BOFS, AND ELECTRIC FURNACES (TWO-FURNACE OPERATIONS)
3
IJ
V
J
V
~ /'
0..,
V ~
~
- -
-
FIGURE VI-26. ELECTROSTATIC-PRECIPITATOR COSTS
AS AFFECTED BY COLLECTION EFFICIENCY
E
.....
o
o
o
o
o
o
LO
01
E .EO: 2
..... '0
o Q)
Q)
~ ~
Q.Q)
~ ~
o E
0.2
'0 ~
":01
~~
U"t:!
C
o
.c.
c
o
"t:!
Q)
VI
o
m
80 82 84 86 88 90 92 94 96 98 100
Efficiency, percent
BATTEL.L.E MEMORIAL. INSTITUTE - COL.UMBUS L.ABORATORIES
-------
15.0
10.0
9.0
8.0
VI 7.0
~
o 6.0
(5
"0
- 5.0
o
:g 4.0
o
E 3.0
-
VI
o
U
2.0
1.0'
80.0
VI-35
~~
/
/
Steel plant precipitators,.
installed cost
1968 costs
/
/
/
/
I
/
/
/
/
/ Cost trend.
I,precipitators
/ only, 1968 costs
II (Ref. 66)
/
90.0
95.0 98.0 99.0
Collection Eff i ciency , percent
FIGURE VI-27. INSTALLED COST FOR STEEL PLANT
ELECTROSTATIC PRECIPITATORS AS
AFFECTED BY COLLECTION EFFICIENCY
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI - 3 6
Effect of Efficiency Specifications Greater Than Current Legal
Requirements on the Cost of Air-Pollution Control Equipment
The remainder of Section VI is adapted from Appendix C in the com-
panion report, "Final Economic Report on A Cost Analysis of Air -Pollution
Controls in the Integrated Iron and Steel Industry", dated May 15, 1969,
and is a result of work done by the Swindell-Dressler Company, Division of
Pullman Incorporated, the subcontractor on the study.
Because of the dearth of experimental and empirical data on the relationship be-
tween collection efficiency and the cost of air-pollution control equipment, resort was
taken to estimation of this relationship from theoretical considerations. The estimates
based on theoretical considerations then were evaluated against the limited amount of
empirical data that could be obtained.
In many localities, current legal requirements specify a permissible particulate
emission at the stack of not more than 0.05 grain per dry standard cubic foot of gas (or
equivalent) exhausted to the atmosphere.. Some facilities have met this requirement (or
even exceeded it) with even fine, submicron sized steelm,aking dust by using high-
efficiency filters, scrubbers, and precipitators. Manufacturers have been able to guar-
antee this performance with their equipment in a variety of applications. Also it is noted
that blast-furnace gas has been cleaned to as low as 0.005 grain/ DSCF':' when nece s sary
for reuse of the gas in high-energy burners and fine checkerwork of the blast stoves
(although this is a coarser dust than from steelmaking).
This index "0.05" is not necessarily an ultimate measure of the effluent quality
that can be obtained. It came into use in the early 19601s, on the basis that an open
hearth furnace stack plume containing fume at such a concentration had an 'Iacceptable"
appearance in many steelmaking areas. The value "0.05" correlated approximately
with the maximum efficiency of electrostatic precipitator s normally offered by manu-
facturers at that time for collecting this fume. However, the rapid growth in the use of
oxygen lancing of steelmaking furnace s has led to larger quantitie s of finer fume in their
waste gases today. .
A stack plume cleaned to this level is not an invisible plume. The very fine steel-
making fume escaping at the stack causes a larger degree of scatter of transmitted light
than the larger particles previously encountered(13), and thus may be visible even in
low concentrations. Yet, visibility of an exhaust plume persists as a means of checking
collector performance, because it involves a simple comparison of "equivalent opacity"
of the plume against the Ringelmann Smoke Chart.
Local code limitations based on Ringelmann opacity judgments may find a concen-
tration of 0.05 grain/DSCF of steelmaking fume unsatisfactory. Where local codes are
based on a schedule of allowable fume emission weight per ton throughput of processed
material, the permissible fume rate customarily decreases for larger production equip-
ment, so that above 30 to 40 tons of proce ss weight per hour, the O. 05 level of control
may not be adequate.
°DSCF = dry standard cubic foot.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORAT,ORIES
-------
VI-37
Thus, the widespread use of 0.05 grain/DSCF. as a general limiting level for
emissions led to its choice as a basis for calculating the size and cost of collectors for
each process in the tabulations in Appendix D of this report. But, in recognition of the
use of more restrictive enforcement methods in some steelmaking areas, and because of
the trend in promulgating air-quality criteria which may suppress the emission sources
in an area to an increasing degree, the following indications are drawn of the difference
in cost for fume-collecting systems capable of an efficiency beyond the currently prac-
ticed or currently attainable level.
Performance Equations
The performance equations of gas cleaners, as currently understood and applied to
select the size and operating parameters for a particular cleaning application have this
in common - they are of the form:
Y) = 1 - e-F(x)
where
Y) = collection efficiency
or l-Y) = penetration, dust loss, or outlet concentration as a
fraction of the inlet concentration to the gas cleaner.
It corresponds to some figure like O. OS, for example;
0.05 (grains/ DSCF)
l-Y) == inlet conc. (grains/ DSCF)
In (l-Y)) = -F(x)
F(x) is a function of the
of the colle ctor .
size and operating parameters
Fabric Filters. The equation for the performance of fabric (bag) filters, has been
shown to be of the form(67):
S Dr
Y) = l-e - 0-
D
D' f . f~
D is the target efficiency and is a unctlon 0 Vf'
where
D = fiber diameter
g = gravitational constant
v = velocity of gas at filter face
- Q - flow rate of gas
- A - area normal to flow = face velocity
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-38
f = settling velocity of particle, as from Stoke's Law.
S = total projected area of all fibers in the filter
o cross section of filter bed '
both normal to the gas flow.
The relationship between the target efficiency ~' and the function W is shown in
Figure VI-Z8.
100
,
,
,
\\
\
\~ ~Spherical obstacles
01/0 (Spheres) = zero -
" at Og/Vf = 24
,~ 0'/0 (Cylinders) = zero -
Cylindrical!>::""I... at Og/Vf = 16 \-
obstacles ,. ~...:::.. ~ J ,
I -
0'/0
00 1
2 3 4 5
16
Q9..
Vf
FIGURE VI-Z8.
. ,
RELATIONSHIP BETWEEN TARGET EFFICIENCY(~)
AND ~ FOR FABRIC FILTERS .
Electrostatic Precipitators. The equation for the performance of an electro-
static precipitator takes the form of the Deutsch equation(lS, 68):
y) = l-e-A'fl/Q
where
AI = collecting surface area
~ Z(k-l)] rEZ
fl = C + (k+Z) J 67TJ.l = drift velocity
k = dielectric constant of particles
E = ele ctric field strength
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-39
r = particle radius
f.l = gas viscosity at temperature.
Wet Scrubbers.
A correlation established for wet scrubbers is
-ofp p )"1 -a(P )\Y
7) = l-e \: G + L = l-e \: T
as fullows(69):
whe re
P G = contacting power of gas stream
=O.157FS
FS = pressure loss across scrubber, in. water, exclusive of
loss due only to velocity changes or friction losses across
dry portions of the equipment .
PL = contacting power of liquid stream
qL
= O. 583PF Q
PF = liquid feed pressure, psig
qL = liquid feed rate, g.p.m.
PT = PG + PL = HP/lOOO CFM, based on Q
Q = actual gas flow at the scrubber, CFM
a, "I = constants for a particular dust, related to particle size
and size distribution.
Theoretical Factors Affecting Performance
To increase the efficiency of a collector, whose performance is describable by
this logarithmic-de cay-type function, it is necessary to increase F(X). The variable
flow and equipment parameters comprising F(X) for a particular dust are respectively:
Bag Filter So, ~ = So, ~A
AIE2 2L W n E2
Precipitator ~ = A V f.l
=
2L W n E2 2LE2
n b W V f.l = bVf.l
where
W = collector surface span normal to flow
L = collector surface length in direction of flow
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI -40
n = number of collecting ducts
b = separation of collecting surface s
Scrubber FS and PF(qL/QI)
where
qL/Q' is the liquid/gas ratio (gal/lOOO CF).
Effects of Propertie s of Particle s
(l) Increasing f increase s YJbag filter
f = 4g r2 p
l8f-.L
where p is the density of particle.
So larger, denser particles are collected more easily.
(2) Increasing r increases YJprecipitator' Again, larger
particles are more easily attracted to the collector. In.
creasing the dielectric constant of the particulate, and
decreasing re sistivity by pre -conditioning via temperature
and humidity (or S02 addition) increases YJprecipitator.
(3) Increasing a or 'Y increase s YJscrubber, as can be
established(69). Both increase with particle size.
Effects of Geometry of Dust Collector
(1) Increasing filter thickne ss or mat density, or decreasing
air/ cloth ratio by using larger bag surfaces will increase
YJbag filter'
(2) Increasing precipitator length in the flow direction, or de-
creasing plate spacing or tube diameter (within limits of
electrical stability) will increase YJprecipitator' Because
the dust loading decreases in the flow direction, it is
possible to achieve an economy by successive stages of
precipitation, each optimized electrically for maximum
efficiency at the respective loading it will see, rather than
simply extending the fir st- stage field.
(3) Decreasing the throat area of a scrubber increase s its
pressure drop and increases YJscrubber. This can be done
by variable geometric arrangement or by increased water
rate.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-41
Effects of Utility Parameters
(1) A partially blinded filter will be more efficient, but at a
cost of higher pressure drop and higher fan horsepower.
(2) Increasing electric field strength increases YJprecipitator
within the limits imposed by the geometry of the collector
and dust properties with respect to sparking. This limit
can be approached more closely with safety if automatic
controls are used to regulate the discharge. Energy use
ri se s.
(3) Venturi Scrubber. Increasing water usage or delivery
pressure in a scrubber increases 7Jscrubber. Increased
gas-pressure drop gives improved efficiency at the cost of
higher fan horsepower.
Effects of Flow
(1) Even though an increase in face velocity (~) gives a higher
theoretical efficiency in the inertial effect range, the effect
is reversed in dealing with small particles «1J..L). For a
filter with a fixed pressure drop and fixed cleaning routine,
the dust buildup will dominate, so that if increased loading
blinds the filter, causing spillage and less net cleaning, then
the following holds. Decreasing the quantity of gas treated
or using a larger filter for lower face velocity increase s
7Jbag filter'
(2) Decreasing the amount of gas treated by lowering precipitator
velocity and increasing residence time increases 7Jprecipitator
if distribution of the gas is maintained uniform between the
plates.
(3) Increasing the quantity of gas treated or increasing throat
velocity increases 7Jscrubber by increasing pressure 10ss
across the constriction, with an increase in PG and fan
horsepower.
Effect of Temperature on Viscosity
(1) Increasing temperature increases J..L gas, decreases
7Jbag filter, and decreases YJprecipitator'
(2) Increasing temperature increases the quantity of gas handled,
again lowe ring the se efficiencie s.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI -42
(3) Besides altering flow and settling or drift velocity, tempera-
ture also endangers the bags, structures, and mechanisms of
the collectors. Filters and dry precipitators must have an
inlet temperature above the water vapor (and sulfuric acid)
dew point to avoid corrosion and dust caking on the collector,
and causing dust handling problems in disposal conveyors.
(4) Temperature affects the scrubber mainly in increasing the gas
flow and increasing the saturation-water requirement.
Control System Cost Change s
It is a property of decay functions of the aforementioned type that, at high effi-
ciency, an increasingly large change in the exponent is required for a small incremental
increase in efficiency.
Electrostatic Precipitators. For example, for an electrostatic precipitator, it
has been stated that the precipitator unit size increases with respect to efficiency
change s as follows(70):
Overall Efficiency
for a Particular Dust
90 percent
99 percent
99. 9 pe rcent
Outlet Loading With 5.0
Grains/ DSCF Input Loading
0.5
O. 05
O. 005
Size of Precipitator
Box and Unit Cost
X
2X
3X
This tabulation excludes ductwork, water sprays, hood with its cooling auxiliaries,
and stack; but include s the precipitator and its electrical components. The fan and
motor size and cost, for a IX increase in precipitator size would be affected by an in-
crement corre sponding to an increased static pre s sure (S. P. ) of about 1-1/2 inche s of
water (the loss through Box X), with the volume remaining unchanged.
For a precipitator increment of X:
Horsepower increment = S. P. + 1. 5 x H P
S.P. ..
Fan-pressure increment = S. P. + 1.5 x S P
S. P. ..
for the total system fan.
Fan volume unchanged.
The effect of increasing the size of an electrostatic precipitator on operating effi-
ciency has also been reported in the literature for an electrostatic precipitator collecting
open-hearth emissions. (71) The relationship developed between the collection efficiency
and size of the precipitator (as shown by the square feet of collecting surface) is shown
in Figure VI-29. (71) The results of the study have shown that removing the dust from
315,000 cubic feet per minute of open-hearth waste gas required 58,300 square feet of
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-43
collecting surface area for an efficiency of 95 percent. An increase in the collecting
surface area to 96,500 square feet (an increase of 66 percent) resulted in an increase in
efficiency of only 4. 3 percent to 99.3 percent.
100
90
+- 80
c
~ 70
~
~ 60
~
~ 50
c
Q)
'0 40
:E 30
w
20
10
00
~.~ .- .~
-.- ~ .- ..- -- ~-- -
...... I
./ ' !
/ I I
I
)/ ! :
/ I !
V I I
1
J g! ~r
I rtI
m,
I 1 I
I ! !
0.5
1.0
1.66
o
o
8
8
o
(\J
~ ~ ~ ~ ~
Collecting Surface, sq ft
o
8
(X)
~
8
o
o
o
FIGURE VI-29.
RELATIONSHIP OF ELECTROSTATIC PRECIPITATOR
COLLECTING SURFACE TO COLLECTION EFFI-
CIENCY FOR OPEN-HEARTH EMISSIONS
The above variation in size corre sponds to Deutsch f sLaw:
(1-1')) =e
AfI'
--
Q
for a particulate of homogeneous size, shape, density, and composition.
inlet loading -R = 1-~. h R th t1 t 1 di
1') = inlet loading 1. L. ' were = e ou e oa ng.
R t t -constant2 x length
= cons an 1 x e
log R = constant3 + constant4 x length
for a given process and precipitator.
Case 1 Illustrative of Deutschf sLaw
R:
5 grains
SCFD
--
Box
X
--
0.5
--
Box
X
--
O. 05
--
Box
X
--
0.005
1')(percent): 90 90 90
net 1')(percent): 90 99 99.9
net size: X 2X 3X
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-44
However, real particulate varies in size, density, and susceptibility to charging
(depending on surface and compositional variables), so that the least collectable particles
remain after each treatment, lowering the efficiency of subsequent treatments. (72)
Case 2 below illustrates this with arbitrary efficiencies:
Case 2. Illustrative of Deutsch's Law
5 -+ Box Box B'ox' Box Box Box
R: -+ 0.5 -+ -+ 0.1 -+ -+ O. 03 -+ -+ 0.012 -+ -+ 0.006 -+ -+ 0.005
X X X X X 5/12X
11 (percent): 90 80 70 60 50 40
net 11 (percent): 90 98 99.4 99.76 99.88 99.9
net size: X 2X 3X 4X 5X 5-5/12X
A body of blast-furnace data(73) for a number of operating furnaces at various
precipitator loadings yields the following progre ssion, which shows this trend,
net y)(percent):
90
95
98
99
99.5
99.9
99.99
net size:
0.55X
0.86X
1.4X
2X
2.65X
4.5X
7.5X
Cost data from precipitator manufacturers indicate a close correspondence to
Case I in variation of cost (= constant x length) with efficiency. Guarantees are made on
efficiency rather than outlet loading, because the precipitator is not adequately adjust-
able for cleaning a higher inlet dust concentration to the same outlet level (say 0.05).
In fact, the higher loading may reach a point where spark-over occurs; so automatic
electrical controls are used to maintain the highest collection efficiency just short of
spark-over. (Large loading differences require design selection of plate spacing and
voltage optimized for the loading and dust properties of the individual process effluent).
The maximum guarantee is presently about 99. 5 percent, although higher efficiencies
(around 99.7 percent) can be reached.
The successive lowering of efficiency found with addition of identical precipita-
tion units can be compensated for. Because each successive unit sees a lower dust
loading, plates can be spaced more closely, and voltage optimized in each succeeding
section, while avoiding spark-over. Still, each type of dust must be te sted to deter-
mine its collectability as a function of precipitator length.
Tables VI-4 through VI-9 show some estimated cost changes for processes cleaned
by electrostatic precipitation to various outlet dust concentrations. The variation is
based on the Deutsch Law. Capital cost changes include:
Materials: precipitator + fraction of electrical.
Labor: corresponding to each of above at standard factors.
Engineering:
scaled fraction materials plus labor.
Annual operating cost changes include 0.24 (Capital change) + fraction of electric
power for precipitator and horsepower increments. Only small variations were noted
for capacity of the cleaner, so that only the central size of cleaners is included for
processes estimated in Appendix D at 0.05 grain per SCFD.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS. LABORATORIES
-------
TABLE VI-4.
VI -45
ESTIMATED COST DIFFERENCES FOR A SINTER
PLANT ELECTROSTATIC PRECIPITATOR AS
AFFECTED BY THE OUTPUT DUST LOADING
FROM THE WINDBOX
Plant Capacity
Gas Volume
Input Dust Loading:
6,000 net tons per day
630,000 ACFM at 325 F
0.8 grain per SCF(a)
Outlet
Loading
(R)
Capital Cost
Difference (6 KR)
KO. OS' percent
Annual Ope rating
Cost Difference (6 CR)
CO. OS' percent
Annual Direct Operating
Cost Difference (6 DR)
DO. OS' percent
O. 125
O. 050
- 29
o
O. 020
+ 29
R
log10 (R )
0.05
log10 (2.5)
- 23
o
+ 23
=
-6K
R
18%
=
-6C
R
15%
- 17. 5
O. 0
+ 17.5
=
-6D
R
10%
(a) 4 grains per SCF effluent precleaned by 80 percent efficient recovery cyclones.
ESTIMATED COST DIFFERENCES FOR A SINTER
PLANT ELECTROSTATIC PRECIPITATOR AS
AFFECTED BY THE OUTPUT DUST LOADING
FROM MATERIAL HANDLING
TABLE VI-5.
Plant Capacity
Gas Volume
Input Dust Loading:
6, 000 net tons per day
250,000 ACFM at 135 F
1. 0 grain per SCF
Outlet
Loading
(R)
Capital Cost
Difference (6 KR)
KO. OS' percent
Annual Operating
Cost Difference (6 CR)
CO. OS' percent
Annual Direct Operating
Cost Difference (6 DR)
DO. OS' percent
O. 125
0.050
- 18
o
0.020
+ 18
R
log10 (R )
0.05
log10 (2.5)
- 15
o
+ 15
=
-6K
R
18%
=
-6C
R
15%
- 10
o
+ 10
=
-.6D
R
10%
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-46
TABLE VI-6. ESTIMATED COST DIFFERENCES FOR A BOF
ELECTROSTATIC PRECIPITATOR AS AF-
FECTED BY OUTPUT DUST LOADING
Furnace Size
Gas Volume
Input Dust Loading: .
200 net tons
785, 000 ACFM at 500 F
4. a grains per SCF
Outlet
Loading
(R)
Capital Cost
Difference (6KR)
KO. 05' percent
Annual Operating
Cost Difference (6 CR)
CO. 05' percent
Annual Direct Operating
Cost Difference (6 DR)
DO. 05' percent
O. 125
O. 050
O. 020
- 9
a
- 10
a
- 11
a
+ 9
+ 10
+ 11
R
log10(R )
O. 05
log10 (2.5)
-6K
R
9%
-6C
R
10%
=
-6D
R
11%
=
=
.TABLE VI-7. ESTIMATED COST DIFFERENCES FOR AN OPEN
HEARTH ELECTROSTATIC PRECIPITATOR AS
, AFFECTED BY OUTPUT. DUST LOADING
Furnace Size
Gas Volume
Input Dust Loading:
200 net tons
85, 000 ACFM at 500 F
5. a grains per SCF
Outlet
Loading
(R)
Capital Cost
Difference (6KR)
KO. 05' percent
Annual Operating
Cost Difference (6 CR)
CO. 05' percent
Annual Direct Operating
Cost Difference (6 DR)
DO. 05' percent
O. 125
O. 050
0.020
- 10
a
- 9
a
- 7
a
+ 10
+ 9
+ 7
R
log 10 (R )
0.05
log 1 a (2. 5)
-6C
R
9%
=
-6D
R
7%
-6K
R
10%
=
=
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-47
TABLE VI-8. ESTIMATED COST DIFFERENCES FOR AN ELECTRIC
FURNACE ELECTROSTATIC PRECIPITATOR AS
AFFECTED BY OUTPUT DUST LOADING
Furnace Size 150 net tons(a)
Gas Volume: 185,000 ACFM at 500 F(b)
Input Dust Loading: 3. 0 grains per SCF
Outle t
Loading
(R)
Capital Cost
Difference (.6KR)
KO.05' percent
Annual Operating
Cost Difference (.6 CR)
CO.05' percent
Annual Direct Operating
Cost Difference (.6 DR)
DO.05' percent
b
O. 125
- 10
- 10
- 9
0.050 0 0 0
0.020 + 10 + 10 + 9
(~) -.6KR -.6 CR -.6 DR
logl 0 R.05
- . = =
log10 (2.5) 10% 10% 9%
(a) Two-furnace system.
(b) Assumes humidification of process fume is capable of maintaining particle resistivity in satisfactory collection
range.
TABLE VI-9. .ESTIMATED COST DIFFERENCES FOR A SCARFING
MACHINE ELECTROSTATIC PRECIPITATOR AS
AFFECTED BY OUTPUT DUST LOADING
Gas Volume : 100,000 ACFM at 100 F
Input Dust Loading: 1. 0 grain per SCF
Outlet
Loading
(R)
Capital Cost
Difference (.6KR)
KO.05' percent
Annual Operating
Cost Difference (.6 CR)
CO.05' percent
, Annual Direct Operating
. Cost Difference (.6 DR)
DO.05' percent
O. 125
- 19
- 18
- 17
0.050
o
o
o
0.020
+ 19
+ 18
+ 17
. (~)
log10 R.05
log10 (2.5)
-.6KR -.6CR
= 19% = 18% =
-.6 DR
17%
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI -48
Wet Scrubbers. In the case of the ventUri scrubber, it has been stated that the
following relate s capital cost to efficiency(70):
Overall Efficiency for a Given
Dust at Inlet Load of
Inlet Grain/ DSCF, percent Outlet Loading Capital
1.0 3.8 5.0 10 Grains/ DSCF Cost
90 97.4 98 99 0.10 X
96.2 99 99.24 99.62 0.038 1.43X
The operating expense s vary similarly for a venturi scrubber as efficiency is in-
creased. This is shown in Figure VI-30(48) for an open-hearth application where a de-
crease in outlet loading from O. 1 to 0.01 grain/SCFD results in more than doubling the
annual operating cost of the fan. For a given size adjustable venturi, the increased
efficiency requires an increase in available horsepower to the fan and selection of a
higher pressure fan. Operating power consumption increases directly with the pressure
drop.
g 0.10
~ '+- 0.08
C'lu
.~:.o 0.06
"0 :J
C u
.5"E 0.04
- c
1/)"0
:J c:
o.E
+- I/) 0.02
:J ...
Q.CV
-Q.
:J'-
o ~ 0.01
~ 0.008
C'I 26
Ore and lime boil
and working period
Charging, melt down
and hot meta I
30 32 34 36
Pressure Drop, inches of water
. . I
45 55 65
Fan Operating Cost (1960),
thousands of doHars per year
38
40
28
.'
35
75
FIGURE VI-30.
RELATIONSHIP OF OUTPUT DUST LOADING TO PRESSURE
DROP AND FAN OPERATING COST FOR A VENTURI
SCRUBBER OPERATING ON AN OPEN-HEARTH
FURNACE(48)
The above venturi-cleaned open-hearth application involves oxygen lancing during
the periods noted on the upper curve. Dust loading was low (0.82 to 0.87 grain/SCFD
during oxygen periods, and 0.35 to 0.45 grain/ SCFD during the charging, melting and
hot-metal addition periods). When these data are corrected to a typical peak of 5 grains/
SCFD loading for today's oxygen-lanced furnaces, it yields the following correlation:
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-49
(grains )
R, outlet loading SCFD
.6P, venturi pressure drop (in. w.)
O. 125
O. 05
O. 02
34.7
41. 0
48.2
6PR -0.178
.6 PO. 05 = (0.~5)
However, the contacting power concept(69) has been applied for the regression line
of a plot of non1anced open hearth gas-cleaning efficiency versus pressure dror at various
operation conditions(46). This give s for a peak 5 grains/ SCFD inlet loading:
R (grains/ SCFD)
O. 125
O. 05
O. 02
.6 P (in. w.)
44
78
136
.6 PR ( R )-0. 62
.6 PO. 05 = 0.05
The numerically higher exponent seems more in line with results from other steelmaking
fume.
Blast-furnace data give the following(46):
(grains )
R, outlet loading SCFD
.6P, venturi pressure drop (in. w.)
0.125
0.05
0.02
15.8
23.0
33.2
.6PR = (~)-O. 403
.6PO.05 0.05
Data for an electric furnace making 20 percent ferrosilicon show(46):
.6PR ( R )-1.53
.6PO.05 = 0.05
For the typical scrap-charged electric-arc furna.ce, wet scrubbing applications are
spar se, and data are not available for a scrubbing power -efficiency correlation.
Venturi gas cleaning data on the basic oxygen furnace have been developed as
follows(74):
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-SO
(grains )
R, outlet loading SCFD
~P, venturi pressure drop (in. w.)
O. l2S
O. OS
O. 02
27
41
60
gi ving
~PR
~ PO. OS
= (~)-O. 417
O.OS
Data from a pilot-size conventional venturi scrubber applied to clean scarfing-
machine effluent have been published as follows(7S):
(grains )
R, outlet loading SCFD
;~P, venturi pressure drop (in. w.)
0.12S
O. OS
O. 02
34
60
108
~PR = (~)-0.63l
~PO.OS O.OS
Tables VI-IO through VI-12 give some estimated cost differences for processes
cleaned by wet scrubbers of the high-energy types. The variation is based on the pre-
ceding scrubber -application data. Capital cost change s include:
Materials:
Fan and motor + fraction of electrical. The venturi
itself is assumed adjustable and of sufficient strength
for the higher pressure difference across its walls.
Water rate s are unchanged. .
Labor:
Corresponds to each of above at standard factors.
Engineering: Scaled fraction of materials plus labor.
Annual operating cost changes include 0.24 (capital change) plus electric power for
hor sepower increments.
An empirical relationship is indicated as follows:
~KR ~CR ~DR ~PR-~PO. OS ~PO. OS r ~PR ]
S.S%=9'%=1.2% = 60-41 = 19 LLPO.OS-l
~ -0.417]
= ~~ L(o,~s) -1
~K ~C ~D [ ~P ] ~ -0.417]
R - R - R - R -1 - ~ -1
l1.8%-19.4%-2S.9%- 6PO.OS - (O.OS)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-51
TABLE VI-I0. ESTIMATED COST DIFFERENCES FOR A BOF WET SCRUBBER
AS AFFECTED BY OUTPUT DUST LOADING
Furnace Size 200 net tons{a)
Gas Volume: 440,000 ACFM at 180 F
Input Dust Loading: 4.0 grains per SCF
Outlet
Loading
(R)
Venturi Pre s sure
Drop (.6 P),
inche s of water
Capital Cost
Difference (.6KR)
KO. OS, percent
Annual Operating
Cost Difference
(.6 CR) CO. OS,
percent
Annual Direct
Operating Cost
Difference
(.6 DR) DO. OS,
percent
0.125
0.050
27.5
41. 0
- 4.0
O. 0
- 6
o
- 8
o
O. 020
60.0
+ 5.5
+ 9
+ 12
(a) One furnace system.
TABLE VI-II. ESTIMATED COST DIFFERENCES FOR AN OPEN-HEARTH WET
SCRUBBER AS AFFECTED BY OUTPUT DUST LOADING
Furnace Size 200 net tons{a)
Gas Volume: 90,000 ACFM at 180 F
Input Dust Loading: 5.0 grains per SCF
Annual Direct
Annual Operating Operating Cost
Outlet Venturi Pre s sure Capital Cost Cost Difference Difference
Loading Drop (.6P), Difference (.6 KR) (.6 CR) CO. OS, (.6DR) DO.05,
(R) inches of water KO. OS, percent percent percent
0.20 32.6 -11.0 - 15.5 - 19.5
0.10 50.2 - 6.5 - 9.0 - 12.0
O. 05 78.0 0.0 0.0 0.0
.6KR .6CR .6DR .6PR-.6PO.05 .6PO.05 [.6PR l
-6.5%= -9% = -12% = 50.2-78 = -27.8 - .6P -J
0.05
It - O. 62 ~
= -2~~8 ~O. ~5) - J
.6KR .6CR .6DR f.6PR l ft R -0.62 ]
18.20//25.1% = 33.5% =[.6 PO. 05 -~ = ~O. 05) -1
(a) One-furnace system.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI - 5 2
TABLE VI-12.
ESTIMATED COST DIFFERENCES FOR A SCARFING MACHINE
WET SCRUBBER AS AFFECTED BY OUTPUT DUST
LOADING
Gas Volume
Input Dust Loading: .
100,000 ACFM at 100 F
1. 0 grain per SCF
Annual Direct
Annual Operating Operating Cost
Outlet Venturi Pressure Capital Cost Cost Difference Difference
Loading Drop (.6P) Difference (.6KR) (.6CR) CO. 05, (6DR) DO. 05,
(R) inches of water KO.05, percent percent percent
0.083 44.3 - 9 - 17 - 23
0.050 61. 0 0 0 0
0.030 81. 5 +11 + 21 + 28
.6KR .6CR .6 DR
11 % = 21% = 28% =
.6PR-.6PO.05
81. 5-61
=
.6P 0.05
20. 5
[.6P ]
R 1
.6PO.05-
61
= 20.5
[(OHOS) -0.631 -I]
.6K .6C .6D
R R R
32.6%= 62.5% = 83.4% =
~ .6P ]
R -1
.6P 0.05
=
[(O.HOS) -0.631 -I]
Doubling the venturi pre ssure drop would cause a 25.9 percent increase in direct
operating costs. Venturi loss of 41 in. w. is 85 percent of system loss, which accounts
for 40 percent of total horsepower (including an unchanged water pumping and treatment
system) in this case. .Power cost is about 72 percent of direct operating costs.
. Fabric Filtration. For an acceptable constant dust penetration through a fabric
filter, the face velocity or air volume -to-cloth area ratio must decrease with decreasing
particle size and density or with increased inlet loading. For a lower allowable penetra-
tion, the face velocity would similarly decrease. Thus, a more difficult or more thori
thorough cleaning job would involve increased cost to provide more filter surface area.
This is exemplified in the extreme case of a reverse jet-cleaned filter where face ve-
locities are the highest encountered. Nomograms that can be used to estimate the size
of reverse jet filters are shown in Figure VI-31 and VI-32. (76)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI - 5 3
40
~ 50
c:
e 60
v
A i
0 70
. 80
o
;;
.D
~ 90
""
co
aP
100
30
2'
20
U
C
10
..:
U
.....
'"
Q
C) ,
z
a
«
9
~
on
:>
Q
The following nO/11ograph is presented
as a convenient means of selecting
Filter Ratio for preliminal'y deter-
mination of the size Aerotul'n Dust
Collector' that will best satisfy the
needs of your installation.
In many instances the nomograph will
pl'Uvide cletenllination of the optimum
Filter Ratio. Because of the great val'i-
ety of possible service conditions and
the effect of the charactcl"istics of
specific dusts. final detenninat ions of
Filtel' Ratio will be made by Buffalo
Forge Company. This (Jl'OCedul'e pro-
vides the gl'catest assunmce of COITect
and economic selection of equipment
fOl' your installation.
IIOW TO I;SJ~
In OJ'der to select Filt.er Ratio, three
conditions pertaining to your specific
dust collection job ar~ needed. They arc:
a. The approximate percentage, by
weight, of dust particles 10 microns
or smalle~.
b. Dust content of the air enterin
-------
36.000
:".000
32.000
30.000
28.000
26.000
24.000
22.000
t
~ 20.000
~
U
~ 18.000
..
u
e-
o'
~
~
~
~
..
~
~
..
~
'"
; 16.000
ii:
14.000
12.000
10.000
8.000
6.000
2.000
°
FIGURE VI-32.
VI-54
FilTER SIZE
200 16.8
400 16.10
16.11
32.8
600 16.14
16.16
16.18
800 32.\0
16.20
~8.8
t
ci
!2
1.000 32.12
1.200 )2.16
48.10
64-8'
32-14
1.400
48.12
. 32-18
1.600 64-10
32-20
48.14
1.800
IIOW TO LOSE
This chart provides a conven-
ient and accurate means for
selecting the applicable size or
sizes of Aeroturn Dust Co 1-
Il'ctors whl'n Filter Ratio and
required Air Cleaning Capacity
are known.
1) Draw a line from the re-
quired Capacity through the
applicable Filter Ratio to in-
tersect the Filter Area scale.
2) From this pOint of inter-
section, draw a ho,.jzontal line
through blocks designating
Filter Size selections for de-
sired Capacity.
64.16
3) If horizontal line passes
through more than one Filter
Size, first size intersected will
be most economical. Subse-
quent selections will be less
economical.
.~
..
::
~
ii:
2.000
64.12
48-16
64-18
4) For capacities larger than
shown: Use % the required ca-
pacity in the above procedure.
Filter Size thus selected must
be doubled for full capacity.
2.200
48.1,8 6.-14
2.400
2.600
48.20
64.20
NOMOGRAM FOR ESTIMATING THE SIZE OF A
REVERSE-JET FILTER
2.800
3.000
3,200
3.400
3.600
A reverse -jet cleaned fabric filter calculated from the nomograms shown in
Figures VI-31 and VI-32 for dust having 70 percent of its weight less than 10 microns
in size and a specific gravity above 2.0 yields the following information on sizing of the
filter:
Efficiency Required for
0.05 Grain/ DSCF
Outlet, percent
Inlet Dust Loading,
grains/ DSCF
5
10
20
25
Filter Ratio,
cfm/ sq it
16.4
13.7
9.6
7.6
99
99.5
99. 75
99.8
Q/ A Filter
Area
X
1.2X
1.7X
2. IX
The effective filtering body is the dust cake layer on the bags.
this time seem amenable to treatment which will improve efficiency.
when new, and to a lesser extent when cleaned, holds little dust cake
This doe s not at
However, the bag,
so that the fabric,
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-55
with its small, dust-laden fibers is the basic filter until the filter cake layer reforms.
As the small fibers break in service, the bag loses filtration capability. Additionally,
the lower flow resistance of a cleaned bag passes a greater volume of air at reduced
cleaning efficiency than when it is dust-coated; but at a higher velocity, which improves
the collectability of larger particles and worsens the diffusiona.l efficiency dominating
small-particle colle ction.
An adequately designed baghouse will have a bag-cleaning cycle suited to the inlet
dust loading from the proce ss to which it is applied. This cycle is often automatically
adjustable, so that the filter maintains the same average (time-wise) efficiency with
variations in inlet dust loading and gas volume. The bag-cleaning period will begin when
the collected dust cause s the pre ssure drop through the filter to reach a set-point
pressure.
In addition, the fabric weave and material are chosen with the special character
of the process effluent in mind (such as particle size distribution). Economic factors
(bag life and initial cost differences) also enter this choice, but increased efficiency can
be achieved only by choosing from a group of fabrics which will give cleaning to the pro-
jected required level. Present practice usually gives efficiencies of 99 percent +, and
bag filters frequently give the highest efficiencies of the applicable cleaning devices con-
sidered for a process. Therefore, this selective optimization does not offer much
potential except as re search may reveal new materials and weave s.
As a case in point, a process having a generally large particulate may be ade-
quately cleaned by a certain bag to 0.05 grain/ SCFD. If lower outlet loading is required,
a suitable bag which gives similar results on a process with finer effluent may be sub-
stituted. The overall cost mayor may not be larger. The choices are presently limited
by limited test results on filtration properties of fabrics, and state-of-the-art in fabric
technology with regard to dust abrasion, flexural durability, and chemical and tempera-
ture re sistance. Electrostatic interactions of various fabrics with dust particle s may
prove to be significant.
No clear correlation has been advanced relating efficiency to operating parameters.
However:
A higher pressure drop may be expected to increase filtration action at
the cost of additional power, but the trend of such variation is not
known.
A higher filter -face velocity (higher air/ cloth ratio) theoretically
yields a higher efficiency of collection for particle s large enough to
be governed by inertial laws , but the se are ordinarily cleaned to
nearly 100 percent efficiency, so the filter size is governed by
loading. The small particles which escape collection migrate under
diffusional impulses, and efficiency here would increase with re-
sidence time (lower face velocity, lower air/ cloth ratio, thicker
filter media). The relative effects of the se coacting collection
mechanisms is not sufficiently understood at pre sent for use in
practical de sign.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LAE!ORATORIES
-------
VI-56
The se foregoing factors are insufficiently defined at pre sent for a useful definition
of the effect on costs of changed efficiency requirements of fabric filters.
Conclusions
A situation of diminishing returns is indicated by performance equations of the
exponential type. In many case s a O. 05 grain/ SCF D outlet concentration become s a
practical maximum level for improving efficiency, even though it is by no means an
absolute limit.
The state of the art, then, allows the gas-cleaner manufacturer to predict per-
formance, design a collector, and guarantee it with some confidence up to about 0.05
grain/SCFD outlet loading for particles greater than 2 microns in size. At lower output
levels, his experience is limited. The large change in size or operating parameters re-
quired for further small increases in efficiency would magnify the uncertainties known
to exist in these simplified exponential relations with their empirical constants.
Measurement techniques used to determine dust loading in the ducted stream be-
fore and after the collector leave much to be desired, especially where small concentra-
tions and even smaller changes in concentration are to be used as evidence of guaranteed
performance. Lack of homogeneity of most dusts from iron and steel processes make
the use of monitored data (light scattering or transmission, for example) difficult to
interpret, or the equipment difficult to calibrate, for all the variations in dust composi-
tion, size, gas flow rate, etc. caused by process changes during a heat cycle, or from
heat to heat. Isokinetic sampling (sampling at stream velocity) with traversing probes
involves much averaging (in time and space) with calculation and readjustment continuing
during the traver se. This is co stly and of que stionable accuracy. Null probe s, too,
operate with a significant degree of error in trying to balance small pressure differences.
Neither approach to isokineticity can give a time history of emission rate during the
course of a rapidly changing heat cycle because only two or three traverses can be run
at best in an hour. Gas density (composition and state) and moisture content data should
be monitored continuously and used as input to sampling-rate determinations during the
course of a sampling test. Deviations here can seriously affect the loading measured
as grains of dust per dry standard cubic foot of carrier gas.
From such quantitative data as can be obtained, control equipment is de signed
(often with a costly excess performance factor built in) and guaranteed somewhat con-
servatively. The guarantee is proven (or indicated) by standard sampling tests, and
no assurance is given that any particular level of Ringleman chart greyness will not be
exceeded. Research is needed to find a method to measure dust concentrations inexpen-
sively; and agreement is needed to correlate de sign and enforcement base s of
measurement.
Very fine particulate matter, because of greatly extended surface area, causes a
much greater scattering of light, even in small concentration. A Ringelmann compari-
son should thus in some way account for the nature of the emission being sampled. If
this correlation can be made, then this economical method of testing might be used to
obtain adequate design data.
In the case of very fine steelmaking dusts from open hearth, electric arc, and
basic oxygen furnaces; the collector performance is difficult to predict because of the
following:
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-57
(1)
The particle - size distribution is difficult to quantify with pre sent
methods for sampled dust, and the correlation of the se data to
"in situ" dust in the furnace effluent gas is in doubt. (Large dis-
crepancies in reported BOF dust sizing are a case in point.) The
smaller the size, the greater the difficulty.
(2) Agglomerative properties of the dust are not well established and
the effect of this on sampled dust sizing and on collection mechanisms
in the gas cleaners is not well understood.
(3) The mechanism of collection upon which the performance equations
are based (inertial and electrostatic forces) tend toward zero ef-
ficiency in the size range of the bulk of steelmaking dusts «2
microns), where molecular interactions dominate the motion of
particle s.
Any attractive interactions or agglomerative tendency would be beneficial to particle
collection on a clean collecting element, but joining of particles into larger interadhesive
masses would tend to blind a filter matrix (lessening gas handling capacity), or to inter-
rupt electrostatic precipitator field propagation about the wires and plates, and make the
collector surface difficult to clean and the dust hard to handle. This in some cases
necessitates close control of temperature and humidity.
For low-velocity collectors (inherently large and thus economically inefficient for
large -particle collection), a diffusional mechanism can give significantly high collecting
efficiencies. (The effect is greatest, in theory, near zero-micron size, and decreases
with increasing particle size.) A middle ground exists around 0.9 micron in a bag
filter where minimum efficiency can be as low as 10 percent; exactly in the center of
concentration of some 70 percent of steelmaking dust. This is shown in an efficiency-
particle size relationship in Figure VI-33. Reference should also be made to Fig-
ure VI-35 for a fabric filter, and to Figure VI-38 for an electrostatic precipitator, where
this effect also seems to be indicated.
2
I
.,. -
10
'\
10 - ~ /
- /'
.0 = '= - - -
=
.
1
I
.
,
"'''C\.I 1111 "'("0'"
s'
..
:
~
~
FIGURE VI-33.
RELATIONSHIP OF PARTICLE SIZE TO COLLECTION
EFFICIENCY FOR A FABRIC FILTER(78)
Further research is needed to:
(1) Develop technique s for reliable particle - size -distribution data
repre senting the dust as it exists in the effluent gas
(2) Determine the extent of agglomerative effects and their effect
in gas cleaning and effluent sampling mechanisms
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-58
(3) Utilize the diffusion mechanism for small particle s in an
optimum way while retaining economical and efficient inertial
mechanisms for large-particle collection. (If the valley of low
efficiency between the size ranges where diffusion and inertia
are effective cannot be narrowed by this development, then
another tack at development must investigate other gas-solid
interaction phenomena for possible use in gas cleaning.
Particle -interaction effects may be important here. )
(4) Develop economical methods to measure dust concentrations.
(Methods should be adequate for design purposes and well
correlated to methods used for obtaining enforcement data. )
In view of the foregoing difficulties, it is concluded that changes in legally required
efficiency levels (to outlet loadings below about 0.05 grain/ DSCF) would at this time have
to be based on questionable design measurement and theory (whose extension into this
range is also questionable.) The cost of such changes, as indicated by present under-
standing of the mechanisms of collection with proven equipment, would become in-
creasingly great for collection efficiency changes of very small magnitude i. e., changes
which can only be measured with an error of the same order as the change sought.
Technological Factors Affecting Gas-Cleaner Performance
(Adapted From Swindell-Dressler Report Given in Appendix C in
Companion Final Economic Report on Cost Analyses)
The processes in the iron and steel industry can and do depart from design capacity
and design operating conditions for a number of reasons that include the following:
(1) Economic pressures dictate the continued increase in productivity
of an installed furnace.
(2) Technological improvements make possible significant increases
in productivity (such as the introduction of oxygen blowing to open
hearth and electric furnace steelmaking) of a new or existing
facility.
(3) Batch handling of special heats, or runs of varying sizes and
treatments.
(4) Slack market conditions may require cutbacks in output.
Changes in rate of production cause effluent quantities to increase or to diminish
both in gas volume and loading. Operating conditions in the gas-cleaning system can
vary with the se conditions, as well as with the weather, gas -utilization program, raw
material charge, etc. Also, noncontinuous or batch-type metallurgical processes vary
during the course of a heat in both quantity and condition to the effluent.
To maintain satisfactory gas-cleaning performance under these conditions, it is
necessary to have anticipated these factors in designing the pollution-abatement system,
rather than specifying for average conditions. Maximum capacity should be anticipated,
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-59
or adaptation to additional capacity provided. Adjustable equipment can often be used to
optimize performance over a range of operations.
Provision should be made also in the initial installation to meet (or to add and
adapt equipment to meet) expected future requirements of the pollution-control code s
both as to dust content of effluent and treatment of objectional gas and solid chemicals
in the effluent.
Assuming proper design and selection of equipment, any variation or variability in
the process, control equipment, or performance would generally require an added cost.
Any unique feature of a particular gas-cleaning application (particle size, dust loading,
corrosion, etc.) would generally require a departure from a system designed for the
general case.
The following exerpt from the British literature in 1963 summarizes the perform-
ance factors required for effective particulate removal (74):
"The Qean Air Act and the increasingly wide use of oxygen in both the classical and the re-
cently developed top-blown converter processes have combined to create an urgent need for
highly efficient cleaning of high -temperature effluent gases containing submicron iron oxide
fume to the visibility threshold of O. 05 grains per cubic foot. In order to satisfy this need,
manufactUrers of gas cleaning equipment had first to find how collectors which had already
been well proved in other fields could be adapted to applications of which they had had no
previous experience. This entailed not only the establishment of the empirical design pa-
rameters concerned with efficiency, but also a very close consideration of the ability of each
type of collector to cope with unavoidable variations in gas volume, temperature, hu-
midity' solids concentration, etc.
The flexibility of any given type of collector (i. e. its ability to operate efficiently
without breakdown over a wide range of conditions) is much more important in practice
than its theoretical efficiency at constant flow -rate and temperature, etc.. and the best
unit for any given application will often not be the one which a comparison of efficiency
and cost based on idealized operating conditions would indicate.
Every manufacturer who can offer a complete range of equipmept must weigh very,
many factors before finally offering one particular type of collector. He may be handicapped,
particu1ar1yin the case of a completely new installation, by a shortage, of basic process data,
but he can usually arrive at a fairly accurate assessment of the relative strengths and weak-
nesses of the possible units. '
Although the size distribution and shape of dust or fume particles are of course the
factors which determine the fundamental suitability or otherwise of any given design of col-
lector for a particular application, other characteristics of the solids, the carrier gas, and
the process itself must be also carefully considered and their effect on the collection device
evaluated before a final selection is made. '
The agglomerating propensities of the solid particles are important because they de-
termine the size distribution of the particles presented to the collector. The extent to which
agglomeration into clusters or chains of particles will have proceeded, and hence what the
effective particle size will be immediately before the process of final collection is begun,
cannot be accurately predicted, and in practice allowance is made for it in the empirical
design constants used by equipment manufacturers. Agglomeration after collection affects
the caking properties of dry material, making it more easily released from filter fabrics,
less liable to re-entrainment during precipitator rapping, and more easily settled from
liquid effluent.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-60
The electrical resistivity of the material to be collected is of the utmost importance
if a dry precipitator is to be used.
If the collected material is not free-flowing when dry it may create dust handling
problems. Hygroscopic dust will give rise to similar difficulties in 'dry' collectors, unless
humidity and the temperature of solids and gas can be maintained at safe levels by control
of the process, lagging, external heating. warm air purging, or by a combination of these.
For the collection of dusts which are corrosive when wet the obvious choice is a dry
type of unit. unless there is a risk of condensation. If the waste gases contain water vapour
which comes from the process itself, or has been added for cooling or conditioning them,
and sudden temperature surges are likely, elaborate precautions against condensation may be
needed, and a more compact wet unit constructed from corrosion-resistant materials may be
more economical as well as more reliable.
The physical and chemical characteristics of the carrier gas must also be carefully
considered when a collector is being chosen. The effect of variations in gas temperature
and humidity. in particular, must be carefully investigated especially if, as is almost always
the case, they accompany or cause changes in gas volume and dust characteristics during
and after collection. These factors are affected by the method of hooding, cooling, and
volume and temperature control, but no matter how carefully these are engineered the char-
acteristics of the process may still cause the collector to be subjected to conditions which
are far from ideal and impair its operation either directly by affecting the collection pro-
cess, or indirectly by hindering dust discharge or causing structural damage. Collectors of
different types are more or less susceptible to different non -ideal conditions, as shown in
Table 1. The table is only intended to indicate some of the fundamental strengths and weak-
nesses of high -efficiency dedusters in relation to fluctuating operating conditions of one sort
or another. and is not intended to be a comprehensive summary; it does, however, demon-
strate the importance of factors which have nothing to do with the properties of particles. ..
Additionally, from the same literature source(79), the following Table I indicate s
operating conditions which affect the dust-collector efficiency at a peak level of equip-
ment maintenance, and factors which require regular attention (cleaning the collector
surface adequately to match dust loading, temperature control in dry collectors to mini-
mize moisture and heat deterioration and maximize dust removal and handling prop-
ertie s) to insure peak efficiency throughout the life of the equipment.
The effect of operating conditions on the effective performance of the gas-cleaning
function, the effect of those conditions which cause maintenance difficultie s and shorten
service life, and the effect of those conditions peculiar to a particular furnace type of
process on the design and selection of gas cleaning equipment are best judged in the light
of operating and design experience. The following quotations are discussions of emission
cleaning for three iron and steel industry proce s se s which pre sent difficult problems of
equipment selection, performance, and maintainability. These discussions were chosen
for their concise and comprehensive consideration from an application point of view of
the critical factors of equipment use. While they center on British practice (where raw
materials, processes, codes, etc. have some variance from general American practice),
the discussion of each factor remains pertinent (with perhaps some difference in degree)
to a consideration of a corresponding American plant. After considering all the condi-
tions existing on a particular job of equipment application, an engineer may find that the
particular situation with which he is dealing is somewhat more difficult or less difficult
than implied in the following quotations.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-61
TABLE I Effect on collector performance of ftuctuating operating conditions (a)
Dry plate precipitator Fabric filter Scrubber
Irrigated precipitator
TemperatUre
Humidity
Flowrate
Corrosive
solids or gas
Inlet
concentration
Normally up to 650°F
with standard. construc-
tion but momentary
peaks of I OOO°F can
be toJ.erated. Tempera-
ture must be selected
to suit electrical
characteristics of dust.
Normal maximum
temperatUre depends on
"fibre used. Up to say
275°F with organic
synthetics, 600°F with
fibreglass. Higher peaks
tolerable but reduce
bag-life dispropor-
tionately.
Operation below dew-
point leads to bag-
Cleaning troubles.
Chemical and physical
damage to fabric likely.
Dust disposal difficulties.
Normally below 200°F
with presaturation.
Surges can be prevented
if maximum water rate
always used in saturator.
Efficiency unaffected by
changes in humidity,
providing gas remains
near saturation.
Water-rate and/or throat
area must be adjusted to
compensate for changes
in inlet volume.
Alternatively volume
may be kept const:mt
by air-addition.
Special materials of
construction will prevent
corrosion. High-pressure
(high top speed) stainless
steel fan impellers can
give trouble.
Efficiency increased by
operation at lower
flowrates.
Insufficient moisture
may lower efficiency
by increasing dust
resistivity. High
humidity with low
temperature may cause
condensation, possible
corrosion, insulator
and plate cleaning and
dust disposal diffi-
culties. Accurate
control of spray
cooling essential.
Efficiency increased
if flowrate reduced,
although gas distribu-
tion may deteriorate.
Efficiency lime affected
by flowrate. Pressure
drop reduced as volume
falls.
Special materials of
construction eliminate
corrosion but price may
rule out.
Initial design must be
based on peak loading.
Initial design must be
based on peak loading.
Corrosion can be
avoided by accurate
temperature control,
insulation, auxiliary
heating, bypassing, or
corrosion-resistant
materials of construction.
Filter fabric may be
damaged. "
Initial design must be Efficiency not affected
based on peak loading. by increased loadings:
effect on pressure drop
depends on duration of
surges, but can be reduced
by temporary increase
"in Cleaning intensity.
(a) Table I taken from Reference 79.
"SINTER PLANT
MAIN STRAND GASES
The gases withdrawn from the main strand of a sinter machine present a fairly difficult gas
cleaning problem, not, as in most other iron -and steelmaking applications, because high
efficiencies must be achieved on very fine particles, but because of other characteristics of
the dust and the gases themsel ves.
Volumes are great and the use of medium and high pressure drop collectors would in-
volve large nonproductive power consumption.
The waste gases contain large quantities of both sulphur oxides and water vapour.
Consequently they have a high (acid) dewpoint so that condensation and corrosion are a con-
stant danger, aggravated by the wide fluctuations of temperature which occur from time to
time.
The coarser fractions of the dust burden are exceedingly abrasive.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-62
Hence, the ideal dust collector will have the following characteristics:
A pressure drop as low as possible.
Ability to operate efficiently over a wide range of temperatures without ill effect from
occasional dampness of dust and collector internal surfaces.
,
A construction which minimizes condensation, lends itself to reasonably economical
corrosion prevention, and is not susceptible to plugging during the occasional but inevitable
periods of operation below dewpoint.
Freedom from abrasion troubles, preferably by complete avoidance of high velocities,
otherwise by pre -collection of the coarse abrasive dust fractions prior to passing the gases
through any collector in which high velocities are used.
DUST CHARACTERISTICS
The particle size analysis of the dust content of sinter strand gases can vary between quite
wide limits. . . The type of dust to be dealt with depends on the mix fed to the strand, L e.
proportions of home and foreign ores and return fines, and also on whether or not the burden
is conditioned in a pelletizing drum. It must be remembered that changes in dust composi-
tion occur as the rate of sintering alters and the relationship between temperature, flame-
front penetration, and position on the strand varies.
DUS T LOADINGS
The general level of dust concentration is affected greatly by the nature of the material fed
to the machine and can vary from plant to plant between O' 1 and l' 0 grains per normal
cubic foot and may occasionally reach l' 2. The rate of solids emission is very sensitive to
variations in the progress of the sintering process along the length of the strand. Dust is
mainly generated early in the sintering process and again when the flame -front reaches the
bottom of the bed. It has been suggested that in the intermediate zone the increased moist-
ness of the lower part of the bed causes it to act as a crude filter and hence to pass less
dust. It has been found that as complete sintering approaches the discharge end of the
strand, Le. as the mean hottest windbox number increases, the dust loading rises noticeably.
G AS TEMPERA TURE
Gas temperatures usually fluctuate between 60°C and 200°C but lOO-150C is the most com-
mon range and 125 C may be taken as a reasonable average figure for UK practice.
(U.S. practice is in the range of 300 F to 400 F)
GAS COMPOSITION
The only constituents of the easte gases which are important from the gas cleaning point of
view are water vapour and sulphur oxides, both of which affect the frequency and severity of
condensation in the cleaning system. There will usually be about lcp/o water vapour by vol-
ume in the gases and the sulphur oxide content, expressed as S02' may be as high as l' 5
grains per normal cubic foot. Unfortunately, no acid dewpoint figures are available, but
water dewpoints as high as 50°C are encountered and acid dewpoints considerably higher than
this must therefore occur. So far as condensation and corrosion are concerned, the relatively
high proportions of water vapour and oxides of sulphur in the gases complicate the design and
selection of gas cleaning equipment. . . ./but they tend to facilitate electrostatic precipitation.)
CHOICE OF DUST COLLECTOR
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-63
In the authors' opinion the sulphur oxide content of the gases rules out wet methods of col-
lection, since these would result in difficult liquid effluent problems, and saturated gases
ha ving hardly any thermal lift and still containing some sulphur oxides would constitute an
air pollution problem worse in some respects than the original one.
The choice of a dry collector will be dictated by the quantity and size range of the
dust in the case under consideration, the space a vailable, the pressure drop which can be
tolerated and the outlet loading required. Generally speaking, particularly for dusts at the
coarser end of the normal range and if an outlet concentration of 10' 15 grains per normal
cubic foot is the highest acceptable outlet loading, or the dust is finer, or a settling chamber
cannot be accommodated within the space available, cyclones may be used, but their pres-
sure drop (up to 6 inwg) is a disadvantage and they must be speciall y constructed to with-
stand erosion by abrasive dust particles. For a stack loading of less than 0.10 grains per
normal cubic foot a more efficient type of collector must be used.
If. outlet loadings down to O' 05 grains per normal cubic foot are required, the only
suitable device is the electrostatic precipitator. Were it not for the constant danger of
condensation, the fabric filter would be a possibility, but a filter fabric would 'blind' when
operated under moist conditions. It is true that the silicone-treated fibreglass fabric (which
would have to be used in any case to withstand the high maximum temperature) is much
less susceptible to plugging than are the natural and organic synthetic cloths, and has been
found to regain its porosity on drying out, but there would always be a risk of the cloth be-
coming 'starched' with soluble salts and failing prematurely through what can only be de-
scribed as cracking. Fibreglass, which has poor flex resistance in the first place, is excep-
tionally vulnerable to this sort of trouble. This type of collector also compares unfa vour-
ably with the precipitator from the points of view of pressure drop, space requirement, and
maintenance cost, and would not be recommended for main strand gas cleaning.
Although, in common with all other collectors, a precipitator for this application has
to contend with occasional condensation, its operation is not unduly affected by moist condi-
tions' providing precautions are taken against corrosion and providing it has efficient rapping
gear which will clear any. . . build -up and prevent progressive deterioration in its perfor-
mance. The water vapour and sulphur oxides in the waste gases 'condition' the dust and, to-
gether with the relative coarseness of the dust,. . . (assist the precipitation process.)
The Head Wrightson sinter machine installed at the works of the Skinningrove Iron Co. Ltd.,
Saltburn-by-the-Sea, is provided with a Head Wrightson/Research Cottrell dry plate precipitator.
The machine was. designed to process a wide variety of home and foreign ore mixes, and experi-
ence indicated that the dust burden in the waste gases could be reduced by a simple settling
chamber from 1. 0 to 0.3 grains per normal cubic foot. (This is lower than typical American loading.)
The gas valume from the 16 x 6ft square windbox machine is 180000 per normal cubic foot.
The precipitator has two treatment zones, energized by a 15 kVA 230 mA transformer-
rectifier set and operates at a treatment velocity of 6.8 ft/s. In view of the expected
intermittent operation, it was thought advisable to fabricate the collector plates in copper-
bearing 'Corten' steel (O'IDfoc max., 0.1-0.3"7oSi, 0.5-1.oDJoMn, 0.3-0.5DJoCu, 0.5-1.5<'/oCr,
0.1-0.2"70 P) and these have withstood the adverse conditions very well without noticeable
deterioration. The interior of the precipitator shell is protected with gunned aluminous
cement and the whole unit is thermally insulated to minimize condensation. The precipi-
tator. . . was designed to operate at an average temperature of 300°F, and at an efficiency
of 86.7"70, corresponding to an outlet loading of 0.04 grains per normal cubic foot The
design performance has been. . . achieved and, although the sinter plant has worked on a
one or two shift per day basis and the precipitator has undergone an abnormal.LImber of start-
ups, there has been no deterioration of its internals. The sinter fan was inspected in August
1963, 20 months after commissioning and showed no sign of wear other than a general
smoothness over the faces of the blades; it is estimated that it will operate for at least another
3 -4 years without requiring maintenance. The machine had produced 250000 tons up to the
time of the inspection. Reduced fan maintenance and plant downtime are two useful in-
direct benefits of efficient main strand gas cleaning.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI - 64
The dust discharged from the precipitator hoppers is conditioned in a pelletizing drum
and the pellets produced are returned to the process via the return fines conveyor.
DISCHARGE END EXHAUST SYSTEM
The whole of the discharge end of the sinter machine is usually completely enclosed; 100
tons or more of dust per day may be released by the equipment in this area (i. e. the end of
the strand itself, the breaker, hot screen, and discharge to cooler). Air volumes vary with
the size of sinter machine and the completeness of hooding. and are between 30000 and
150000 cubic feet per minute, Gas temperatures are usually between 40° and 150°C. Both
the loading and the size range of the entrained dust are affected by the designs of hoods em-
ployed' and the exhaust volumes allocated to them, but dust burdens are typically in the
range 4-6 grains per normal cubic foot of which 8fJ1/0 might be <100].J m and IfJ1/o <10 ].Jm.
Careful hood design. combined with adjustment of individual exhaust rates during com-
missioning' can reduce both grain loadings and the proportion of coarse abrasive particles
carried in the gases. 1t is relatively easy to obtain collection efficiencies of 90 -95"/0 by
means of simple high -efficiency cyclones, and the stack discharge in such cases will con-
tain about 0.5 grains per normal cubic foot of dust, 9fJ1/0 of which is <10 ].Jm. At.this sort
of grain loading the stack plume does not appear offensive; all the same it represents a very
high rate of solids emission (up to 700 pounds per hour on a large plant), and more and more
interest is being shown in alternative higher-efficiency collection methods.
For cleaning the tip end emission the fabric filter and dry plate electrostatic precipi-
tator are two obvious possibilities. Wet methods can be employed (self-induced spray units
are fairly often used in the USA), but are notto be recommended because they introduce
a secondary (liquid) effluent problem, and are liable to suffer from wet-dry interface troubles
and sometimes from sludge discharge problems. At first sight, the fabric filter would appear
to be ideally suited to this application, providing it is designed so as to a void excessive
scouring of the bags by the abrasive dust, and properly maintained so that a small leak in
one bag cannot' grit -blast' a hole into an adjacent one and start a rapid and messy chain
reaction. The first requirement is quite easily satisfied, but the second is not so straight-
forward and a short period of neglect could have expensive and inconvenient consequences
in the form of extensive bag replacements and operation at reduced capacity. From the
point of view of efficiency and capital cost, the fabric filter is a 'good buy', but running
costs are a most important factor, and cannot be accurately forecast.
While the operating characteristics of the dry plate precipitator are quite predictable
for this application, discharge end precipitation is difficult because of the high resistivity
of the dust at the gas temperatures normally encountered, when the moisture content is less
than about 1. 5 % by volume. In cold, dry weather the water vapour content may be as low
as O. 5"/0 by volume, and under these conditions unstable precipitator conditions are liable
to occur at temperatures around 60°C.
The addition of relatively small quantities of water vapour, sifficient to raise the vol-
ume percentage to 2. 0, leads to a marked improvement in precipitator performance, as
does the addition of 100 ppm of S02. If a guaranteed efficiency. is to be maintained under
every circumstance, and at all times, and if water vapour or S02 cannot be added, the
precipitator will be perhaps three times as large as a unit which will operate satisfactorily
under all but the driest conditions. It is therefore well worthwhile either to mix in gases
from some other part of the sinter system or to add steam. If the problem of conditioning
can be overcome this is a very straightforward precipitator application,. . . (Application of a
dry system at the discharge end when water cooling at the sinter strand is employed could lead
to the reintroduction of moisture control problems.)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-65
QUENQ-l GASES
The quenching of hot fines in pug mill or drum gives rise to large quantities of fine dust,
particularly during periods of erratic plant operation.
In a typical installation the volume of gas vented from the drum was 7600 normal
cubic feet per minute at 40-120.C, containing between fP/o and 24= water vapour by volume.
It was found that the dust loading was greatly affected, not only by the quantity and distribu-
tion of spray water, but also by the quality of sinter being made. During normal operation
of the machine the loading was found to vary between 1. 3 grains per normal cubic foot when
sintering was complete, and 4.7 grains per normal cubic foot when incompletely sintered
material was being discharged from the strand. Shortly after commissioning, before the
sprays had been adjusted and while the operation of the machine was abnormally erratic,
the mean dust concentration had been 4'8 grains per normal cubic foot (corresponding to a
rate of discharge of nearly 400 pounds per hour) and the peak loading 33.2 grains per normal
cubic feet in gas volumes of 8500 -11000 normal cubic feet per minute. This illustrates the
effect of plant operation on stack emissions. The final emission rate averaged 140 pounds
per hour compared to 280 pounds per hour from the main stack and 135 pounds per hour from
the tip end cyclone stack.
The quench stack dust is rather fine (9f11/o <100].J m, 300/0
-------
VI-66
be pressurized with warm air to keep the insulators dry. Dust should preferably only be
stored in the hoppers in an emergency because it tends to bridge, and it may be advisable to
heat the hopper sides. Some condensation is bound to occur at start-up, and it is advisable
to clear as much collected dust as possible from the interior of the precipitator while it is
shut down. If this is not done conveyers and dust discharge val ves may become clogged
with moist dust. If possible the precipitator should only be energized when it has reached its
normal operating temperature, so that little dust is collected in it when it is sweating. If
these. . . precautions are observed the dry plate precipitator will operate continuously. . .
if not, severe build -up, electrical and operating difficulties, and corrosion will be experienced.
A dry plate precipitator installation on a 250 ton tilting OH furnace. . . follows a waste
heat boiler and ID fan, and is designed to clean 78900 cubic feet per minute of furnace
gases at a maximum temperature of 280°C. The design inlet loading is 5 grains per normal
cubic foot during oxygen lancing and the outlet cleanness O. 04 grains per normal cubic foot.
The precipitator is insulated, the hoppers are steam -heated, and the insulator compartments
on top of the unit are pressurized with 600 cubic feet per minute of air at 200°F to prevent
outward leakage of dirty gas and to keep the insulators both dry and clean. The precipitator
has three treatment zones each of which is energized by a 21 kV A, 250 mA transformer
recitfier set. This is quite a good example of a precipitator fitted into a very restricted site,
utilizing turning vanes to reduce inlet and outlet duct sizes without detriment to gas
distribution.
The fabric filter may be used for OH gas cleaning but is more susceptible than the
precipitator to condensation troubles, has a much higher power consumption, and requires
more space. A filter serving one of the Ajax furnaces was reported to operate at a pressure
drop of 8 inwg and to have a bag-life on only 20 weeks. There seems to be no reason why
filters of modern design using improved high -temperature fabrics should not operate satis-
factorily at a pressure drop of 4-5 inwg with a baglife of a year or more, but prolonged
pilot -plant testing would be needed to prove the durability of the filter fabric.
Both the irrigated electrostatic precipitator and the high -energy scrubber are capable
of cleaning OH fume to O. 05 grains per cubic foot or better. but they would ha ve to be
constructed from expensive corrosion -resistant materials and would create secondary prob-
1ems of liquid effluent treatment and loss of stack gas buoyancy.
ARC FURNACES
FURNACE PRESSURE CONTROL
For consistently good fume control at minimum rates of extraction, automatic control of
furnace pressure is essential. The indicated pressure which it is necessary to hold within the
furnace depends on the position of the pressure pick -up. The accuracy of control required
is of the order of :1:1. 0 nwg for furnaces melting OH grades of steel but may be as fine as
:1:0.03 inwg for a furnace producing alloy steels. The control system used must have a
high speed of response if it is to cope with sudden fluctuations within the furnace.
GAS COOLING OR CONDITIONING
I-
I
Temperature at the outlet of the combustion chamber may be upwards of 1000 C and
the gases must be cooled before they can be cleaned. The methods available are air dilu-
tion' indirect cooling by heat exchanger, and evaporative cooling. It is considered that
the latter is often the best compromise on the grounds of simplicity, final gas volume,
space requirements, and initial cost.
However, the type of collection device used will often dictate the manner in which
cooling is carried out. With wet methods of collection, a comparatively'small spray tower
may be used (without fine control of the cooling sprays) and air dilution or indirect cooling
would be pointless. If dry precipitation is preferred, the gases must be conditioned (most
simply with water) and if a spray conditioning tower is required for this reason the gas will
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-67
be spray cooled to the desired precipitator operating temperature. The fabric filter does not
require pre-humidification of the gases for efficient operation and is, moreover, exception-
ally vulnerable to condensation. The preferred method of cooling in this case will depend
upon whether the filter fabric is organic-synthetic (e. g. Orlon or Tery1ene) and therefore
not suitable for operation at over 130°C, or fibre glass, which will withstand up to 250°C.
In the former case air dilution or indirect cooling may be used, but in the latter spray cool-
ing should present no difficulties providing a good control system is fitted.
SYSTEM CAPACITY AND SAFETY
. . . The details of safety require that. . . very conservative assumptions are made. The
problem of explosion hazards has been considered in recent papers.
Air may enter the system at the air break between elbow and fixed fume pipe and at
the combustion chamber, as well as through the furnace openings. The volume of air
entering by each of these routes is unimportant providing (a) that control of fumes is ob-
tained and (b) that the final waste gas volume is such that even if combustion has been in-
complete an explosive mixture cannot be formed.
The combined effects of combustion and dilution have been calculated for the lancing
period, and are shown in Table III. However, the rate of evolution of combustion, follow-
ing the addition of oily scrap cannot be predicted, and it must be remembered that in
practice the operation of a fume cleaning system must take second place to the production
of steel; allowance must also be made for occasional deficiencies in the standard of both
operation and maintenance of cleaning systems. Hence, although under ideal conditions
an 02 to waste gas ratio of 10: 1 would no doubt be adequate, it is recommended that a
ratio of not less than 15: 1 be used.
Ratio of waste
gas/oxygen
injection flowrate
TABLE III Effects of combustion and cWutionn
Approx. %
combustion for safe
operation (based on
100% oxygen
utilization)
Nil
Nil
5%
10%
32%
50%
55%
% Carbon
monoxide if no
combustion occurs
22: 1
16:1
15:1
14: 1
10: 1
6: 1
5: 1
9.1
12'5
13,3
14.3
20.0
33,3
40,0
Current understanding of the explosion problem is incomplete; explosions have been
reported even in conservatively designed systems following errors in operation, and it is
considered more prudent to use theory to predict the magnitude of apparent safety margins
rather than to reduce these to the point where (due to the intrusion of incalculable factors)
they do not exist, and a variation in the process or a mistake by an operator can cause
an explosion.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-68
GAS CLEANING
The furnace gases may be cleaned to 0.05 grains per cubic foot by precipitator
(wet or dry), fabric filter, high -energy scrubber, or combination scrubber -precipitator.
Dry plate precipitation is relatively straightforward providing the gases are properly
conditioned. It is therefore ideally suited to direct extraction systems but much less so for
hood or conventional hood vent installations.
(A) 75 ton furnace. . . has been fitted with direct extraction fume control equipment
and fume is to be collected by a dry -plate electrostatic precipitator ('B' unit referred to
below). The lancing rate of this furnace is 1200 cubic feet per minute and the volume
during lancing. after combustion and cooling 49200 cubic feet per minute. The precipitator
is designed to clean a total of 83200 cubic feet per minute from the existing furnace and
another which is to be added in the future, from 6.5 to 0.05 grains per normal cubic foot.
Furnace gases will pass through a water-cooled elbow and refractory-lined fixed duct
connected by a power-operated movable sliding sleeve. into a gas burner followed by a
combustion chamber. They will be cooled and conditioned in the rectangular spray
tower and will enter the precipitator at a temperature of 500'F.
The fabric filter is theoretically ideal. having a uniformly high efficiency irrespective
of throughput but it must be carefully designed and protected against condensation. Filtering
velocities may be as low as two feet per minute so that space limitations will often exclude
this type of cleaner.
The high-energy scrubber operating at a pressure drop of 30 inwg or more will do a satis-
factory fume-cleaning job (U.S. air pollution regulations would require about 4S inches water gage.)
and its compactness is a great advantage, particularly when the available space is limited.
Power may be saved by regulating the fan in an efficient manner to suit the rate of exhaust
required for fume control and the pressure drop needed at different periods of the melt to
give the statutory final gas cleanliness. but this is only practicable if the pressure drop of
the scrubber can be adjusted to the desired level over a wide range of flow-rates.
It must be stre ssed that in the long run regular maintenance and attention to op-
erating conditions affect the cost and effectivene ss of any gas-cleaning unit. The in-
corporation of automatic controls, operator-proof controls, scheduled preventive main-
tenance, anticipation of adverse process conditions and raw material possibilities are
important to the continued performance of gas-cleaning equipment after the guarantee
period.
Attention is now directed more specifically to the following parameters that affect
gas cleaner performance:
(1) Effect of gas-volume changes on collection efficiency of a dust
collector
(2) Effect of pressure drop within the gas cleaner on efficiency and
capacity of the collector
(3) Effect of dust loading, and effect of collector-surface renewal
on pressure drop, volume, and collecting efficiency
(4)
Effect of particulate as generated in each metallurgical proce s s
(particle density, particle density, particle size, size distribution)
on efficiency of each applicable dust-removal device
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-69
(5) Effect of temperature on efficiency of and gas volume to
collector, and required gas conditioning for cooling and
humidification before dust removal
(a) Gas analysis as it affects conditioning required
prior to cleaning and exhausting
(b) Corrosion and the use of water
(c) Abrasion and chemical effects of dust
(6) Adaptability of the particulate-removal system to removal
of gaseous pollutants.
""~fect of Gas Volume Changes
As previously indicated, the volume of effluent gas emitted by a metallurgical
proce ss may vary greatly during one heat, or according to changing production level of
the proce s s. Because the efficiency of dust removal change s when volume change s, it
becomes necessary to
-- -operate at constant volume with air substituted for effluent gas
deficiency,
---or, use a gas-cleaning device which adjusts itself to volume
changes, or is adjustable to satisfactory efficiency over a range
of volume.
Self-induced or orifice washers and certain fluidized-bed scrubbers can adjust
themselve s, essentially at constant efficiency. Adju stable -throat venturi s, orifice-
wedge and flooded disk scrubbers can be adjusted to suit a range of gas flow. The se and
other wet scrubbers can also be flooded (uneconomically) to achieve the same effect.
Multiple units (nested cyclones, parallel scrubbers, precipitator tubes or ducts,
multiple venturis, baghouse filter tubes) can be partially blocked off to maintain high
(design) efficiency at reduced volume, with economy of water and power use.
-- -or, de sign for maximum possible effluent volume, and lIover-cleanll
at reduced volumes.
The following quotation from the British literature in 1964 illustrates one sug-
gested method for determining good design in relation to gas flow(l3):
"It will be appreciated that the efficiency of a precipitator is greatest when the velocity of the gas
through the cross-section of the electrode system is uniform, and no gas is bypassing the electrode system.
This is ensured by the construction of . .. models... of the precipitator and inlet flue system. The flow
conditions in the model are adjusted to give the same Reynolds number as the full-scale plant, allowance
being made for scale factors, gas viscosity. and density. The flow pattern in the model is corrected using
splitters and baffles, their position being determined by experiment. Such model tests permit requirements
to be worked out in advance, and a void the difficulties in carrying out such work on site on the finished
plant. "
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-70
Effect of Pressure Drop
Because of large cross section and control of build-up conditions (temperature and
humidity) and regular rapping for dust removal, electrostatic precipitators usually will
show negligible change in resistance to flow in operation.
Bag filters, when new, have low resistance and low efficiency. Sometimes a pre-
coat of dust is applied to make the initial cleaning of process fume more effective. The
buildup of dust increase s both efficiency and pre ssure drop until the cleaning (by shaking
or reverse flow of air) cycle is initiated (often by a pressure signal). Then efficiency
will be at a lower (but still effective) level until the dust layer reforms on the fabric. (55)
Wet scrubbers increase in efficiency with increased resistance due to mechanical
constriction of the throat area or added water input. The proportionality of change as
attributed to Semraufs correlation is described later.
Efficiency of cyclones also depends upon pressure drop. Cyclones are used only
with coarse, easily collected dusts, however, and usually with a view to product re-
covery as much as to gas cleaning. As such, they may usually be regarded as process
equipment. The rules relating pressure drop, capacity, and efficiency are available in
the Air Pollution Engineering Manual. (80)
Effect of Dust Loading
An electrostatic gas cleaner is essentially a constant-efficiency device, so that
any change in inlet loading will be reflected proportionally in the outlet stream loading.
However, in actuality, changes in dust build-up occur and adversly affect the propaga-
tion of a uniform electric field. Plate spacing must be designed to accommodate the con-
dition of heavie st expected dust loading. Automatic controls are often required to main-
tain an optimum electric field without spark-over.
A well-designed bag filter will be unaffected by a change in inlet loading, except
that automatic cycling of the bag-cleaning system will adjust to the change (within cer-
tain limits of variability).
A wet scrubber will yield constant efficiency for a given pressure drop. There-
fore, a change in inlet loading will be reflected proportionately in the outlet loading.
However, wet-scrubbing systems can be adaptable to changing conditions, provided suf-
ficient power is applied. A venturi throat can be closed to maintain a given effluent
level with increased dust generation in the process. A process whose fume output varies
widely with time could be handled by making frequent adjustments of the cleaner to
maintain a constant acceptable output of fume.
Particulate Characteristics From Different
Proce s s Se gments
In Appendix C, "Characteristics of Emis sions", of this report, some data are
presented on the nature of particulate material generated by various processes and
conveyed by gases emitted from the process vicinity. This dust is generally nonuniform
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI -71
from one particle to another and from process to process. The differences may be
categorized as particle size, shape, density, and composition.
The mechanisms of particle collection on which gas-cleaning equipment are based
vary in collecting efficiency, generally with particle physical properties. The chemical
nature of the dust may affect its susceptibility to electric charging. This would pri-
marily affect electrostatic precipitation; but could be a second-order effect in wet
scrubbing and in fabric filtration. Solubility and chemical activity in water would affect
the water cycling, dust handling, and collector-surface maintenance in wet collectors.
Some general variations in the efficiency of collectors with these particle properties
can be drawn. Grade -efficiency curve s typical of industrial collectors in the mid-1950' s
have been presented for various types of dust-collecting equipment. (78) These show
the efficiency of collecting particle s of a given size. The curve s were based on te st
results using a standard dust (Table VI-13) with a 2.7 specific gravity. These curves
can be used to indicate in a general way the relative applicability of each type of equip-
ment to different process fumes. As shown in Figure VI-33, the efficiencies generally
are lower (often dropping abruptly) for finer grade s of dust. Some device s are more
economical to operate, but they generally do not clean fine particles from gases as
well as others.
TABLE VI-13.
GRADING OF W. C. 3 TEST DUST(a)
Size
of Grade,
microns
Pe rcentage by
Weight in
Grade
Percentage by
Weight Smaller
Than Top Size
of Grade
104-150
75-104
60-75
40-60
30-40
20-30
15-20
10-15
7-1/2-10
5-7-1/2
2-1/2-5
2-1/2
3
7
10
15
10
10
7
8
4
6
8
21
100
97
90
80
65
55
45
38
30
26
20
12
(a) From Reference 78.
Relationship of Particulate as Generated by
Different Proce sse s to Collecting Efficiency
Stairmand(78) has presented grade efficiency curves for various cleaning equip-
ment at specific conditions using a standard dust. These curves indicate the relative
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
1---
VI-72
applicability of each type of equipment to different dusts. Some devices are more eco-
nomical to operate but do not clean fine particulate from gases as well as others.
The curves are applied to various processes with their given particulate size
di stributions.
By making a density correction, the curves can be applied to dusts for which
particle-size distribution data (as in Table VI-13) are known. Some distribution data
are given in Appendix C to this report. No quantitative data are available upon which to
base corrections for particle shape, composition, and surface differences; so such an
application of the curves will not be quantitatively precise.
Particle - size -distribution data available for thi s kind of analysis are inadequate in
some measure. The size ranges reported are usually too large and require excessive
averaging in the region of greatest variation in efficiency on the grade-efficiency curve
(the fine-particle-size region). Steelmaking dust is largely concentrated in this region.
Whether or not averaging according to the log-probability distribution would be applicable
to distributions having given data ranges such as 0-1 micron or 0-5 microns is not
known.
The shape of the grade-efficiency curve may be affected somewhat by (1) process
variables which alter the properties of the dust, (2) conditioning of the dust by humidity
and tempe rature control (in the case of electrostatic precipitation), (3) collector
geometry (affecting treatment time) and (4) energy input (see Table I). The develop-
ment of a family of curves should be undertaken to develop grade -efficiency variations
at different levels of pertinent operating variables (such as scrubbing-energy level or
electrostatic-precipitation treatment duration). Note that steelmaking fume can gener-
ally be adequately removed with a scrubber pressure drop of 40 plus inches of water,
or by electrostatic precipitators whose geometry, control, and energization are spec-
ifically selected for that application. Using the given curves, however, to analyze the
effect of each type of treatment on actual proce ss dusts will give a rough comparison of
collectors for a given task, and a general comparison of dust-property effects on per-
formance of each collector. At least, relative indications may be drawn. Efficiencies
measured in the field of equipment that is collecting dust from the actual processes
dusts whose properties would also be tested under the collecting conditions) would allow
precise comparisons, more precise (and probably more economical) designing, and
dependable predictions of performance. Such data generally are not available, not very
good, or undisclosed. Therefore, the theoretical treatment while limited, is the be st
alternative available for relative comparisons.
The fabric-filter curve has been deleted from Figure VI-33 because it is based on
a theoretical calculation for a new filter. A more typical practical grade -efficiency
curve given in Figure VI-34(82) is based on actual test results. However, both indicate
that the efficiency would not go to zero as particle size approaches zero. This is dis-
cussed in the literature(67) in terms of a diffusional collecting mechanism which comes
into play at an increasing rate as particle size diminishes. The zero drop-off in the
other grade-efficiency curves represents the failure of the inertial impaction mechanism
to collect small particles. Discussions in the literature(67) include diffusion of particles
to water-droplet targots as well as to filter media. Therefore, this effect should also
apply to wet scrubbing. As indicated by the U-shape of curves in Figure VI-38, the
mechanism of diffusion seems also to apply to electrostatic precipitation. This mech-
anism, then, would suggest an upward alteration of the grade-efficiency curves.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
1---
SlOO
U
i 10
t
: .0
~
. .0
o
.- '-
/'
/'
/
/
I
I
I
II
I
;
.. .0
:s
II
u
10 40 .0 10
'.UTIC",. 1111. N.C:AO...I.
.... II
Medium ~lfidcnC)'. 1b"'.d':rouehp'.IIC')'tlont.
Erficiency II , rn~ .. l1~.
.100
=
..
r 10
..
./ - - - ,..-
I , I
I ,/
I
,
I - U"II'-AUO
----- DRy ,- -
I
, I I
.
..
~.o
~
.00
.
~
... .0
a
..
10 10 JO 40
,....""C\,.I '.'U.,W'CAONI.
F.. 17
urcr Jiametu dry .and iuigucd c}'I.:Iont.
Edic~y II , nhnunl .. '7'/.
Fabric filter
curveomi tte d
for reasons
given in the
text.
SIOO
..
r 10
o
,.....- r
I
I
I
,.....-
..
:60
u
~
.00
~
::; J 0
:s
~
. '0
'."TlCLI
IS 20
1111. "'CAO"'.
Fit- =
Sdl.induced \f'tIY cullector.
I!tkiency .1 , mn:ron. - 9 J ~~
VI-73
~IOO
~
r .0
.
..
~ .0
u
~
.'0
.
~
.. '0
:s
II
u
I
I
I
I
",,100
:,
...
% 10
.
u
~ 10
U
~
.00
o
---
I
I
I
o
100
ao 40 60 10
MATICL. 1'11. WIGAONI.
Y.. 15
Hilh efficiency (Ion, cone) cyclone.
Efticicrw:y II , l!'i\,"fOfU - 7) '/,
.
g
u '0
:s
~
lOG,
10 20 )()
ItA"TlC:U 1111. WICRONS.
Fit- IS
Small di.meter. rubulu c)'clones.
ElRcicncy u.' ",i.:ronJ - &91.
'0
10
,..'00
=
u
~ 10
~
..
u
~ 60
o
I --
V
/
/
I I
I
I
,..100
=
u
r 10
/' - - - - 'r- - - -
-
,
,
,
I
- UIAICAaO
__n 0" I
I
I I
u
~
. '0
.
g
u '0
:s
g
10
10 20 JO 40
PARTICLI SIU. W'CRONS.
Fla. 18
Low prcssure drop cdlul" .:ydone..
Efficiency al , :ni.."TOfts - 042"l.
.
...
: 60
U
~
.00
.
g
u 10
~
II
u
o
10
5 10 II JO
MATltLI IUI!. I ~ ICR.?t4 s.
Fit- II
EI~tic preripiutOt. IrriS3Itd-(ffid-:n\:y ae " mkrons .. 9a~/.
Dry-efficienC)' al ~ microes .. 9: ~;.
~ 100
~ I
~ 80
.
u
~'O
~
.'0
.
2
~
~ 30
~
0
I 10 I'
"'.TleLI Sill, NI(:-ONI.
""21
Spray lOWer,
Efficiency II , micron. .. 9<'%
10
..
~ 100
=
U
~ 10 I
.
u
~ .0
U
~
.00
.
2
~
uJO
~
0
.s
I I' 1
'.A~TICLI 'III, ,..I(IIOH,.
Fit- u
"'" inpi"l..:mem scrubber.
Elftciel'Cy JI 1 mkroos - 97 :',
.0
..100
=
u
r 80
- -
.'
~'OO
u
i 10
1/
I
f
I
I
.
..
~ 60
~
woo
.
g
u '0
:s
II
u
o
I 4 I . , .
'"""TlCLI 11'11. !tit r CROM I.
. F.:U
Vanu" tc:nJbbcr.
EfficiaKy 81 , mil,.Tons ... ~,9,6'1.
.
u ,
~.o
U
~
"'0
.
g
u '0
~
II
u
o
10
.1 " 50 . 7 .
P." TlCLI SUI. NICRONS.
. FIa- ZS
Di.intecrltOt ISS washer.
Effiocncy at " mi.:rons .. 981.
'0
..
FIGURE VI-34.
GRADE-EFFICIENCY CURVES FOR DIFFERENT TYPES OF
AIR-POLLUTION CONTROL EQUIPMENT(78)
(These curves were determined on the relatively coarse
dust described in the text, and do not necessarily apply
directly to metallurgical fume. )
BATTELLE
MEMORIAL
INSTITUTE - COLUMBUS
LABORATORIES
-------
VI-74
Variations in this effect will occur with temperature and particle concentration. By
designing low-flow-rate collectors, and by optimizing inlet conditions, one could take
advantage of thi s me chani sm with fine du sts.
99.99
After 10 shakes
c:
.
u
c:
-------
VI 75
TABLE VI-14. CALCULATED RELATIVE EFFICIENCY OF COLLECTING EQUIPMENT FOR VARIOUS PROCESS
DUSTS AND COLLECTORS
(Results reported here are subject to the limitations given in the text. )
Flux Fraction
(as in self-
Sinter fluxing sinter Basic Oxygen Electric Arc Pressure Drop
Strand making) Furnace Open Hearth Furnace (in. water)
Particle Specific Gravity 4.0 2.7 5.0 5.2 . 3.93
grains 4 4-8 2-7 3-6
Inlet Loading, SCFD
Required Efficiency to Attain 98.73 98.9-99.38 97.5-99.28 98.33-99.17
0.05 Grain per Standard
Cubic Foot, U/O
Computed Efficiency, U/o, for:
Cyclones
High Throughput 65 59 Not Applicable 3.7
High Efficiency 91 90 4.9
Multicyc10ne 98.5 98 4.3
Wet 97 96 3.9
Wet Scrubber
Low Energy
Spray 97 97 1.4
Wet Impingement 99.75 99.52 6.1
Self-Induced 98.25 98 6.1
High Energy
Disintegrator 99.32 98.95 72 88 86
Venturi 99.98 99.95 85(a) 94.5(a) 94(a) 22(a)
Electrostatic Precipitator
Dry 99.00 95.5 64(b) 83(b) 81(c) 0.6
Wet 99.86 98.95 86(b) 0.6
Fabric Filter 99.99 99.99 (97.8) 99.5 99.5 4.0
(a) A pressure drop of 40+ inches of water gage normally is required to clean steelmaking fume to the specified 0.05
grain per standard cubic foot with a venturi scrubber.
(b) The codes in the mid-1950's (when these grade-efficiency data were typical in Britain) were less restrictive.
Precipitator designs today can be guaranteed to 99.5 percent efficiency if required, but the curves in
Figure VI-34 still can be used for comparative collection efficiencies.
(c) With proper humidification.
(d) Underlined values for computed efficiencies meet the required efficiency for outlet bonding of 0.05 grain per
standard cubic foot.
BAT.TELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-76
(1) The fan-power requirement is proportional to volume because
of this, many systems cool the gases to a minimum practical
temperature before entry into the fan, unless thermal lift
must be maintained to get rid of noxious gases in the effluent.
Since positive pressure ordinarily makes for simpler struc-
ture in a precipitator, and be st efficiency requires several
hundred degrees of temperature, this benefit is ordinarily
not available for a forced-draft precipitator fan. However,
in the wet scrubber, where efficiency depends on a high gas-
stream energy, at a high power-consumption level, the cooling
of gase s is particularly beneficial.
(2) Precipitators and especially bag filters are limited in the
temperature at which they operate. Thus, cooling as a
preconditioning step is required before entry to the gas
cleaner on many processes. The method of cooling also
affects the volume of gas to be handled, as shown in
Figures VI-36(l3) and VI-37(82).
1000 --.
900
800
~ 700
~: 600
~ ~oo
~~ 400
~v"
lool'
100
-L
o reo 100
----.--.-----.---
LU
a:
2 S
«
ITi
"-
~
<./),
« ,..~
t~ L-J 4
0'
CJ 5
'Z 0
o t)?
~
15 1-'
- «
h 0 3
) ~ 1.-:
1.-!.. <~
(, LJ..J
. CI,
ILJ t.:.
..,- L
J => 2
G V!
:>
-------
VI - 77
Effect of Humidity
Additions of water to the gas stream or to the gas-cleaner system is a frequently
required preconditioning step., The humidity of the gas entering a baghouse must be
sufficiently below the dewpoint to preclude corrosion of the structure and clogging of the
bags with moist cake. On the other hand, quenching is an economical way to cool,
avoiding expensive radiation ductingor excessively large components to handle dilution
air.
Humidity is sometime s critical in the electrostatic precipitation of certain dusts
of high resistivity. Again, it can be coupled with cooling quench in the pre-conditioning
zone, but care must be taken to stay above the dew point temperature. No liquid effluent
results from these cases. It should be noted that in both of the above systems, the acid
dew point is also critical from a corrosion point of view in those cases where sulfur
dioxide is a significant process effluent component, such as the open hearth and some
coal-burning processes. While it has been noted that sulfur dioxide can be beneficial in
the precipitation of some proce s s dusts, this benefit is likely to be lost as sulfur dioxide
regulations take effect. ' .
Water flushing of elbows, fan blades, and other parts of the systems has been ef-
fectively used to inhibit impingement abrasion and to prevent dust buildup.
Corrosion, abrasion, dust buildup and excessive temperature are the most frequent
maintenance problems on a gas-cleaning system. Where water is used as a remedy,
careful pH control is important, as is control of solids buildup in a recirculating-water
system.
Except for this preventive maIntenance, water effluents usually are the result of
wet scrubbing or wetted-surface precipitation. This later finds application to electric-
arc-furnace fume-cleaning where satisfactory particle resistivity for collection is dif-
ficult to achieve, and in sinter plant and blast furnace applications where coarse,
abrasive dust must be removed.
""'ffect of Electrical Resistivity of Dust
The effect of electrical resistivity on the collection of dust is discussed as follows
in a British source published in 1964(13):
Much has been written on the subject of the effect of the resistivity of the dust on precipi-
tator efficiency and it is not proposed to go deeply into this aspect of the subject;. .. It can be
shown, however, that for dust of a very high resistivity, precipitation efficiency can be. seriously
reduced (see Figure VI -38 and section on particle sizing), and, for resistivities higher than 1011
ohm-cm, difficulties are likely to be encountered. There is some divergence of opinion between
different investigators on the value of resistivity at which difficulty is likely to be encountered
and it is thought that this is due to a number of factors difficult to control, such as the degree
of packing of the dust, so that in practice different forms of apparatus can disagree to a con-
siderable extent. At the same time resistivities measured by anyone form of apparatus, when
used by an experienced qperator, can be related to precipitator performance.
The electrical resistivity of most dusts and fume depends on the nature and condition of
the surface of the dust particles, rather than on the material of which the dust is composed; the
resistivity is in practice often determined by adsorbed layers of vapour, such as water, sulphuric
acid, or ammonia. These usually arise from reactions taking place in the furnace or vessel to
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-78
which the precipitator is attached; for instance, high-sulphur fuel oil used in firing OH furnaces
can produce sulphuric acid, and this in turn is adsorbed by the dust. Where the dust resistivity
is high, suitable layers to reduce the resistivity of the dust can be provided by the injection of
one of the conditioning agents listed above into the flue before the precipitatOrs. In practice,
however, in this country, it has not so far been found necessary to supply any artificial condi-
tioning agent to red oxide dust plant, although difficulties have been reported from abroad.
Figure VI-39 shows the resistivity plotted against temperature for fume originating from LD
converters, OH furnaces, arc furnaces, and ladle desiliconization processes.
It will be seen that the resistivity is below 1011 ohm-cm in all cases except for fume
from the arc furnace. In the case of OH furnaces the dust is normally 'conditioned' by the
water vapour and sulphur trioxide resulting from the combustion of the fuel used to fire the
furnace.
In the case of the arc furnace, there is normally no such supply of conditioning agent in
the gases leaving the furnace as curve I, which is the resistivity of a dust sample taken
immediately at the furnace outlet and is typical of a highly resistive dust without the condi-
tioning surface layer. When a precipitator is attached to an arc furnace it is necessary, in
view of the high temperatures involved, to cool the gases, usually by means of a water spray
tower, to an economical level for the precipitator. This has the effect of reducing the gas
volume to be treated; and at the same time, the dust is 'conditioned' by water vapour and the
resistivity curve assumes the shape shown for the other fume with the peak value below the
limit for efficient precipitation.
Effect of Particle Size on Precipitator Efficiency
The same British report(13) describes the following:
It can be shown from calculations on the forces acting on charged particles that the
efficiency of an electrostatic precipitatOr should decrease with decreasing particle size; this,
however, is not normally borne out in practice and many commercial applications of precipi-
tation are on processes in which much of the fume is submicron, as for instance, blast-furnace
gas cleaning and the red oxide fume evolved from oxygen blowing processes.
Figure VI-37 shows the relationship with precipitation efficiency and particle size for a
number of dusts, the first three relating to dry precipitation, and number 4 to a wet precipi-
tator. Curve No.1 was obtained on a dust whose electrical resistivity was of the order of
1013 ohm -em, and the precipitator was exhibiting signs of severe reverse ionization; it is
interesting to note that the fall-off in efficiency becomes increasingly serious for particle
sizings below 20 ].1m. Curve No.2 was obtained upon the same dust when the resistivity had
been decreased by the use of a conditioning agent, while curve No.3 was obtained on a dust
whose resistivity was well below the limit of 1011 ohm -cm quoted in the section on
resistivity.
Curve No.4, obtained on the wet electrofilter, has a fall-off comparable with the lowest
resistivity dry dust (curve No.3); in this case, although the dust was initially highly resistive, the
resistivity of the dust in the precipitator was reduced to a safe level by the cooling and natural
conditioning effect of the spray rower preceding the precipitator; in addition, since the particles
were deposited on a moving flow of water, the effect of high resistivity would be of no conse-
quence in any case. Since this fume consisted of non-magnetic particles of spherical form it
was possible, using the electron microscope, to continue the grading beyond the limit of most
forms of grading apparatus; these gradings indicated that tpere was no serious fall-off of precipi-
tator efficiency for particle sizings down to the order of 100 ].1m,
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-79
While the theory of the motion of the dust particles under the effect of the electric field
assumes that the dust is deposited as individual particles, there is in practice a strong tendency
for very fine fume to agglomerate into masses consisting of hundreds of fine particles, such
agglomerates behaving as single, much larger particles in the electric field, with the result that
efficiency is higher than would be theoretically calculated for such a fume. This is normally
considered to be one of the explanations why the precipitator fails to obey the basic theory. It
is also one of the difficulties of carrying out dust gradings and limits their value, since clearly
what is required of a dust grading apparatUs is the grading including the effect of agglomeration,
and one of the debatable points in dust grading methods is the energy which should be used to
disperse the agglomerates formed in the precipitators in a dust grading apparatUs. An interesting
featUre is the action of the conditioning agent, as illustrated by curves 1 and 2, since it would
appear that, in addition to reducing the resistivity of the dust, the agglomerating properties
are also materially improved. The authors consider that for efficient precipitation it is neces-
sary, particularly in a dry precipitator, for the dust to have the correct agglomerating properties
in addition to a suitable electrical resistivity value.
Adaptability of Particulate Removal Systems
to Removal of Gaseous Pollutants
Studies on the injection of dry, powdered limestone, dolomite, maganese dioxide,
alumina, and other metal oxides to process gases containing sulfur dioxide indicate that
some 30 to 60 percent of the SOZ can be absorbed by the additive and removed in the
particulate-removal system (as an added inlet loading).
A bag filter could do this effectively. A wet scrubber system gives the added
benefit of a liquid-:absorbtion stage, and has yielded good test re suIts. Alkaline solutions
may be used without the powder injection.
A catalytic oxidation process unit could be inserted in series following a precipi-
tator so the high temperature at which the precipitator operates could be used in the
oxidation. The process scrubbing would preclude following with a baghouse or
precipitator.
These systems are in the development phase, and their use is contingent on eco-
nomics and competitive-process developments.
BATTELLE MEMORIAL INST.ITUTE - COLUMBUS LABORATORIES
-------
FIGURE VI-38.
VI - 8 0
-.---"."..--...
I Arc furnace
1 lD conyett~r .
~ Open. hearth furnace,
AU11toilo
. Open-heonh furnace
~ Destllcon1zotlon
ladle proce n
o Open.heorth furnoce
90
,
l
o
~
>'
U
~
U
~
z
o
~
a:
U
~
8l
I
JL'
2 "
Hiqhly re.isti-.e dusl Od3ohm-cm)
qivinq serious reverse ionization
>-
~
;;
:;; .
~ to
'"
~
2 DJst as I after use of conditioninq
oqeni to reduce re.istivity
Dry type precipitator
3 Normal well conditioned dust
~
" Hiqhly re.isti-.e dust usinq wet
type precipitolor
10'
.1. J ....-!-.. _.J_~,. .1..--:- I -_.'--_J --- L
8 ~ ~ ~ ~ ~ M n ~
PARTICLE SIZE . Jim
J
b
100
200 JOO 400
TEMPERATURE. .C
100
600
VARIA TION OF PRECIPITATION EFFICIENCY
WITH PARTICLE SIZE
FIGURE VI -39.
ELECTRICAL RESISTIVITY OF RED OXIDE
FUME FROM VARIOUS OXYGEN-BLOWN
STEELMAKING PROCESSES
..BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-81
REFERENCES FOR SECTION VI
(1) Kudlich, R., (Revised by Burdick, L. R.), "Ringlemann Smoke Chart", U. S.
Bureau of Mines, Information Circular 7718 (1955).
(2) I'Maximilien Ringlemann - Man of Mystery", Air Repair, ~ (2), 4-6 (November
1952).
(3) McShane, W. P., and Bubba, E, "Automatic BOF Stack Monitoring", 33/The Mag-
azine of Metals Processing, ~ (5), 97-104 (March 1968).
(4) Communication to John Varga, Jr., Battelle Memorial Institute, Columbus Labora-
tories, December 20, 1968.
(5) Silverman, L., "Predicting Performance of Collector s in Air Pollution Control",
Journal of the Air Pollution Control Association, ]2 (12), 573 (December 1963).
(6) Communication from R. B. Engdahl to John Varga, Jr., Battelle Memorial Institute,
Columbus Laboratories, February 3, 1969.
(7) 11 Dust Collectors: A Look at Wet vs Dry Systems", 33/ The Magazine of Metals
Producing, 4 (6), 57-64 (June 1966).
(8) Vajda, S., "Blue Ribbon Steel With Blue Skies", Iron and Steel Engineer, 45 (8),
71-75 (August 1968).
(9) Langer, G., "Ice Nuclei Generated by Steel Mill Activity", Proceedings of the First
National Conference on Weather Modification", 220-227, April 28 - May 1, 1968,
Albany, New York.
(10) Gottschlich, C. F., "Removal of Particulate Matter from Gaseous Wastes-
Electrostatic Precipitators", American Petroleum Institute, New York, 42 pp.
(1961). ------
(11) White, H. J., "Industrial Electrostatic Precipitation", Addison-Wesley Publishing
Company, Inc., Reading, Mass. 1963.
(12) Stern, A. C., Editor, "Air Pollution" Volume III, Academic Press, New York,
437-456 (1968).
(13) Watkins, E. R., and Darby, K., "The Application of Electrostatic Precipitation to
the Control of Fume in the Steel Industry", Special Report 83, Fume Arrestment,
The Iron and Steel Institute, 24-35 (1964).
(14) Sproull, W. T., and Nakada,
Moisture and Temperature",
1350-1358 (June 1961).
Y., "Operation of Cottrell Precipitators-Effects of
Industrial and Engineering Chemistry, 43 (6),
(15) Lagarias, J. S., "Predicting Performance of Electrostatic Precipitators", Journal
of the Air Pollution Control Association, ]2 (12), 595-599 (December 1963).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-82
(16) Frame, C. P., and Elson, R. J., "The Effects of Mechanical Equipment on Con-
trolling Air Pollution at No.3 Sintering Plant, Indiana Harbor Works, Inland Steel
Company", Journal of the Air Pollution Control Association, 13 (12), 600-603
(December 1963). -
(17) Young, T. A., "Gary Steel Works Experience With Dust Control at Number 3 Sinter
Plant", Blast Furnace and Steel Plant, 56 (12), 1057-1063 (December 1968).
(18) Akerlow, E. V., "Modifications to the Fontana Open Hearth Precipitators", Pro-
ceedings, Semi-Annual Technical Meeting, Air Pollution Control Association,
Houston, Texas. 59-72 (December 1956).
(19) Speer, E. B., "Operation of Electrostatic Precipitators on O. H. Furnaces at
Fairless Works", Special Report No. 61, Air and Water Pollution in the Iron and
Steel Industry, The Iron and Steel Institute (1959), pp. 67-74.
(20) Schneider, R. L., "Engineering, Operation and Maintenance of Electrostatic Pre-
cipitators on Open Hearth Furnaces", Journal of the Air Pollution Control Associa-
tion, ~ (8), 348-354 (August 1963).
(21) Dickinson, W. A., and Worth, J. L., "Waste Gas Cleaning at Sparrows Point
Plantls No.4 Open Hearth", AIME Open Hearth Proceedings, 47, 214-225 (1964).
(22) Smith, W. M., et al., "The Use of a Flow Model in the Design of an Electrostatic
Precipitator", Blast Furnace and Steel Plant, 55 (12), 1097-1102 (December 1967.).
(23) "Inland Steel Completes $7 Million Air Pollution Control Facilities", American
Metal Market, p. 27, June 26, 1968.
(24) "Joy Building Youngstown Precipitators", American Metal Market, p. 8, February
7, 1969.
(25) Peterson, H. W., "Gas Cleaning for the Electric Furnace and Oxygen Process Con-
verter", AIME Electric Furnace Proceedings, ~, 262-271 (1956).
(26) Smith, J. H., 'lAir Pollution Control in Oxygen Steelmaking", AIME Open Hearth
Proceedings, 44, 351-357 (1961).
(27) Rowe, A. D., et al., "Waste Gas Cleaning Systems for Large Capacity Oxygen Fur-
nace Plants", Second Interregional Symposium on the Iron and Steel Industry, United
Nations Industrial Development Organization, Moscow, September 19 -
October 9, 1968. 35 pp.
(28) Wheeler, D. H., "Fume Control in L-D Plants", Journal of the Air Pollution Con-
trol Association, ~ (2), 98-101 (February 1968).
(29) "u. S. Steel BOFs at Gary Outstrip Their Design", Steel, 202 (10), 12
(September 5, 1968).
(30) "Pittsburgh Steel Co. to Maintain Regular Deliveries Despite Blast", American
Metal Market, p. 9 (May 15, 1968).
(31) Stairmand, C. J., 'IRemoval of Grit, Dust, and Fume From Exhaust Gase s
Chemical Engineering Processes'l, . The Chemical Engineer, No. 194,
pp. CE 310-CE 324 (December 1965). .
BATTELLE MEMORIAL INSTITUTE - COLlIMI=III~ I Al=ln~AT("'~I~~
From
-------
VI-83
(32) Campbell, W. W., and Fullerton, R. W., l1High-Energy Scrubbers Can Satisfac-
torily Clean Blast Furnace Top Gas", AIME Blast Furnace, Coke Oven, and Raw
Materials Proceedings, ~, 329-334 (1959).
(33) Air Pollution Manual, Part II-Control Equipment, American Industrial Hygiene
Association, Detroit, Michigan, p. 63 (1968).
(34) Semrau, K. T., "Dust Scrubber Design - A Critique on the State of the Art",
Journal of the Air Pollution Control Association, .!2 (12), 587-594, December 1963.
(35) Lapple, C. E., and Kamack, H. J., "Performance of Wet Dust Scrubbers",
Chemical Engineering Progress, ~ (3), 110-121 (March 1955).
(36) Bowman, G. A., and Houston, R.. B., "Recycled Water Systems for Steel Mills",
Iron and Steel Engineer, 43 (11), 139-147 (November 1966).
(37) Communication from R. B. Engdahl to John Varga, Jr., Battelle Memorial Institute,
Columbus Laboratories, February 28, 1969.
(38) Communication from P. D. Miller to John Varga, Jr., Battelle Memorial Institute,
Columbus Laboratories, February 28, 1968.
(39) Harris, E. R., and Beiser, F. R., "Cleaning Sinter Plant Gas With a Venturi
Scrubber", Journal of the Air Pollution Control Association, 15 (2), 46-49
(February 1965). -
(40) "McLouth Steel Using Scrubbers at Sinter Plant", Blast Furnace and Steel Plant,
54 (11), 1051-1052 (November 1966).
(41) Lowe, J. R., "An Orifice Gas Washer", AIME Blast Furnace, Coke Oven, and Raw
Materials Proceedings, .!i., 28-30 (1957).
(42) Reid, G. E., "Experience in Cleaning Blast Furnace Gas With the Orifice Washer",
Iron and Steel Engineer, E. (8), 134-137 (August 1960).
(43) Hipp, N. E., and Westerholm, J. R., "Developments in Gas Cleaning-Great Lakes
Steel Corp.", Iron and Steel Engineer, 44 (8), 101-106 (August 1967).
(44) Morgan, E. R., et a1., liThe Rejuvenated Blast Furnace", Blast Furnace and Steel
Plant, ~ (7), 625-631 (July 1962).
(45) Eberhardt, J. E., and Graham, H. S., "The Venturi Washer for Blast Furnace
Gas", Iron and Steel Engineer, 32 (3), 66-71 (March 1955).
(46) Basse, B., "Gases Cleaned by the Use of Scrubbers", Blast Furnace and Steel
Plant, 44 (11), 1307-1312 (November 1956).
(47) Morgan, M., "Industrial Waste Treatment-Steel Plants", Iron and Steel Engineer,
37 (7), 70-74 (July 1960).
(48) Bishop, C. A., et a1., "Successful Cleaning of Open Hearth Exhaust Gas With a
High Energy Scrubber", .Journal of the Air Pollution Control Association, .!..!. (2),
83 -87 (February 1961).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-84
(49) Johnson, J. E., IIWet Washing of Open Hearth Gases", Iron and Steel Engineer,
44 (2), 96-98 (February 1967).
(50) Broman, C. U., and Iseli, R. R., "The Control of Open Hearth Stack Emissions
With a Venturi Type Scrubber'l, Blast Furnace and Steel Plant, 56 (2), 143-147
(February 1968). -
(51) Pettit, G. A., "Electric Furnace Dust Control Systemll, Journal of the Air Pollu-
tion Control Association, .!2 (12), 607-621 (December 1963).
(52) "Swindell-Dressler to Furnish Armco Furnaces", 33/ The Magazine of Metal Pro-
ducing, ~ (7), p. 18 (July 1967).
(53) 11 Armco Plans Expansion Program at Houston", Iron and Steel Engineer, 43 (3),
p. 180 (March 1966).
(54) Finney, Jr., J. A., and DeCoster, J., IIA Cloth Filter Gas Cleaning System for
Oxygen Convertersll, Iron and Steel Engineer, 42 (3), 133-139 (March 1965).
(55) First, M. W., and Silverman, L., IIPredicting the Performance of Cleanable
Fabric Filters", Journal of the Air Pollution Control Association, 13 (12), 581-586
(December 1963). -
(56) Herrick, R. A., 11 Theory, Application of Filter Drag to Baghouse Evaluation",
Air Engineering, .!.Q (5), 18-21 (May 1968).
(57) 'ISinter Line Baghouse Collector Still Going Strongll, Iron and Steel Engineer, 45
(2), p. 124 (February 1968).
(58) Herrick, R. A., IIA Baghouse Test Program for Oxygen Lanced Open Hearth Fume
Controlll, Journal of the Air Pollution Control Association, 13 (1), 28-32
(January 1963). -
(59) Herrick, R. A., et al., "Oxygen-Lanced Open Hearth Furnace Fume Cleaning With
a Glass Fabric Baghouse", Journal of the Air Pollution Control Association, 16 (1),
7 -11 (January 1966). -
(60) Campbell, W. W., and Fullerton, R. W., IIDevelopment of an Electric Furnace
Dust Control Systemll, Journal of the Air Pollution Control Association, 12 (12),
574-590 (December 1962). -
,
(61) Bintzer, W. W., IIDesign and Operation of a Fume and Dust Collection System for
Two 100-Ton Electric Furnaces'l, Iron and Steel Engineer, 41 (2), 115-123
(February 1964). -
(62) Wood, R. M., and Burcham, J. 0., IIArc Furnace Steel Production, Kansas City
Works-Armco Steel Corp. 11, Journal of Metals, ~ (12), 1005-1008 (December 1964).
(63) Bintzer, W. W., and Kleintop, D. R., IIDesign, Operation and Maintenance of a
150-Ton Electric Furnace Dust Collection Systemll, Iron and Steel Engineer, 44 (6),
77-85 (June 1967). -
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VI-85 and VI-86
(64) Wilcox, M. W., and Lewis, R. T., "A New Approach to Pollution Control in an
Electric Furnace Shop", Iron and Steel Engineer, 45 (12), 113-120 (December 1968).
(65) Venturini, J. L., "Historical Review of the Air Pollution Control Installation at
Bethlehem Steel Corporation's Los Angeles Plant", Pre print No. 68-134, Air Pol-
lution Control Association Annual Meeting, St. Paul, Minnesota (June 23 -2 7, 1968),
19 pp.
(66) Stastny, E. P., "Choosing Your Electrostatic Precipitator", Power, 104 (1),
61-64 (January 1960).
(67) Stairmand, C. J., "Dust Collection by Impingement and Diffusion", Transactions
of the Institution of Chemical Engineers, 28 (1950) pp.
(68) Robinson, M., "A Modified Deutsch Efficiency Equation for Electrostatic Precipi-
tation", Atmospheric Environment, Permagon Press, Vol. 1, (1967) pp. 193 -204.
(69) Semrau, K. T., "Correlation of Dust Scrubber Efficiency", Journal of the Air
Pollution Association, 10 (3), (June 1960), pp. 200-207.
(70) Communication from the Pangborn Corporation to Swindell-Dressler Company.
(71) Elliott, A. C., and Lafreniere, A. J., "Collection of Metallurgical Fumes
Oxygen Lanced Open Hearth Furnaces", Journal of Metals, .!..!!. (6), 743-746
( June 1 96 6 ) .
From
(72) Penney, G., "Symposium on Gas -Solids Separation", Carnegie-Mellon University,
Pittsburgh, Pa., January 14, 1969.
(73) Berg, B. R., "Development of a New, Horizontal-Flow, Plate-Type Precipitator
for Blast Furnace Gas Cleaning", Iron and Steel Engineer, 36 (10) 93-101
(October 1959).
(74) Willet,. H. P., and Dike, D. E., "The Venturi Scrubber for Cleaning Oxygen Steel
Process Gases", Iron and Steel Engineer, 38 (7), 126 (July 1961).
(75) American Air Filter Company Bulletin 294-1 OM-3 -65 -CPo
(76) Buffalo Forge Company Bulletin AP650.
(77) Robinson, M., "Turbulent Gas Flow and Electrostatic Precipitation", Journal of
the Air Pollution Control Association, .!..!!. (4) 235-239 (April 1968).
(78) Stairmand, C. J., "Design and Performance of Modern Gas -Cleaning Equipment",
Journal of the Institute of Fuel, 29 (181) 58-76 (February 1956).
(79) Punch, G., "Gas Cleaning in the Iron and Steel Industry, Part II: Applications",
Fume Arrestment, Special Report No. 83, Iron and Steel Institute, p. 10 (1963).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VII-l
SEC TION VII
PROBLEMS AND ASSOCIATED
OPPOR TUNITIES FOR RESEARCH
An objective of this study is to determine where resea'rch and development may be
undertaken to achieve the desired control of emissions to the air from the integrated
iron and steel industry. As a means for screening the various process segments with
respect to research needs, an identification key was developed to select the subjects
requiring research effort to resolve its air-quality problems, and to establish some
indication of priority. The key is given in Table VII-I.
TABLE VII-I.
IDENTIFICATION KEY FOR
PROBLEM IDENTIFICATION
Factor Key
. ~ D QJ
Level of emission Severe Moderate Minimal
Is it controlled? No Partially Yes
Can it be controlled? Partially Yes Uncertain
The key is explained as follows:
Level of emission -
Seve re -
Emission is (1) large in volume, (2) detrimental regardless
of the volume, or (3) particularly obnoxious with respect
to odor. As examples: (1) a large volume of emissions
would be the iron-oxide fume generated in uncontrolled
BOF operation, (2) detrimental in small amounts would
be applied to fluoride emissions, and (3) the evolution
of hydrogen sulphide from quenched blast-furnace slag
would be an example of a particularly obnoxious odor,
Moderate
- Emission is (1) modest in volume, (2) somewhat detri-
mental regardless of volume, or (3) slightly obnoxious or
disagreeable with respect to odor. As examples: (1) the
particulates generated during the quenching of coke,
(2) acid fumes from partially hooded pickling operations,
and (3) benzol odors downwind from a coke plant.
Minimal -
Emission is (1) at a low volume, (2) slightly detrimental
when in large volumes, but generally tole rabIe, and
(3) any odors classed as a minor nuisance. As examples:
(1) particulates generated during the indurating of pellets,
and (2) gases from mold coatings during ingot pouring.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VII-2
Is it controlled?
- This factor in the identification key is concerned with the
apparent average for the entire integrated iron and steel
industry.
No
The industry on the average does not control emissions to
the air from the process or from the process segment.
Partially - The steel industry generally does some control of emis-
sions to the air.
Yes
- The process or process segment usually appears to be
controlled adequately.
Can it be controlled? -
No
Present pollution-control technology is not adequate for
successful control.
Partially - Present pollution-control technology has some limitations
that interfere with successful control.
Yes
- Emissions can be controlled with present control-
equipment and technology.
Uncertain - Insufficient information available to determine if control
is pos sible.
It should be pointed out that under the factor "Can it be controlled? I', no process
or process segment is considered as absolutely uncontrollable with respect to emissions.
Priority for the conduct of research and development was evaluated on the basis
shown in Table VIl-2. The priority rating is based on a combination of the level of
emis sion, whether it is controlled, and whether it can be controlled. The first priority
is for processes or proces s segments having a severe level of emissions, no control or
only partial control, and the control technology is, for practical purposes, unknown or
only partially effective. Second or third priority is determined on the basis of emission
level, with the same considerations for control and the possibility of control. Any pro-
cess or process segment listed as controllable with respect to emissions by use of
present technology is not considered as an appropriate area for research and develop-
ment activities.
Ratings for the various process segments considered in this study are shown in
Tables VIl-3 through VII-8. Research and development priorities are determined on the
basis of the indicated ratings. Emissions from the process segments are designated as
follows: ~ - particulates, @) - gaseous, and ~ - aerosols.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VII-3
TABLE VII-2. PRIORITY FOR RESEARCH AND
DEVELOPMENT EFFOR TS
Level of Is It Can It Be
Emission Controlled? Controlled? Priority Rating
. . [?J
. . ~ First
. ~ [?J
~ . [?J
~ . ~
Second
~ ~ [?J
~ ~ ~
0 . [?J
0 . ~ Third
o ~ [2]
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VII-4
TABLE VII-3. RESEARCH AND DEVELOPMENT EVALUATION FOR MAKING
PELLETS AND FOR RAW-MATERIAL HANDLING
-------
VII-S
TABLE VII-4. RESEARCH AND DEVELOPMENT EVALUATION FOR COKE MAKING
iJ
r?
~
~
~
l::
:£
~
Coke Plant Coal transfer - stOrage to coke plant 0 Wind . . OJ III-41, V-2, VII-12,
A-4, C-16, C-26
Internal transfer to crushing plant 0 Handling . ~ 0 III-41, VII-12, C-16
Crushing and grinding 0 Process . ~ 0 III -41, IV -10, VII-12,
C-16
Internal transfer to stOrage 0 Handling . ~ 0 III -41, VII-12, C-16,
C-26
Coal transfer - storage to larry car m Handling . ~ 0 III-41 , V-3, VII-12,
C-16
- larry car to coke oven 0 Handling . ~ IJJ III "41, IV-10, V-3,
VII -12
~ Open oven . ~ OJ V-3, C-16
@] Lids and ~ ~ E&3 III-41, IV-10, IV-n,
Coking operation - oven door seals V-2, VII-12, C-14,
C-18, C-27
- underfiring @] Fuel. rn ~ 0 III-41, V-2, V-31,
C-18
Abrasion, III-41, IV-10, V-3,
Pushing coke 0 thermal . . OJ VII -12, C -19
draft
@] Incomplete . f2?J 0 C-19
coking
Quenching coke 0 Thermal ~ . ~
draft III -41, V-3, VII-12
Transfer to stOrage 0 Handling 0 ~ IJJ III -41, V-4, C-20
Crushing and screening m Process ~ 1221 0 III -41, VII-12, C-20
Coke-oven gas system @] Leaks 0 0 III -41, VII-12, C-21
By-product plant @] Leaks . ~ 0 III -41, V-4, VII-12,
C-21
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VII-6
TABLE VII-5. RESEARCH AND DEVELOPMENT EVALUATION FOR MAKING
SINTER AND MAKING IRON
~
~
...
-------
VII-7
TABLE VII-6. RESEARCH AND DEVELOPMENT EVALUATION FOR MAKING STEEL AND POURING INGOTS
,fJ
u
~
~
i}
~
:::
~
~
Open-Hearth Refractory maintenance 0 Handling D ~ D V-l1
Steelmaking m f22] ~ D
Charging scrap Handling III-43, V-lO, C-54
III-43, V-lO, V-38, VI-8,
Preheating and melting m Material ~ l88! D VI-18, VI-2l, VI-26, VI-27,
VI-28, A-13, C-62, C-66
@] Fuel ~ .~ D III -43, V-lO, C-62
Hot-metal addition m Chern istry . ~ D III -43, V-lO, C-62
Oxygen lancing 0 Process . ~ D III -43, V-l1, C-64
@] Process ~ ~ D V-l1
Tapping m Handling ~ . D \1-12
IT] l22J 0 III-43, IV-15, VI-9, VI-l1,
BOF Charging scrap Handling VI-18, VI-26, VI-27,
Steelmaking VI-28, A-16, C-66
Charging hot metal m Chemistry . ~ D III-43, V-12
Oxygen lancing m Process . ~ D III-43, V-12, C-69, C-70,
C-7l. C-73
@] Process . ~ D V-12, C-70, C-73, C-77
m ~ ~ D III-43, IV-16, IV-17, IV-18,
Electric - Furnace Charging scrap Handling V-14, VI-2l, A-18, C-79,
Steelmaking C-88
Charging hot metal (if used) 0 Chemistry . f82j D III -43, C -91
m ~ ~ D III-43, V-14, VI-12, VI-20,
Melting Process VI -26, VI -27, VI -28,
o . B23 D VI-29, C-9l
Oxygen lancing Process III-43, C-83
@] Process ~ ~ D V-15, C-84
Pouring Ingots Pouring without hot tops 0 Oxidation D . D III -43, III -44, C -104,
C-l06
@] Mold ~ . D V-12, A-24, C-l06
coating
@] Process D . D V-12, C-l06
Pouring with hot tops @] Material ~ . D C-l06, C-112
m Material ~ . D C-l06
Leaded steel additions m Mater{al . D C-l08
Ingot stripping 0 Process n D C-l08
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VII-8
TABLE VII-7. RESEARCH AND DEVELOPMENT EVALUATION FOR PRIMARY ROLLING,
CONTINUOUS CASTING, PRESSURE CASTING, AND CONDITIONING
qJ
~
~
~
~
!::
£
~
Rolling Billets, Maintenance - soaking pits [E] Handling ~ ~ 0
Blooms, and
Slabs Firing - soaking pits @] Fuel ~ ~ 0
Charging and removing ingots [E] Process D . IT]
Primary rolling [E] Oxidation 0 0
Continuous Flow of molten steel into machine [E] Oxidation 0 ~ 0 IV-19
Casting @] 0 0
Oxidation
Torch cutOff 0 Process D . 0 C-110
@] Process D . 0
Pressure Casting Flow of metal into mold [E] 0 0 IV-19, C-1l2
Flow of metal into hot top 0 Material D . 0
@] Material D . 0 C-1l2
Torch cutting to remove riser 0 Process D . 0 C-113
Auxiliary Preparation of ingot molds [E] Process ~ . IT]
Operations [E] ~ . IT]
Material C-I06
Preparation of pressure casting mold 0 Process 0 0
m Material D . IT]
Pigging of molten iron 0 Chemistry . ~ IT]
Conditioning Grinding and chipping [E] Process 0 ~ 0
Semifinished
Products Spot scarfing 0 Process ~ ~ 0
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VII-9
TABLE VII-8. RESEARCH AND DEVELOPMENT EVALUATION FOR HOT ROLLING, COLD ROLLING, COATING
OF FINISHED PRODUCTS, WASTE INCINERATION, AND POWER GENERATION
,if
(J
~
'lJ"
;}
Qf
:::
~
~
~ ~ D III-44, III-45 , IV-23,
Finished Heating for rolling Fuel V-41, V-18, C-1l4
Products IlI-44, V-18, VI-33, A-27,
Hot scarfing prior to rolling 0 Process . ~ 0
C-114, C-116, C-117
Hot rolling IT] Oxidation ~ ~ 0 III-45 , lV-20, V-18, A-27,
C-114, C-117
Pickling for scale removal 0 Acid mist ~ 0 IV-23, V-18, C-117. C-118
Shotblasting for scale removal 0 Proc ess ~ D C -115
Cold rolling 0 Rolling oil 0 ~ 0 IlI-45, IV-22, V-18, C-117
Hot galvanizing 0 Cover flux ~ 0 III-45 , IV-22, V-19, C-121
Electro-gal vanizing IT] Cover flux D 0 IV-23, V-20, C-120
Electro-tin plate 0 0 0 III -45, IV-22, V-20
Paint coating 0 . Proc ess ~ 0 IIl-45 , I V -22
Plastic coating m Process ~ 0 IlI-45 , IV-22
Waste Disposal of in-plant generated wastes 0 ~ ~ 0 A-28, A-29, C-124
Incineration
@] ~ ~ 0
Power In-plant generation of electric energy 0 ~ ~ 0
Generation
@] r?&J ~ IT]
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VII-IO
First Priority for Research and Development
Process or process segments rated to warrant first-priority efforts are char-
acterized by the following key designations:
Level of Is It Can It Be
Emission Controlled? Controlled?
. . !2]
. . ~
. ggJ !2]
Examination of the processes in the tables shows the following to fall into the
first-priority classification:
(1) Unloading and transfer to storage of fine ore or pellets
(2) Unloading and transfer to storage of coal
(3) Coke plant - oven charging (particulates and gases)
(4) Coke plant - pushing coke
(5) Sinter plant - ignition and firing of sinter
(6) Ironmaking - (a) casting iron, (b) flushing slag, (c) slag disposal
(7) Pigging of molten iron.
Second Priority for Research and Development
Process or process segments rated to warrant second-priority efforts are char-
acterized by the following key designations:
Level of
Emission
Is It
Controlled?
Can It Be
Controlled?
~
~
~
~
.
.
~
~
!2]
[8B
!2]
~
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VII-ll
Examination of the processes in the tables shows the following to fall into the
second-priority designation:
(1) Unloading coarse ore
(2) Coke plant - gaseous emissions from lids and door seals
(3) Coke plant - particulates during quenching of coke
(4) Sinter plant - gaseous emissions from fuel
(5) Making iron - transfer of bulk materials from storage to stockhouse
(6) Preparation of ingot molds
(7) Gaseous emissions from in-plant generation of electric energy.
Third Priority for Research and Development
Process or process segments rated to warrant only third-priority efforts are
characterized by the following key designations:
Level of Is It Can It Be
Emission Controlled? Controlled?
D . IT!
o . ~
D ~ [2]
Examination of the processes listed shows the following to fall into the third-
priority designation:
(l) Transfer of coarse ore
(2) Unloading and transferring of limestone
(3) Coke plant - handling of coke
(4) Coke plant - coke-oven-gas system
(5) Sinter plant - transfer of sinter to storage
(6) Making iron - charging from skip hoist to blast-furnace top
(7) Charging and removing ingots from soaking pits
(8) Preparation of pressure-casting molds.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VII-12
Evaluation of Research and Development Priorities
The problems, possible solutions, and implications of the changes that are re-
quired and which may occur as a result of research and development, are discussed in
Table VII-9. Making of coke is considered first because it presents major problems
requiring solution at three levels of priority. Sinter-plant operation is considered
second because it also presents problems at all three levels of priority, but these
appear, overall, less severe than those for coke making.
Coke Making
The unloading of coal from barges and movement of coal to the storage areas pre-
sents a serious particulate-emission problem because of the fineness of the coal dust and
the ease with which it can become airborne once it is in a stockpile. The unloading op-
eration can possibly be improved by the development of continuous bucket unloaders that
would transfer the coal to shrouded conveyor belts for transport to storage areas. Such
an approach would keep the particulates contained, but in all probability the shrouded
conveyor belts would also have to be hooded and exhausted to remove any airborne par-
ticulates from the system. If this were not done, excessive wear of the conveyor-belt
rollers would probably occur, with resulting high maintenance costs. This type of sys-
tem would also have a tendency to restrict the expanse of storage space. Moisture
additions at the first transfer point, which would be stationary, should tend to minimize
dust generation at subsequent transfer points.
The containment of coal in the stockpiles and prevention of coal dust from becom-
ing airborne is a serious problem. Attempts have been made to keep dust down by
moisture additions to the piles, as well as by spraying the piles with plastic films.
Neither method has proved to be satisfactory. An ideal additive to the coal stockpiles
would be one that would keep a protective film over the stockpile, would heal itself if
broken, would not be detrimental to the crushing, grinding, and coking of coal, and
could be used to control the bulk density of the coal.
The transfer of coal from storage to the coal-grinding plant is a problem similar
to that of unloading and transfer of coal to stockpiles. The Japanese have reportedly
resorted to a reclamation system that utilizes underground systems to recover the coal
from the stockpiles, thereby eliminating the problem of agitating the coal pile.
Particulate emissions from the crushing, grinding, and transport of coal to the
storage bins above the larry cars usually are under control, and particulate emissions
from this source should not be present. Grinding mills and conveyor systems can be
serviced with efficient emission-control equipment. The recovered coal dust is a plus,
because any lost dust would mean a loss in yield for each ton of coal ground, with an
attendant increase in the cost of coke.
Charging of the coke ovens from the larry car is a persistent problem, and one of
constant concern to the coke-oven operator. All lids must be open during the charging
operation, with the result that dust which becomes airborne inside the coke oven due to
the thermal draft finds an easy path to the atmosphere via any of the open ports. Con-
stant maintenance of automated lid-lifting and closing equipment and rapid charging
have been the major means of minimizing emis sions from this source. The new
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VII-13
TABLE VII-g. ANALYSIS AND POSSIBLE SOLUTIONS TO COKE-PLANT AIR-POLLUTION PROBLEMS
Process and
Technological Equipment Factots
Process Segment Problem Possible Solution Advantages Disadvantages Requiring Solution
Making coke Insufficient control of Complete enclosure and Centralized control and Create industrial hygiene Design, construction,
particulate and gas- building ventilation exhausting of emissions problems safety, and industrial
eous emissions to allowable limits Possible explosion hazard hygiene problems
Unloading of coal Particulate emissions Continuou's bucker un- Contain and minimize Restrict area of operation Design and construction
from barges and caused by clam -shell loading and discharge particu.lates and extent of storage
transfer (0 storage unloading, free-fall to shrouded belt Permit all-weather area
dumping, and wind operation
Unloading of coal Particulate emissions Enclosure and exhaust- Contain and minimize Restrict area of operation Design and construction
from railroad cars, caused by car dumping ing of car-dumping particulates and extent of storage
and transfer (0 or bottom dumping or bortom -dumping area
storage area, and shrouded
belt to storage
S rorage of coal Particulate emissions Unknown Retention of particulates
generated by wind in srorage piles
Transfer of coal Particulate emissions Underground recovery Minimize or even elim- Extensi ve rebuilding of Design and construction
from storage (0 generated by clam -, and conveyor transfer in ate emissions due (0 coal-storage facilities
coal crushing and shell recovery and to grinding plant coal recovery
grinding plant free -fall loading of
cars
Coke-oven charging P articulate and gaseous Double aspiration Reduce emissions during Unknown New charging methods
emissions generated lines charging
during charging Shrouded charging Reduce emissions during Possible explosion or
pipes and collection charging fire hazard
equipment on
larry car
Pipe -line charging Eliminate emissions
Co king Leakage of gaseous Improved maintenance Reduce emissions None Design, construction,
emissions around Improved design in Minimize emissions None and maintenance
doors, seals, and lids lids, seals, and doors
Pushing of coke Particulate generated by Hooded exhaust over Contain and exhaust Possible interference Method and/or equip-
abrasive action of coke push side ,particulate and gas- with pushing operation ment for capturing
on oven brick, and eo us emissions and collecting
thermal draft causing emissions
dispersion of particu - Fine spray nozzles to Reduce temperature of Possible interference Mechanics of quenching
lates spray coke and re- the coke as it is with pushing operations and particulate
duce temperature pushed, reduce the and create a safety emission
intensity of the ther- problem
mal draft, and mini-
mize dispersion of
particulates
Quenching of coke Emission of particulates Use of baffles Reduce particulates Quench tower operation
during the quenching Redesigned quench Minimize or even elim- anct characteristics
of hot coke tower inate particulates
Coke-oven-gas Leakage of gas Improved gas-plant Minimize leakage Design and maintenance
recovery system components
High HZS content in Develop economical Minimize or even elim-
coke -oven gas method of removing inate sulfur oxide due
HZS to burning of coke-
oven gas
By-product re- Leakage of gas and Improved by-product- Reduce emissions Design and maintenance
covery plant ae rosols plant components
Making coke Insufficient control of Develop new coke- Eliminate emissions Unknown New process and
particulate' and gas- making process equipment
eo us e missions in
present coke plants
BATTELLE
MEMORI AL
INSTITUTE - COLUMBUS LABORATORIES
-------
VII-14
European larry car which exhausts the emissions into a scrubber system located on the
larry car has not been fully successful because of fire and explosions which have re-
portedly occurred in two Canadian installations. >:< It has been reported that cooperative
work among several steel companies on the improved larry car-emission control system
has been somewhat successful. Details are not yet available, but are expected to be dis-
closed to NAPCA soon, probably by the AISI, which has informally expressed its inten-
tion to request research funding from NAPCA on this subject.
An American company (Allied Chemical Corporation, Wilputte Division) that de-
signs and builds coke ovens has under development a method of preheating coal followed
by charging into the coke oven by means of a pipeline system. The developers feel that
the method will control completely the emissions of particulates and gases to the atmo-
sphere during charging. The company is installing the system on a commercial battery
of coke ovens of its Semet-Solvay Division at Ironton, Ohio. Further information has
not yet been released by the company, but is expected to be released soon. The com-
mercial installation probably will be operational about mid-1970, unless unexpected dif-
ficulties are encountered. Assessment of the advantages and disadvantages of the sys-
tem must await operation on a commercial scale.
Leakage of gaseous emis sions around the lids, doors, and seals is a problem that
can be minimized only through constant maintenance. As coke ovens become older, the
leakage problem becomes aggravated. Coke-plant-construction companies have been
working on the development of improved designs, but under the conditions imposed by
the coking proces s, there has not yet been a substantial return on the effort put into
this part of coke -oven construction.
Pushing of coke results in particulates generated by the abrasive action of the
stove refractories on the coke as it is pushed from the oven. The thermal draft created
by the exposure of the hot coke to the atmosphere carries the particulates into the atmo-
sphere. Various types of hoods and exhaust systems have been tried on the European
continent, but none have operated successfully. The possibility of a prequenching
operation between the coke oven and quench car may be a means of reducing the temper-
ature of the coke, and thereby reducing the intensity of the thermal draft.
The quenching of coke is 'also a source of particulates; installation of baffle!) in the
quench tower has reduced this emission from 60 to 80 percent. Further work is being
carried out on this aspect of particulate control. Another subject of research could con-
cern itself with new designs for quench towers that would be a combination of quench
tower and heat exchanger to (1) prevent the steam generated from going into the atmo-
sphere, and (2) contain the particulates within the restricted area of the quench tower
itself.
New coking processes have been under development for many years, but none have
been able to compete with the existing by-product coke plant. However, the fact that the
coke by-product market has to a great extent been usurped by the petrochemical industry,
and markets for coal by-products have deteriorated, the attraction of the by-product
coke plant has diminished. Coking processes that do not recover the by-products are of
considerable interest in the integrated iron and steel industry today, and a number of new
coke-oven installations may in the future dispense with much of the by-product recovery
that has been traditional until now.
"This report has not been confirmed.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VII-IS
When new coking processes are developed, evaluation of them and of the usefulness
of the metallurgical coke they produce is usually hampered by two facets of scale. First,
the coking pilot plant often is too small to produce large amounts of coke for commercial-
scale evaluation, and second, the integrated steel industry includes only extremely lim-
ited capability to evaluate the performance of the new coke sufficiently to qualify it for
acceptance on a commercial scale. Without such acceptance, it usually is impractical
to authorize construction of a multimillion dollar coking plant based on a new proces s.
The blast-furnace operators usually are not satisfied with anything les s than fdl-scale
evaluation of a new coke. Experience has shown that such a test must be carried out for
a minimum of 1 week to provide valid results for comparison to existing practice. This
means that a blast furnace producing 2000 net tons per day of hot metal at a coke rate of
1200 pounds per net ton requires 1200 net tons of usable coke per day, which is 8400 net
tons for the complete test. Pilot facilities are seldom in a position to supply such re-
quirements. An experimental blast furnace operated by the Bureau of Mines was avail-
able for such tests and required only about 300 tons of coke to carry out the required
trial. Unfortunately, this furnace is no longer operative, and probably will soon be
declared surplus. This will leave in the United States only one small experimental blast
furnace suitable for such evaluations - the furnace owned by United States Steel.
For about a decade or more, the FMC Corporation has had under development a
new process for making coke, and at various times has operated a pilot plant at
Kemmerer, Wyoming, to produce their FMC coke. I This coke is made by a continuous
process involving fluidized beds and briquetting. One application of the coke has been
as a reducing agent in phosphorus furnaces - an application les s demanding than the
blast furnace in terms of mechanical strength of the coke. Recognizing the need for
new methods to produce coke for blast furnaces, United States Steel cooperated with
FMC to manufacture some metallurgical coke by this proces s and evaluate it in one of
the small experimental blast furnaces. The results were reasonably good, but were not
sufficiently convincing with respect to performance and cost to encourage U.S. Steel or
other steel companies to invest the much larger amounts of money needed to expand this
evaluation to a larger scale. Subsequently the Kemmerer pilot plant was closed down,
but there are today unconfirmed rumors that some thought is being given to reopening of
that plant for purposes that are unknown at this time.
At least one American coal company (probably two) have developed new methods
for making coke or char on a continuous basis under conditions that are thought to con-
trol emissions much better than those for conventional coke ovens. Island Creek Coal
Company is thought to have a process that might be of interest for application to the steel
industry, but little is known about this process at the present time. The activities of
Peabody Coal Company along these lines are of more interest. Peabody has a process
that involves pretreatment of coal on a traveling grate in a dutch oven, followed by
treatment in a shaft furnace. The coke is self-agglomerating during the process (like
coke in a conventional oven), rather than briquetted as is done for FMC coke. Through-
put time is understood to be between 1 and 2 hours, rather than about 16 hours as in a
conventional oven. Peabody installed in 1961 a demonstration plant rated at 100 tons
per day at Columbia, Tennessee; they use the plant for process development and sell
the coke for chemical purposes in the vicinity. Most of this coke was not intended to be
a metallurgical grade, but it is understood that some work was done on blast-furnace
coke. The results appeared good to Peabody, but have not been evaluated by blast-
furnace tests. The process looked good enough for some purposes that Monsanto
Chemical Company built a plant in 1966 at Decatur, Alabama. This plant is rated at
500 tons per day; and the coke is used as a reducing agent in phosphorus furnaces.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VII-16
Making Sin te r
Problems, possible solutions, and implications of changes in sinter-plant opera-
tions are given in Table VII-IO. Particulate and gaseous emissions are a problem in
the operation of a sinter plant, because each of the various proces s segments that are
included in the over-all operation has its own particular air-pollution problem that must
be handled. Sinter plants will undoubtedly be in use for many years to come, if for no
other reason than to recover the iron values that are generated as dusts in steelworks
operations.
Crushing and grinding operations associated with the preparation of materials for
sintering are straightforward, and particulate emis sions can be controlled with presently
available equipment. The transport of materials to and from sintering plants falls into
the same category. The areas of major importance in minimizing emissions is in the
ignition and sintering operation of the sinter strand. Particulates are drawn from the
sinter bed by the strong air flow required by the sintering operation. The primary fuel
used to make the sinter is coke in the form of fines. The sulfur content of the coke is
a major source of sulfur emissions to the atmosphere. Coke is the usual fuel in this
particular application, because the coke fines are to a great extent generated during
the crushing and sizing of the coke for blast furnace use. The fines are, therefore,
available at a reasonable cost for the production of sinter. Major problems exist in the
development of existing air-pollution control equipment to handle high-lime particulates
that are generated in the production of highly-fluxed sinters required to meet increased
demands for higher blast-furnace productivity. Wet scrubbers offer a possible solution
to the recovery of the high-lime particulates, as do bag houses. Wet scrubbers also
offer the possibility of recovering the sulfur gases, because the lime cont.ent of the par-
ticulates may act somewhat as a scavenger for the sulfur dioxide. The possibilities
could be explored as a combined method for recovering particulates and sulfur gases in
the same equipment. Work also is advisable to determine if other additives can be intro-
duced into the water to improve the recovery of sulfur dioxide by the lime. Particulates
generated during cooling, crushing, and screening of the sinter can be handled by exist-
ing air-pollution control equipment, and probably do not require further development in
their application. However, all of the foregoing depends to a considerable degree on
better knowledge of the nature and characteristics of particulates and gases evolved from
sintering plants. Therefore, the first step in further research aimed at better control of
emis sions from sintering plants is improved characterization of the amount and nature of
the emis sions. This improved characterization will require measurement, sampling,
analysis, and evaluation of emissions from several sintering plants being operated on a
commercial scale.
Raw Material Storage and Handling
Problems, possible solutions, and implications of change in the handling and
storage of the vast amounts of raw materials required in making iron and steel are
given in Table VII-II. The problem here is one of the most difficult to solve. Gener-
ation of particulates during the handling of raw materials from receiving to final transfer
for processing is a never-ending air-pollution problem. Particulates become air-borne
from the storage piles due to dry and windy weather conditions. The more-apparent
solutions such as complete enclosure and underground recovery systems are certainly
technically feasible, but appear to be economically impractical at this point in time.
Research and development may fruitfully be carried out to produce some type of wetting
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
VII-17
TABLE VII-10. ANALYSIS AND POSSffiLE SOLUTIONS TO SINTER-PLANT AIR-POLLUTION PROBLEMS
Process and
Technological Equipment Factots
'rocess Segment Problem Possible Solution Advantages D isad vantages Requiring Solution
Making sinter Insufficient control Complete enclosure Centralized control Limited access for Design, consrruction,
of particulate and and building and exhausting of maintenance and maintenance
gaseous emissions evacuation emissions to allow - Industrial hygiene Industrial hygiene
able limits problem for working problem
. personnel
::rushing and Collection of iron
grinding oxide, limestone,
and coke dust
fr ansport to Iron oxide, limestone, Use of water sprays Minimize generation Actual level of control Design, construction,
sinter strand and coke dusts gener- and detergents at of particulates subject to variation and maintenance
ated at transfer points transfer points
Complete enclosure Collection of particu - Creates secondary dust-
and evacuatiori of lates to allowable handling problem
transfer system limits
:gnition and Iron oxide, limestone, Ducted control of Possible reduction of Restrict direct observation Redesign of sinter
sinterihg and lime d usrs combustion air to the volume of air of sinter strand -probably strand
strand handled with re - require remote
suIting increased monitoring
collection efficiency
Oxygen enrichment Further reduction in Increase maintenance Effect of oxygen enrich-
of combustion air volume of air handled requirements ment on process
and further improve- S pace restrictions
ment in collection around sinter strand
efficiency
Sulfur dioxide gener- Reduce sulfur content Corresponding reduc- No known method for Reduction of sulfur
ated by burning coke of coke tion in S02 reducing sulfur in in coal used to make
coke coke
Wet recovery system Minimize or eliminate Corrosion resistant Mechanism of S02
with additives to S02 emissions to materials required removal
combine with S02 atmosphere
Fabric collectors Minimize or eliminate Close control required Mechanism of S02
with additives that S02 emissions to to monitor additions entrapment
combine wirh S02 atmosphere
Cooling, crushing, Generation of Can be controlled by Design and construction
and screening particulates application of exist -
ing equipment
Transfer to storage Generation of Enclosure and evacu- Eliminate particulate Increased maintenance Design, construction,
particulates ation of transfer emissions problems and maintenance
system
BATTELLE
MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
In
»
-t
-t
!11
r
r
!11
~
!11
~
o
~
»
r
Z
III
:!
-t
C
-t
!11
I
o
o
r
c
~
In
C
III
r
»
In
o
:u
»
-t
o
:u
!11
III
TABLE VII-ll. ANALYSIS AND POSSmLE SOLUTIONS TO RAW MATERIAL STORAGE AND HANDLING AIR-POLLUTION PROBLEMS
Process Segment
Problem
Process and
Technological Equipment Factors
Possible Solution Advantages D isad vantages Requiring Solution
Better methods of Minimize particulate Design and
unloading boats emissions construction
and barges
Enclosure and Minimize particulate Increased maintenance, <:
......
evacuation of emissions secondary dust handling ......
I
railroad car problem ......
00
dumps
Complete enclosure Eliminate emissions None Retention of par-
Self-healing to the atmosphere ticulates in
coating Minimize emissions Unknown storage piles
Underground re- Minimize particulates Increased maintenance Design and
covery systems construction
Unloading and
transfer to
storage of coal,
fine ore, and
pellets
Storage
Recovery from
storage
Generation of particulates
due to handling
Generation of particulates
due to drying of mate-
rials and windy weather
Generation of particulates
due to handling
-------
VII-l9
agent that will minimize the generation of air-borne particulates during storage and
handling. Fortunately, gaseous emissions are not a problem during handling and stor-
age, except for the handling and storage. of blast-furnace slag, which is a special prob-
lem now receiving research attention by the AISI, and which is discussed in the next
paragraph under "Iron Making".
Iroq Making
Problems, possible solutions, and implications of change for emissions generated
during the making of iron are given in Table VII-l2. The most pressing problems
deserving of attention in the making of iron are (l) the generation of hydrogen sulfide
from the quenching of slag, and (2) improved methods for the control of "kish".
Some research has been conducted on the mechanisms controlling the generation
of hydrogen sulfide from quenched blast furnace slags. However, the problem is far
from solved, partially because of the limited funding that has been available to carry out
the research. Further research is required on rapid analytical methods for determining
the concentration of hydrogen sulfide in the air, as well as continued work to further
investigate the mechanisms of formation and possible methods for control.
The evolution of "kish" during the cooling of hot metal is a persistent problem in
the handling of hot metal. The greasy platelets of graphite are difficult to collect and
equally difficult to remove from collecting equipment. The only available solution
appears to involve hooding of the areas where kish is evolved, and connection of these
hoods to high-velocity ducts connected in turn to particulate-collection equipment.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
TABLE VII-12. ANALYSIS AND POSSffiLE SOLUTIONS TO IRON MAKING AIR-POLLUTION PROBLEMS
Process and
Technological Equipment Factors
Process Segment Problem Possible Solution Advantages Disadvantages Requiring Solution
Transfer of bUl;: Generation of particulates Underground recovery Minimize emissions Increased maintenance Design, construction,
m materials from due to handling and maintenance
»
-i storage to
-i
III stockhouse
r
r
III Stockhouse inter- Generation of particulate More complete applica - Minimize particulate Secondary dust handling Design and installation
3: nal material emissions due to handling tion of available emissions problem
III transfer to equipment
3:
0 shiphoist
:II
» Material transfer Generation of particulate Enclosure and evacu~ Minimize emissions Increased maintenance, Redesign of blast furnace
r
shiphoist to emissions due to handling ation of blast furnace secondary dust handling top, dust handling, and
z
1/1 blast furnace top top problem, and potential safety problems
-i explosion hazard
-i <:
c: H
-i Casting of iron Evolution of "kish" from Shorter troughs Reduction in time of None Redesign and reconstruc- H
III I
N
I cooling iron iron exposure to tion of cast house 0
o atmosphere
0
r
c: Flushing of slag Evolution of hydrogen Unknown Mechanism of hydrogen
3:
m sulfide sulfide evolution and
c: means of control
1/1
r
» Slag disposal Evolution of hydrogen Unknown Mechanism of hydrogen
m
0 sulfide sulfide evolution and
:II means of control
»
-i
0 Transfer of iron Evolution of "kish" from Application of existing Minimize particulate None Design and construction
:!!
III to steelmaking cooling iron equipment emissions
1/1
furnace
Pigging of molten Evolution of "kish" from Redesign of pig machines Minimize particulate None Design and construction
iron cooling iron machines and appli- emissions
cation of available
pollution control
equipment
-------
A-I
APPENDIX A
PROCESSES IN THE INTEGRATED IRON AND STEEL INDUSTRY
Because the manufacture of many products depends on the use of steel, the iron and
steel industry has grown to be one of the largest basic industries in the United State sand
the World. The production of steel consists of making metallic iron from iron ores, con-
verting the iron into steel, casting the molten steel into shapes that are solidified, and
then further processing of the solid shapes into semifinished products such as sheet,
strip, bar, rod, plate, slab, billet, bloom, or ingot. The following section provide s a
description of the various processes involved in the production of steel.
Manufacture of Iron and Steel
The principal steps in the manufacture of iron and steel are (1) preparation of
raw materials, (2) making iron, (3) making steel, (4) casting of steel, (5) rolling into
semifinished products, and (6) manufacture of finished products. Some of the descriptive
material in this section is taken from the work of J. J. Schueneman, et al. (1 )':, For more
detailed descriptive information on the various p:rocesses, it is recommended that such
well-established references such as "The Making, Shaping and Treating of Steel"(2) be
consulted.
Preparation of Raw Materials
The major raw materials used in the production of iron are iron ore (or agglom-
erates such as pellets and sinter made from ores), limestone, coke, air, and energy
in the form of heat. One factor that contributes to the economy of the integrated pro-
duction of iron and steel is that gases produced in the making of iron and coke fre-
quently are used to meet other energy requirements in the plant.
Iron Ore. Prior to the end of World War II, the United States iron and steel
industry was essentially self -sufficient in high-grade iron ores. However, because of
the high demand for steel, these high-grade ore deposits were depleted. The industry
instituted research to develop methods of utilizing low-grade taconite ores for blast-
furnace use. These developments were instrumental in the adoption of highly bene-
ficiated burdens in blast furnaces. The change in the makeup of iron constituents
charged to blast furnaces in the United States in the period from 1957 to 1966 is shown
in Figure A-I. During 1967 the iron and steel industry consumed a total of about 136
million net tons':":' of iron are and recycled mill scale and dust in the making of pig
iron. (3) Agglomerated materials such as pellets, sinter, and briquettes accounted for
about 97 million net tons, and the remaining 39 million net tons was in the form of
lump ore, most of which was imported.
"References for this appendix are given at the end of Appendix A.
~*Net ton = 2000 pounds.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
A-2
~ 90
~
8. 80
-
.s::.
.!2I 70
CI)
~
~60
::::I
CI)
~ 50
~
o
1957 58 59 1960 61
62 63 64 1965 66 67 68
Year
FIGURE A-I.
CHANGE IN BURDEN CHARACTERISTICS FOR
UNITED STATES BLAST FURNACES
Very little crushing and grinding of ore is done at the blast-furnace plant.
Usually the only crushing and grinding located at the steel-plant site is associated with
s inte r -plant ope rations. Pellets are made in plants located at the mine site s . High-
grade ores are crushed and sized at the mines within very narrow size ranges for
shipment to the blast-furnace plants(4). Fine materials that cannot economically be
processed further and used in the production of pellets near the mines are shipped to
steel plants for use in the manufacture of sinter.
Sinter. Sintering plants are dedgned to convert iron ore fines and blast-furnace
flue dust into a product more acceptable for charging into the blast furnace. This is
achieved by burning a mixture of ore-bearing fines plus a fuel consisting of coke dust or
coal. Combustion air is drawn through the flat porous bed of the mixture. The prin-
ciple of sintering is to supply just enough fuel to the material to be sintered so that a
sticky mass will be produced, but the material will not be melted sufficiently to run.
The bed is formed on a slow-moving grate composed of receptacle elements with per-
forated bottoms, known as pallets. The assembly of such pallets end to end in a hinged
or linked arrangement comprises an endless metal belt with large sprockets at either
end. The ignition furnace is either gas or oil fired, <:'l1d its purpose is to bring the fuel
in the charge to its kindling temperature, after which the down draft of air through the
bed keeps it burning.
The sintered material is dumped from the grate as it passes over the head
sprocket upon a screen, the undersized material becoming the return fines, and the
oversized material, which is still at a red heat in the center, passing to a sinter
cooler. The cooler can be a large rotating apron or a linear grate upon which the
sinter is deposited while cool air is blown through louvers located in the apron or grate.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
A-3
As the cooler reaches a certain position, stationary scraper bars push the sinter off
the apron into cars or conveyors. A sinter plant is illustrated schematically in
Figure A-2.
FIGURE A-2.
SINTER PLANT
Modern sintering plants have capacities ranging from 2000 to more than 6000 tons
of sinter per day. One plant of the latter capacity has a sinter -bed width of 12 feet and
a bed length of about 150 feet. In 1967 a total of about 51 million net tons of sinter were
made in the United States for use in the blast furnaces. By far the largest portion of
this sinter was made within the pe rimeter of blast -furnace plants. In contrast to fired
oxide pellets (which are very strong), sinter is relatively friable and does not stand up
well physically during shipment for long distances.
Oxide Pellets. The recovery of the iron components from taconite ores can only
be done if the ores are ground to a very fine powder. Sintering of the fine taconite con-
centrates was unsuitable as a method of agglomeration, so extensive efforts were
directed toward the development of pelletizing processes to agglomerate the concen-
trates into useful sizes. The first successful commercial pelletizing plant was placed
into operation by the Reserve Mining Company in 1955. The estimated annual pellet-
making capacity in the United States in 1968 was 56.3 million net tons.
The pelletizing process was originally developed to agglomerate the fine magnetite
concentrates, but since its initial development the process has been used for the
hematite ores as well. The process consists of two main operations: (1) rolling the
fine concentrates into damp balls of a suitable size (much like making a tiny snowball),
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
A-4
and (2) drying and firing the balls to make hard pellets. The pellets made are roughly
spherical in shape and about 1/2 inch in diameter. The types of equipment for making
the balls and for hardening them vary from plant to plant. One type of pellet plant,
such as used at the Empire Mine in Palmer, Michigan, is illustrated in Figure A-3(5).
CONCINnAn IIN
IY-IA!!
S1ACK
,',;, : ," ,,""'~,."
-~:
:,;;;.,;" . "',','"
~iR).,~ :~:.~~:::.
x':,.r~J '
fAN NO, 3 {COOUI} . -,'
---.-
(NO. I'AN 150 'I
FIGURE A-3.
GRATE-KILN PELLETIZING PLANT AT THE EMPIRE MINE
Limestone and Lime. Limestone (commercial CaC03) is the major fluxing mate-
rial used in producing metallic iron in a blast furnace. It is also used in some open-
hearth furnaces. The major role of limestone is to flux silica from ores, and to com-
bine with sulfur to lowe r the sulfur content of the iron or steel. To reduce the amount
of energy required to achieve the desired chemical reactions in the making of iron and
steel, technologists have adopted the use of burnt lime (commercial CaO) as an additive
in many steelmaking processes, but not in the blast-furnace process.
Limestone is crushed and screened to the desired size at the quarry site and only
the correctly sized stone is shipped to the blast-furnace plants. Burnt lime is also
prepared, usually at the quarry, by calcining the limestone to produce a high-quality
lime for use in basic oxygen furnaces, open hearths, and electric furnaces. There are
about eight different processes for making lime for use in the steel industry. (6) Many
steel companies have their own limestone quarries and related lime -producing facilities,
while others purchase their requirements from commercial operators.
Coke. Coke, the chief fuel used in blast furnaces, is the residue after distillation
of certain grades of bituminous coal. It is made in two types of ovens: (1) the beehive,
and (2) the recuperative or by-product oven. In either type of oven, the distillation or
coking process consists mainly of driving off certain volatile matter, leaving in the
residue a high percentage of carbon mixed with relatively srnall amounts of impurities.
The beehive oven is the older of the two types of oven, and on a national scale is
unimportant in comparison to by-product ovens. On a local scale in several small
areas, however, beehive ovens can be a matter of local concern. Their use at iron
and steel works has nearly disappeared, although a few are used occasionally during
times of maximum steel production when supplies of by-product coke may be in short
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
A-5
supply at the blast furnaces. The dome -like structure is built of refractory brick. It
has a flat floor sloping slightly toward the front. In the roof is an opening through which
coal is charged and the products of distillation and combustion escape. A door in the
front permits both the regulation of the amount of air admitted during the coking proces s
and the discharge of the coke after the process has been completed. A typical beehive
oven is about 12 feet in diameter by 8 feet high, and will hold about 6.5 tons of coal.
It is insulated with loam or clay to prevent loss of heat.
Beehive coke ovens are operated continuously to conserve the heat that has been
absorbed by the oven refractories. The ovens are charged as soon as practicable after
the coke from the previous cycle has been removed or "drawn" from the oven. The
heat stored in the oven refractories is enough to start the coking cycle for the following
charge of coal. The door to the beehive oven is partially bricked up and the coal
charged into the oven through an opening ("trunnel head") in the top of the oven. After
the total charge is in the oven, the coal is leveled to provide more or less uniform
treatment of the coal. The heat stored in the oven refractories starts the coking
process very soon after the charge has been leveled. Volatile matter from the coal is
driven off by the heat and starts to burn, thereby providing more heat to continue the
coking process. Coking takes place from the top to the bottom of the coal in the oven.
The rate of evolution of volatile materials and their subsequent combustion is controlled
by regulating the amount of air entering through the opening in the oven door. After the
coking process has been completed, the door is opened by removing the sealing brick-
work, and water is sprayed over the coke to quench or "water it out". By-products
are not recovered in the beehive process. Beehive ovens almost invariably are
located in coal fields; not within the perimeters of integrated steel plants.
In the by-product coking process, coal is heated in the absence of air. The vola-
tile matter is not allowed to burn away, but is piped to special equipment that extracts
its valuable ingredients. After the extraction process, some of the gas (heating value
about 550 Btu per cubic foot) returns to the ovens for use in heating the coking chambers
and for heating in other processes in the steel plant. These ovens are rectangular in
shape. Older ovens may be from 30 to 40 feet long, 6 to 14 feet high, and 11 to 22 inches
wide. Coke ovens built since 1967 are usually about 50 feet long, 16 to 17 feet high, with
coking chambers having an average width of 18 inches. As many as 100 of them may be
set together in a battery for ease in charging and discharging the coal and coke. A mod-
ern by-product oven can receive a charge of 16 to 20 tons of coal through ports at the
top. The ports are then sealed and coal begins to fuse, starting at the walls of the oven,
which may generate heat from 1600 to 2100 F. The fusing works toward the center of
the charge from both walls, and meets in the center, causing a crack down the middle of
the mass. This crack and the porous structure of the by-product coke are its distin-
guishing features. When coking is finished (16 to 20-hour carbonizing period), doors at
the ends of the oven chamber are opened, and the pusher ram shoves the entire charge
of coke into railway cars. The load is taken to a quenching station, where it is watered
by an overhead spray. After this, it is taken to a wharf to cool prior to screening.
The volatile products that have passed out of the ovens are piped to the chemical
plant where they are treated to yield gas, tar, ammonia liquor, and light oil. Further
refinement of the light oil produces benzol, toluol, and other complex chemical com-
pounds. However, in recent years, competition from the petrochemical industry has
made the recovery of coke by-product chemicals marginally economical or uneconomical,
unless coke-plant installations are of such a large size that processing costs can
compete with petrochemicals.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
1- --
A-6
The fundamental features of a coke battery cannot be changed during its lifetime,
which amounts to 20, 30 or more years. At the end of its life it is completely razed
and a new structure embodying current technological ideas is erected to replace it.
Fuels. An integrated iron and steel plant uses a great variety of fuels, some of
which are generated as part of the plant's own operations and some of which are
purchased.
Coke -oven gas is produced during the manufacture of coke from coal in the coke
ovens. The exhausted gas does not contain any particulates and does not require
separate cleaning because particulates are trapped in the by-product recovery system.
The major impurity is hydrogen sulfide, which in burning is changed to sulfur oxides.
For some in-plant use, the hydrogen sulfide is removed. Coke -oven gas typically has
a heat content of about 500 to 550 Btu per cubic foot.
Blast-furnace gas is a product from the ironmaking process in the blast furnace.
It is a rather low-heat-content gas, with an average value of about 80 Btu per cubic
foot. The gas when exhausted from the top of the blast furnace, is laden with particu-
late materials which are cleaned from the gas before its use in the steel plant. Blast-
furnace gas is usually burned to heat blast-furnace stoves, normalizing and annealing
furnaces, foundry core ovens, gas engines for blowing, firing of boilers, and gas engines
and gas turbines for power generation. Preheated blast-furnace gas combined with pre-
heated air has been used succes sfully for heating coke ovens, soaking pits, and reheating
furnaces. (2) The present trend in the use of higher blast temperature for blast furnaces
has increased the requirement for cleaner blast-furnace gas in the heating of blast
stoves. A cleaner gas is required to prevent clogging of the checker work in the blast
stoves, because clogging decreases stove efficiency and increases maintenance problems.
Tar, one of the by-products of the production of coke, is often used as a fuel for
firing open-hearth furnaces. The tar that is burned is not cleaned, and the sulfur con-
tained in it (usually about 0.60 percent) becomes sulfur oxide in the combustion process.
Commercial fuels (principally oil and natural gas) are used in the as -received
condition at the steel plant.
Air. Air is a necessary material in the production of metallic iron in the blast
furnace. It is used as it exists in the surrounding atmosphere without any treatment
except for preheating to temperatures varying from 1000 to 2000 F, before it is blown
into the blast furnace. Air requirements for the blast furnace may vary from 45, 000
cubic feet to 60, 000 cubic feet per ton of iron, depending on the type of practice used.
Heated air is used to supply thermal energy to the blast furnace, but it can also act as
a direct replacement for coke that would normally be burned to supply this thermal
energy inside the blast furnace. A typical blast stove is illustrated in Figure A-4,
with various types of checker brick used in it.
The use of high blast temperatures permits the blast-furnace operator to inject
auxiliary fuels (such as oil, natural gas, coal, or coke -oven gas) through the tuyeres.
The injection of these auxiliary fuels results in the generation of increased amounts of
carbon monoxide in the blast furnace. (Carbon monoxide is the active reducing gas in
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
A-7
ACtUS 00011
'--
~
II'
IiiP
~j
CH,M"'[YIJAlvyii
CO",,,'C'I'IO,,. ~
COLD flL"S1 '
CO......,(.TIO'"
FR£YN TYPE
CHlIiIINf.'I' "'oHVf.
COJrol"'[CTlO"
FIGURE A-4.
111
~ "16" HI.A' DuTT
i'l
"
::
:',
iil
!
::
;:1
:.;111
::
:'
.:1
:,1
~:1 5~"P' Dun
':1 5[W1' ~tLiC'"
;:1
:d
:'
::
:'
::
j'l
::!
':i
'i'l.
~ :
:!I
ii
:'
CO""U$1'IOIi
,........f...
1
SI(I",ULl
$ltllo WALL
SUPE" OuT"
''IIGH "LVIII!N'"
""I'"
:1
. - !
, ,
j ~
H
'I
,/ jl
. il
" IL"", '.He
II'!: i!
'I :1
j. Ii!
I II "'.IUL "'IOff
II~~.'..".'..'.""'"
j t~~~~~:LA" .".,
t:l:'r:.
n! r<,,'~~,
j:: .
Ii: .
I:: :
"
::
ii I
:: I
;: I
i jj Iii I
'I:: I
Ii ii,
;: I:
Iii ;
I':
f ~
~
ii
.:
II
TYPICAL BLAST FURNACE STOVE
BATTELLE
MEMORIAL
INSTITUTE - COLUMBUS
LABORATORIES
-------
A-8
the blast-furnace process.) This use of injected fuels also results in a lowering of the
amount of coke that must be charged to a blast furnace. Typical effects of fuel-oil
injection and coal injection on the amount of coke required to make one net ton of pig
iron is shown in Figure A-5(7) and Figure A-6(8). The injection of auxiliary fuels into
the blast furnace is economically attractive to blast-furnace operators. In 1968 in the
United States, about 50 blast furnaces were using injection of natural gas; about 15 were
using fuel oil; about 3 were using coke-oven gas; 3 were using tar; and 1 was using in-
jection of coal.
1400
~
8.c:
~ e 1300
C:'-
~.[
a;-- 1200
+-0
0c:
a:o
Q)+- 1100
~ +-
OQ)
UC:
1000
Blast Temper ure
0- 1697F
. --- 1877 F
.-- 2057F
"
" "'.........
" ....,.
....... ... ""-
.... " ....
~'" -..-
......-. --
.-- --.
-...
"".
970
o 40 80 120 160 200 240
Fuel Oil Injected, pounds per net
ton of pig iron
FIGURE A-5.
EFFECT OF FUEL-OIL INJECTION AND BLAST
TEMPERA TURE ON COKE RATE
c: 1250
o
~
-
0
c: 1150
o
-
-
Q)
c: 1100
~
Q)
a.
Q)
~ 1050
u
-
0 1000
VI
"0 .
c:
:J
0 950
a.
~
Q)
C 900
a:
Q)
~
o
U
.
.
8500 50 100 150 200 250 300
Coal Injected at Tuyeres, pounds
per net ton of pig iron
FIGURE A- 6.
EFFECT OF COAL INJECTED AT THE TUYERES ON COKE RATE
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
A-9
Making Pig Irori
Pig iron has been manufactured in the United States for a great many years in
blast furnaces, and this is still the process used to make almost lOO percent of the pig
iron. Pig iron from a blast furnace is saturated with about 4 percent carbon. The iron
is tapped while molten from the blast furnace and usually is not allowed to solidify be-
fore it is delivered to the steelmaking process. The molten iron in steelmaking parlance
usually is called "hot metal'!. Only a small portion of molten iron from blast furnaces is
allowed to solidify into "pigs" for distribution in the solid state. The name "pig iron" is
generic and includes "hot metal'! and "iron pigs". Much research has been directed to-
ward the development of processes that would bypass the blast furnace as an iron pro-
ducer. The first such direct-reduction process for pig iron is expected to be operational
in the United States in 1969 near Mobile, Alabama.
Blast-Furnace Practice. The first step in the conversion of iron ore into steel
takes place in the blast furnace. The blast furnace is a large cylindrical structure about
100 feet high, lined with heat-resistant bricks. A blast furnace is shown schematically
in Figure A-7. Iron ore, coke, and limestone are charged through sealing "bells" at
the top, and heated air under pressure is blown into the lower section through the tuyeres
to burn the coke. The air is heated in stoves (described in the section on Air and illus-
trated in Figure A-4), which typically are 26 to 28 feet in diameter and over 100 feet
high. Three or four blast stoves are used per blast furnace, depending on the method
of blast heating developed in the various plants. As the solid materials (known coHec-
tively as the "burden") pass down the furnace from the top to the bottom, the reducing
gases (carbon monoxide and hydrogen) rising through the burden react with oxygen in
the ore to start the formation of iron. This reaction continues as the burden materials
flow toward the middle of the furnace, at which point the coke acts to take out still more
of the oxygen in the ore, and the limestone begins to crumble and react with impurities
in the ore and coke to form a molten slag. As the charge enters the zone of fusion, all
the materials but the coke become pasty or fused. The iron becomes a porous mass.
It then pas ses through the melting zone and becomes liquid. In this zone the ash from
the burned coke is absorbed by the liquid slag, while the iron absorbs silicon from the
slag and carbon from the coke.
The iron and slag form a molten mass in the hearth, the slag floating on a pool of
iron 4 or 5 feet deep. About every 4 or 5 hours iron and slag are drawn off. The slag
is removed more frequently than the iron. From 100 to 300 (or more) tons of iron are
drawn off at each time. The hot-metal or ladle cars which receive the iron range in
capacity from 40 to 160 tons. The ladle car usually is a special type of tank car that
makes it possible to deliver hotter iron to the steel works, even though it may be 20
miles away. Most of the metal produced in the blast furnace is used in molten form for
the manufacture of steel in open hearth and other types of steelmaking furnaces.
To produce 1 ton of pig iron requires, on the average, 1.7 t.ons of iron ore,
0.9 ton of coke, 0.4 ton of limestone, 0.2 ton of sinter, scale, and scrap, and 4. 0 to
4.5 tons of air. In addition to the pig iron, the furnace yields about 0.5 ton of slag and
about 6 tons of gases per ton of pig iron produced. Air constitutes over one -half of the
material entering the furnace, whereas gases constitute more than three -quarters of
the materials leaving the furnace. The difference is due to the fact that much of the
carbon and oxygen entering as solids, in the coke and ore, respectively, emerge as
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
A-10
:jl
II:
:' :11
~!:'Ii
:' ,I
~ !
"... :
If
~~'
~~
,";
,,'.
,.:'
:!:;
~':;
FIGURE A-7.
TYPICAL BLAST FURNACE
BATTELLE
MEMORIAL
INSTITUTE - COLUMBUS LABORATORIES
-------
A-II
gases. These gases, piped from the top of the furnace, are rich in carbon monoxide,
which can be burned. They are used to heat stoves and generate power. About 30 per-
cent of the gas is required to heat stoves, and the remainder is used for steam genera-
tion, unde rfiring of coke ovens, or he ating soaking pits.
Direct Reduction. Processes which bypass the blast furnace as a means of pro-
ducing pig iron usually make use of much smaller equipment and are not dependent on
handling such large quantities of materials as required in blast-furnace plants. The
first direct-reduction process to be constructed for commercial production of pig iron
in the United States is the Dwight-Lloyd-McWane (D-LM) process. The plant (under
construction near Mobile, Alabama) will be very small by steel-industry standards
(only about 200,000 tons per year). A typical flow sheet for the process is shown in
Figure A-8.
1- 0.. C..I PI..
~~~
~~:
~.
Q ~AAAAA/\A4
(........i..
..cJCI.
'.C.If;...
FIGURE A-8.
DWIGHT-LLOYD-McWANE DIRECT REDUCTION PROCESS
The D-LM Process makes use of a balling operation to prepare powdered ore,
coal, and flux for partial reduction on a sintering machine, after which the partially
reduced pellets are charged into an electric smelting furnace where the pellets are
further reduced and melted to make pig iron.
Although the D-LM Process is mentioned here by way of illustration, and because
it is a "first" for the production of pig iron in the United States by a means other than
the blast furnace, this particular process does not have any unique worldwide impor-
tance among processes of this general type. For example, by far most of the pig iron
that is produced throughout the world by means other than the blast furnace is made in
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
A-12
electric smelting furnaces of the Tysland-Hole type; sometimes with some type of
preheating or pretreatment of the burden. The largest installations of this type are in
Norway and Venezuela where prices of electrical energy are low in comparison to
prices for coal. Such processes for pig iron have not yet been adopted in the United
States, nor does there seem to be significant pressure to adopt them on a large scale
in the near future.
In addition to direct-reduction processes that produce molten pig iron (as dis-
cussed in the preceding paragraph), other direct-reduction processes perform the
reduction of ore to metallic iron without going through a melting step. Examples
include (1) the HyL Process now operational in several plants in Mexico, which uses
natural gas as the main source of heat and reductant, and (2) the SL-RN process now
being implemented by plant construction in New Zealand and Korea, which uses coal as
the main source of heat and reductant. Processes in this category produce "metallized
ore" or "sponge iron" (without going through the pig iron stage), and these solid
products are then melted and refined in steelmaking furnaces in a manner generally
similar to melting and refining of scrap. The first plant of this general type in the
United States presently is under construction near Portland, Oregon, using the Midland-
Ross "Midrex" process, but as in the cas,e of the D-LM plant, rated output will be
small in comparison to even a small blast-furnace plant.
Making Steel
Steel in the United States is made by three major processes, (1) open-hearth
practice, (2) basic oxygen (BOF) practice, and (3) electric-furnace practice. At one
time, steelmaking in the Bessemer converter was one of the predominate processes
used for making steel. By 1948, the production of Bessemer steel in the United States
had decreased to about 4.2 million net tons per year. This production decreased further
to about 1.4 million net tons by 1958, and by 1967 the total production of Bessemer steel
was only 0.3 million net tons. During 1968, Jones and Laughlin Steel Corporation shut
down the last Bessemer converters in the United States integrated iron and steel indus-
try. These converters were located at their Aliquippa, Pa., plant. The only Bes semer
converter remaining in operation in the United States is located at the A. M. Byers
Company, Ambridge, Pa., and is used in the manufacture of wrought iron. (9)
Bessemer converters are not considered in this study.
The production of carbon raw steel':' from 1954 through 1968 is shown in Fig-
ure A-9. The rapid increase in production of carbon steel in basic oxygen furnaces
(BOF) and the simultaneous decrease in tonnage made in open hearths is quite evident.
The relationship for the full year 1968 for the major steelmaking processes in
the United States was as follows:
°AISI definition. Raw steel is steel in the first solid state after melting and suitable for further processing or sale and includes
ingots, steel castings, and continuous or pressure-cast blooms, billets, slabs, or other product forms.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
A-13
Process
Production of Raw Steel in 1968
Percent
of Total
Millions of Net Tons
Open hearth furnace
Total
66. 1 50.4
48.6 37. 1
16.4 12.5
131.1 100
Basic Oxygen furnace (BOF)
Electric furnace
130
120
110
(/)
"0100
C'I
c
;: 90
o
(/) 80
c
o
~ 70
.....
~ 60
g 50
Carbon - Steel Ingots
~ 40
30
20
Oxygen Converter
10
o
Electric furnace
Bessemer
1965 1970
Year
1975
1955
1960
FIGURE A-9.
PRODUCTION OF CARBON RAW STEEL IN THE
UNITED STATES BY VARIOUS PROCESSES
Open-Hearth Steelmaking. The open hearth furnace at one time accounted for
about 90 percent of the steel made in the United States. During recent years, increa!:!.ed
use of the basic oxygen furnace and electric furnace has decreased the production of
steel in open hearths to about 55 percent of the industry total by early 1968.
Open-hearth steel is made usually from a mixture of scrap and hot metal in vary-
ing proportions, depending on relative cost and availability of these two main raw
materials. The object of the operation is to lower the impurities present in the scrap
and pig iron, which consist of carbon, manganese, silicon, sulfur, and phosphorus, to
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
A-14
the limits specified for the particular grade of steel. This refining operation is
carried out by means of a slag that forms a continuous layer on the surface of the
liquid metal. This slag consists essentially of lime combined with the oxides of
silicon, phosphorus, manganese, and iron, which are formed or added during the
operation.
Open-hearth furnaces are of two types, depending on the character of the refrac-
tory material that forms the basin holding the metal. Where the refractory material is
mainly siliceous fireclay or another silica-rich refractory, the furnace is described as
"acid", and where the basin is lined with dolomite (or magnesite), it is termed a
"basic" furnace. Acid open hearths are used mainly for making steel castings in the
foundry industry. Steel in the integrated iron and steel industry is made mostly in
basic open hearths.
Open-hearth furnaces in the integrated steel industry are large massive struc-
tures. The open-hearth furnace proper consists of a shallow rectangular basin or
hearth enclosed by walls and roof, all constructed of refractory brick, and provided
with access doors along one wall adjacent to the operating floor, as shown in Figure
A-IO. A tap hole at the base of the opposite wall above a pit is provided to drain the
finished molten steel into ladles. Fuel in the form of oil, coke -oven or natural gas, tar
from coke making, or producer gas (a gas rich in carbon monoxide manufactured by
blowing a limited quantity of air through a hot bed of solid fuel) is burned at one end.
The flame from combustion of the fuel travels the length of the furnace above the charge
resting on the hearth. Upon leaving the furnace, the hot gases are conducted in a flue
downward to a regenerative chamber called checkerwork or checkers. This mass of
refractory brick is systematically laid to provide a large number of passageways for
the hot gases. The brick mass absorbs heat, cooling the gases to about 1200 F. All
the elements of the combustion system burners, checkerwork, and flues are duplicated
at each end of the furnace, which permits frequent and systematic reversal of flow of
the flame, flue gases, and preheated air for combustion. A system of valves in the
flue effects the gas reversal so that the heat stored in checkers is subsequently given
up to a reverse -direction stream of air flowing to the burners. In some plants, the
Checker
Chambers
FIGURE A-IO. CROSS SECTION OF A BASIC OPEN-HEARTH FURNACE
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
A-15
gases leaving the checkerwork pass to a waste -heat boiler for further extraction of
heat so as to lower the temperature from around 1200 F to an average of about 500 or
600 F. Open-hearth furnace capacities span a wide range. The median is between 100
and 200 tons per heat (batch of finished steel), but there are many of smaller capacity
and an increasing number of larger capacity. Time required to produce a heat is com-
monly between 8 and 12 hours without the use of large amounts of oxygen.
The open-hearth process consists of se'veral stages: (1) tap to start, (2) charging,
(3) meltdown, (4) hot-metal addition, (5) ore and lime boil, (6) working (refining),
(7) tapping, and (8) delay. The period between tap and start is spent on normal repairs
to the hearth and plugging the tap hole used in the previous heat. During the charging
period, the solid raw materials (which usually include a combination of pig iron, iron
ore, limestone, scrap iron, and scrap steel) are dumped into the furnace by special
charging machines. The melting period begins when the first scrap has been charged.
The direction of the flame is reversed every 15 or 20 minutes. When the solid material
has melted, a charge of molten pig iron is delivered direct from the blast furnace in
large ladles and poured into the open hearth through a spout set temporarily in the
furnace door. This is the normal sequence for a. "hot-metal" furnace; but for a cold-
metal furnace, only solid materials (pig iron and/ or steel scrap) are added, usually in
two batch charges.
The hot-metal addition is followed by the ore and lime boil, which is a bubbling
action much like the boiling of water and is caused by the oxidized gases rising to the
surface of the melt. Carbon monoxide is generated by oxidation of carbon and is
characterized by a gentle boiling action called "ore boil". When carbon dioxide is
released in the calcination of the limestone, the more violent turbulence is called the
"lime boil".
The aims of the working pe riod are (l) to lower the phosphorus and sulfur content
to levels below the maximum level specified, (2) to eliminate carbon as rapidly as
possible and still allow time for proper conditioning of slag and attainment of proper
process temperature, and (3) to bring the heat to a condition ready for final deoxidation
in the furnace or for tapping. At the end of the working period the furnace is tapped,
with the temperature of the steel at approximately 3000 F.
The delay period includes waiting time during the heat cycle (e. g. equipment
breakdown, tapping equipment in use on another furnace, etc.) plus repair work not
usually done during the tap to start period.. For normal operation of a 10-furnace shop
as a whole, the following breakdown of the heat stages has been made:
Period
Percent of Time in
Indicated Period
Tap to start
Char ging
Meltdown
Hot-metal addition
Ore and lime boil
Working (refining)
Tapping
Delay
6
12
12
3
38
19
2
8
100
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
A-16
The use of consumable lances to inject gaseous oxygen into the bath during the
refining period and speed the oxidation reactions, shorten heat time, save fuel, and
increase production has become more or less standard practice over the last 10 to 12
years. Since 1957 water -cooled lances inserted through the furnace roof have come
into prominent use, Frequently, oxygen lances are used throughout the heat with the
exception of the charging and hot-metal-addition periods, By use of high oxygen flow
rates from hot metal to tap, production rates of 90 to 100 tons per hour are conceivable
in a 300 -ton furnace. Oxygen consumption under these conditions ranges from 600 to
1000 cubic feet per ton (900 to 1667 scfm during the period oxygen is being added),
There has been some experimentation with oxy-fuellances, i. e" the use of
oxygen in combination with the fuel. This procedure plus the substitution of burned
lime for limestone has increased the steel output of a 200 -ton furnace from 20 to
approximately 30 tons per hour.
Upright Basic Oxygen (BOF) Steelmaking. A process for refining molten pig iron
("hot metal") to make steel was developed in 1952 in Linz-Donawitz, Austria, in which
a top -blown oxygen converter was used to refine the pig iron. Although there are now
several variations in practice, the general technique worldwide is known as the "basic
oxygen" or. BOF proce s s. The furnace is a pear -shaped steel shell lined with refrac-
tory brick as shown in Figure A-II. The usual charge for this type of furnace consists
of hot metal (molten pig iron), steel scrap and flux. The ratio of hot metal to scrap
conventionally is about 70/30, The steel s crap can be replaced with iron ore or pre-
reduced iron pellets. A water -cooled lance is used to supply high-purity oxygen at
high velocity to the surface of the metal bath. The high velocity of the oxygen results
in impingement on the liquid-metal surface, which in turn produces violent agitation
and intimate mixing of the oxygen with the molten iron. Rapid oxidation of the dissolved
carbon, silicon, and manganese produces a heat of steel. In the blowing process,
some of the iron is oxidized as well and pas s e s off as fume, The B OF p roce s s diffe r s
from open-hearth practice in that external heat does not have to be supplied to facilitate
the refining of the iron. The only sources of heat are (l) the sensible heat from the
hot metal, and (2) the heat released by the exothermic reactions between the oxygen
and metalloids in the charge (primarily silicon and carbon), In March, 1969, there
were 27 steel plants in the United States with BOF installations with a total rated annual
capacity of 57 million tons. The 60 existing vessels have capacity ratings from 75 to
325 tons per heat. An additional 19 million tons of annual capacity is under construction
or planned for operation through 1970, The time required to make a heat of steel in
the BOF is much shorter than in the open hearth. Heat time for a typical 150-ton BOF
operation is as follows:
Charge scrap
Charge hot metal
Oxygen blow
Chemical tests
Tapping time
1 minute
2 minutes
20 minutes
5 minutes
5 minutes
Total time
33 minutes
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
A-17
TYPICAL SASIC OXYGEN FCE.
IS.QF)
FIGURE A-II.
BASIC OXYGEN FURNACE
~
~
~
Z
o
u
~
o
~
~
~
Z:
0,
~'
~
~
~
c
..
ZZ
00
~~
..~
Rotary Basic-Oxygen Steelmaking. A steelmaking process developed in Sweden
makes use of a rotating vessel that is operated in a nearly horizontal position as shown
in Figure A-12(10). The process is known as the Stora-Kaldo Oxygen Process.
Only one steel plant in the United States has a rotary-oxygen-furnace installation;
Sharon Steel Corporation has two Kaldo vessels with a nominal capacity of 150 net tons
of steel per heat. Operation of the Kaldo converters is somewhat similar in principle
to that for the upright BOF vessels, but there are significant differences that increase
heat time over that of the BOF and permit the use of more scrap than in the BOF.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
A-IS
OXYC:EN
CHARGING POSITION llM[ - ORt J7
-----"--
~." '- ;.:;:-....
A/' ,," -;'--:-...
(,'.:" , ." :' .,....~
. : \ . 1/
'.";"- '"'.'.c " J
.' '.."; ". ,'/. .
FIGURE A-12.
STORA-KALDO ROTARY OXYGEN CONVERTER
Electric -Furnace Steelmaking. Whereas the open-hearth and BOF steelmaking
processes conventionally use a charge that contains a high percentage of molten pig iron
(hot metal) obtained from blast furnaces, electric steelmaking furnaces conventionally
have no hot metal in their charge (although there has been some limited use of hot metal
in electric steelmaking furnaces). Generally, electric steelmaking furnaces depend on
metallic scrap for most of their charge. Because electric steelmaking furnaces permit
a high degree of control over their operations, expecially with regard to ability to hold
the steel for long refining periods and to control temperatures to high levels in the
furnace, they are generally preferred for the manufacture of alloy and stainless steels.
Although electric furnaces account for only about 12 percent of the total raw steel made
in the United States in 1967, they accounted for about 36 percent of the alloy and stain-
less steel. Of the total steel made in electric furnaces in the United States in 1967,
about 41 percent was in alloy and stainless grades. Comparable alloy and stainless
fractions for the other two major types of furnaces were about 9 percent of open-hearth
production and about 6 percent of BOF production.
The furnaces employed in electric -arc melting practices in the integrated steel
industry are refractory-lined cylindrical vessels with large graphite electrodes passing
through the furnace as shown in Figure A-l3. Electric energy is supplied to the elec-
trodes by transformers ranging in capacity from 4,000 to 85,000 kilovolt-amperes.(ll, 12)
The trend in recent years has been to provide electric-arc furnaces with larger trans-
formers than previously thought feasible. By installing larger transformer capacity,
the productivity of a given electric-arc furnace can be doubled. The largest installa-
tions to date are a 200 -ton direct-arc electric furnace powered by a 76, OOO-kV A trans-
former at the Laclede Steel Company, Alton, Illinois(l2); four 200-ton furnaces at the
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
A-19
Port for third electrode
FIGURE A-13.
DIRECT-ARC ELECTRIC FURNACE
Republic Steel Corp., Canton, Ohio, plant and a 250-ton furnace at Northwestern Steel
and Wire Company which was placed in operation in early 1969. The relationship be-
tween melting capacity of direct-arc electric furnaces in the United States and their
transformer capacities are shown in Figure A-14. (11, 12)
80
.
II)
~70
~
E
c 60
I
.
-
o
>
o 50
~
.
. .
.
g' 40
-
c
a:::
~30
E
~
o
~ 20
c
e
.....
.
.
..
.
.
.
..
10
..
.
I~....
:/'
.. .
..
o
o
FIGURE A-14.
50 100 150 200 250
Furnace Capaci t y, net loos
RELATIONSHIP BETWEEN ELECTRIC-ARC FURNACE
CAPACITY AND TRANSFORMER RA TING
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
A-20
Only basic -lined electric -arc furnaces are used to make steel in the United States
integrated iron and steel industry. Some acid-lined electric-arc furnaces are still in
use in the United States, but these are used to produce special steels in the foundry
indus try.
The metallic charge for direct-arc electric-furnace melting of steel usually con-
sists of mostly steel scrap, along with some cast-iron scrap or solid pig iron. Charge
preparation consists of the selection of the proper grades of scrap for the steel to be
made, and the addition of alloying elements to achieve the desired composition. High-
purity gaseous oxygen usually is used today to carry out refining more rapidly than
would be done with the older practice of using additions of iron ore as the source of
oxygen. .
Electric -induction furnaces are used in the integrated iron and steel industry only
to melt special alloys and stainless steels on a scale that is very small and scattered
when compared with methods for melting high-tonnage steels. Because induction-
furnace melting involves no products of combustion, and lacks the high-temperature
arc of the steelmaking electric-arc furnace, it is the cleanest method for melting steel.
If the scrap placed into an induction furnace is clean, emission from the furnace is
minor, with respect to both quantity and density, and is easily collected in simple equip-
ment. Because induction furnaces are used only in specialty situations, the scrap
charged to them almost invariably is selected with care as to composition and cleanli-
ness. In those cases where contaminated (e. g., oily) scrap is charged, the scrap
usually will emit fume and smoke until the contaminant is burned off. To minimize
fume emission during melting in such furnaces, contaminated scrap sometimes is pre-
heated, thus moving the point of evolution of fume from the melting operation back to
the preheating operation. In general, however, induction-furnace melting of steel is
a miniscule contributor to air-pollution problems.
Scrap Preheating. Preheating of steel scrap for charging into steelmaking fur-
naces is not a common practice, but when it is done it is accomplished by three tech-
niques: (1) heat exchange by the removal of sensible heat from gases not undergoing
combustion as part of the preheating cycle, (2) by the use of air-fuel burners, and.
(3) by the use of oxy-fuel burners. The use of air-fuel burners is most common. With
the use of such burners (as with the employment of noncombustion processes), tem-
peratures attained by the scrap are rarely above 1800 F. At these temperatures, the
only appreciable potential for particulate emission from the preheating step rests in
the presence of oil, paper, rubber, and other combustibles in the scrap. If the scrap
contains such combustibles, a considerable amount of fume can be generated during
preheating. In comparison to these first two methods of preheating, oxy-fuel preheating
adds additional p'roblems because of the higher temperatures that can be attained during
preheating. These higher temperatures extend into and past the melting range of the
steel.
A new steelmaking technique developed in the United Kingdom in 1962-63 made
use of an oxygen-fuel burner instead of an oxygen lance to achieve melting and refining
of low-carbon and low-alloy steels. (13) This fuel-oxygen scrap (FOS) process, has
not gained acceptance in the United States, but the idea of fuel-fired burners and oxygen-
fuel burners has been adapted to a limited degree in the United States for the preheating
of scrap in open-hearth, BOF, and electric steelmaking practices.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
A-21
Oxygen-fuel burners are used as preheaters to shorten the time required to melt
scrap in the steelmaking furnace, and thereby reduce the overall tap-to-tap time in open-
hearth practice, and to permit use of greater amounts of scrap in the BOF's. Tap-to-
tap times in open hearths have been decreased by 12 percent with a corresponding pro-
duction increase of 15 percent. (14) Scrap charges in BOF's have been increased from an
average of 28 percent of the metal charge to 36 percent of the charge. (15) Republic Steel
Corporation, at its Chicago Works, has four open hearths operating with oxygen -
natural gas roof burners (16); Inland Steel Company at East Chicago, Indiana, has in-
stalled oxygen - natural gas burners in all open hearths of their No.3 open-hearth
shop. (14) Wisconsin Steel Works, Chicago, Illinois, has used oxygen - natural gas
burners in their two 120.:.ton BOF's, (17) and the Pittsburgh Steel Company has used
oxygen - oil burners in their 200-ton BOF. (15)
Vacuum Degassing of Molten Steel
In the early 1950's several catastrophic failures of large electric-generator turbine
rotors were traced to the presence of hydrogen in the steel. These events were quickly
followed by research and development efforts directed toward developing methods of
. eliminating hydrogen from steel. In the United States, the first vacuum degassing instal-
lations were placed into operation in 1956. Steel technologists were not long in finding
that vacuum degas sing could be used also as a means of deoxidizing steels. The rapid
reaction between carbon and oxygen under reduced pressures produces a cleaner product
than when oxygen is removed by use of conventional deoxidizing additions such as silicon
and aluminum. This carbon-deoxidation technology was rapidly placed into practice by
the steel industry, as evidenced by the number of units installed in the years that fol-
lowed. The number of units installed from 1956 through 1967 is shown in Figure A-IS.
en
-
'2
::> 70
0'1
,~
~ 60
o
l50
I
E 40
::::I
::::I
°30
~
Cumulative
total installations
'020
~
~ 10
E
::::I
Z
1956 1957 195819591960 1961 1962 1963196419651966 1967
Year
FIGURE A-IS.
NUMBER OF VACUUM DEGASSING INSTALLATIONS
IN THE UNITED STATES
(Source: Battelle compilation. )
Vacuum degassing processes can be divided into three general groups:
degassing, (2) circulation degassing, and (3) ladle degassing.
(1) stream
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
A-22
Stream Degassing. The stream-degassingprocess was the first to be placed into
operation for the treatment of steel that is cast into large ingots for subsequent forging
into rotors for electric generators. In this process the vacuum-treated steel is collected
directly in the ingot mold that is inside the vacuum chamber as shown in Figure A-16.
If the degassed steel is collected in a ladle that is located inside the vacuum tank, the
OBSERVATION PORT
OR TV CAMERA
WATER
COOLING
STOPPER ROD
OBSI!RVATION PORT
OR TV CAMERA
VACUUM-
DEGASSING
CHAMBI!R
INGOT MOULD
FIGURE A-16.
INGOT STREAM DEGASSING
process (called ladle -stream degassing) is conducted, as illustrated in Figure A-17. The
ladle of vacuum-treated steel is removed from the tank and transported to the ingot-
pouring area where the steel is then cast into ingot molds. There are other variations
that perform the same type of operation by slightly different mechanical means, but the
end results are generally the same as for the processes illustrated.
To vacuum - -
system
FIGURE A-17.
LADLE-STREAM DEGASSING
Circulation Degassing. Some steelmaking technologists believed that the treat-
ment time in stream degassing was too short to take full advantage of the potential of
vacuum degassing. Several processes were developed to extend the vacuum-degassing
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
A-23
cycle and permit a longer treatment time. The circulation processes involve taking a
part of the molten steel from the ladle into the vacuum treatment chamber for treatment
and then returning the steel to the ladle. The cycle is repeated until the oxygen and
hydrogen contents of the steel are reduced to acceptable levels. Representative
circulation-degassing processes are shown in Figure A-18.
. -- ---
" if A
a.
D-H Process
Refractory lined
vacuum chamber
~
AltOON
P'RI!HI!AT
I!IUltNI!It
ASSI!MI!ILV
Process gas.
b.
Thermo-Flow Process
R-H Process
c.
FIGURE A-18.
CIRCULATION-DEGASSING PROCESSES
a.
Purge Degassing
Ladle Degassing. Ladle-degassing processes provide agitation or stirring the
molten steel in the ladle that is positioned in a vacuum tank. One process bubbles argon
gas through the steel to agitate the molten metal, and another process stirs the metal by
means of an induction coil. These processes are illustrated in Figure A-19.
Alloy
additions
Outlet to 1
vacuum p.JI'f1)S
Allo, t""" 1
Observation port
-'
..- -
Vacuum seol
Argon tu be
Furnace ladle
Sight
port
Vacuum
jets
::!J
Stainless
steel ladle
(non-magne-
tic)
Ladle-Induction Degassing
b.
FIGURE A-19. LADLE-DEGASSING PROCESSES
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
A..24
Manufacture of Semifinished Products
An essential step in the preparation of molten steel for further processing into sal-
able products is the solidification of the molten steel into shapes that can be proces sed
into the desired products. The traditional method has been to pour (teem) the steel into
ingot molds, permit the steel to solidify, remove the ingots from the mold, reheat the
ingots, and roll them into the desired semifinished products such as billets, blooms, or
slabs. In recent years, two new methods have been developed and placed into USe by the
integrated iron and steel industry. These processes are (1) continuous casting and
(2) pressure casting, both of which eliminate much of the processing associated with con-
ventional ingot practice. Conventional ingot practice accounted for about 94 percent of
the raw steel produced in the United States in 1967. It is estimated that continuous cast-
ing accounted for 5.5 percent and pressure casting 0.5 percent.
The teeming of molten steel into conventional ingot molds at one time was accom-
panied by much evolution of smoke and fume, primarily because tar and other bitumens
were used as mold coatings. During the last decade the use of such coatings has been
drastically curtailed, so that visible emission during teeming is less than formerly.
However, under some teeming practices, evolution of air contaminants is high enough to
restrict visibility at the teeming station. This degree of evolution does not occur in the
newer continuous-casting and pressure-casting processes.
Some grade's of free- machining steels involve intentional additions of lead or sulfur
to the steel shortly before or during teeming. Because of the volatility of these elements
at the temperature of molten steel, fuming of the additives represents an emission prob-
lem. The tonnage of such steels represents a small fraction of total steel produced
nationally, but can be substantial at particular steel plants that make such grades.
Conventional Ingot Practice. Conventional casting and rolling require a large
amount of plant area to accommodate the teeming area, soaking pits, and roughing mill.
In addition, extens ive transport facilities are required to handle ingots and ingot molds.
Conventional ingot-casting practice is illustrated in Figure A-20. Molten steel is trans-
ported in ladles to the teeming station for pouring of the ingots. Molds are transported
~I
Teem Solidify
~ ,,,"'~, '''"'~~
,," ~ ",;,
lWJ Soaking pil
FIGURE A-20.
CONVENTIONAL INGOT-CASTING PRACTICE
to the same area on special cars holding two or three molds to the car. One string of
cars holds enough ingots to receive all of the molten steel from the heat. After the teem-
ing operation, the ingots are permitted to solidify in the mold and cool to a selected
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
A-25
temperature, which is dependent on the type of steel produced. After the proper hold-
ing time, the ingots are transported to the stripping station, where the molds are re-
moved. The ingots are then transferred to soaking pits where they are soaked to equal-
ize the temperature throughout the ingot and heated to the desired rolling temperature.
The heated ingot is removed from the soaking pit and transported to a roughing mill
where it is rolled into billets or blooms, or to a slabbing mill where it is rolled into
slabs. Billets, blooms, and slabs differ in size and shape. Billets are usually square
and measure 2 x 2 to 5 x 5 inches. Blooms are usually square or slightly oblong and
measure 6 x 6 to 12 x 12 inches. Slabs are always oblong and measure 2 to 12 inches
thick and 20 to 70 inches wide. Billets are used to produce bar and light merchant prod-
ucts, blooms to make heavier merchant products and structural products, and slabs to
make strip, sheet, and plate products.
Continuous Casting. For many years, steel producers recognized that continuous
casting was possibly the ultimate method for the conversion of molten steel to semi-
finished products. The first United States patent was issued to Sir Henry Bessemer in
1865. There was a period of experimentation on a pilot-plant scale in the United States
about 1940. Mechanical and material problems prevented early development of this
process, and it was not until 1943 that the first continuous- casting installation was suc-
cessfully operated in Germany. This was followed by further work in the United States
in 1946, Austria in 1947, the United States again in 1949, and Germany in 1950. In the
following years the Russian technologists devoted a great amount of effort to the process
and succeeded in placing several commercial plants into operation. Efforts in Europe
and the United States resulted in commercial installations in the 1950 IS. The estimated
capacity for continuous casting of raw steel in the United States in 1968 was about 7 mil-
lion net tons. Capacity now under construction is expected to increase this figure to an
estimated 14 million net tons in 1969, and 16 million by 1971.
Continuous-casting machines are of three general types: (1) vertical machines,
(2) vertical machines with bending rolls, and (3) curved-mold machines. These are
illustrated in Figures A-21 (18) and A-22(l9).
Pressure Casting. Pressure casting is a relatively new method for converting
molten steel into semi-finished products. The process was originally developed for the
manufacture of cast-steel freight-car wheels. Additional research and development led
to applic:ation of the proces s for making cast slabs. The first commercial installations
were for the production of stainles s steel slabs. Construction on the first plant des igned
to make plain-carbon-steel slabs was started in 1968 with start-up scheduled for early
1969. Total production of pressure-cast steel in 1967 is estimated to be 500,000 net
tons, all of it essentially stainless steel. With the addition of the carbon-steel facility
in 1969 this figure should increase to 1 million net tons annually.
The principal parts of a pressure-casting unit are shown schematically in Fig-
ure A-23. Operation consists of (1) placing a ladle of molten steel in the pressure tank,
(2) covering the tank with a special cover that includes a special preheated ceramic tube,
(3) positioning 'the mold over the tank and pouring tube, (4) introducing air pressure into
the tank and forcing the molten steel up the ceramic tube into the mold cavity, (5) seal-
ing the pouring tube with a refractory plug to prevent the flow of the molten steel from
the mold back into the ladle, (6) releasing pressure in the tank, and (7) moving the mold
from the tank and positioning another mold for another casting cycle. The mold is held
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
A-26
.,. -- .I..aollom-pour lad I.
J'
Siopper......
.. ..... Trou;h tun dish
.......u Woter-cool.d copper mold
..m.Osciliotin; mold tobl.
..''''''''.S.condory spray cool in;
.".'-""Withdrowol rolls
, ............Clomp
........Counterw.i;hltd cutoff from.
-------
A-27
in the closed position until the slab has solidified, after which the mold is opened, the
hot slab removed, and the mold prepared for another cycle of casting.
Entry for
pressur ized
air
Grophite molds
Steel slab
Mold-clomping
mechonism
Top of pressure
tank
Ceramic tube
Molten steel
Ladle
Base of pressure
tank
A 47"2
FIGURE A-23.
INSTALLATION FOR PRESSURE CASTING A SLAB
Manufacture of Finished Products
Rolling of ingots into billets, blooms, or slabs rarely yields a defect-free product.
Consequently, additional work must be done to condition the semifinished products be-
fore they are processed further. Conditioning is done by grinding, chipping, or scarf-
ing, depending on the type of steel involved and the kinds of defects that must be re-
moved. Grinding is done with conventional abrasive grinders, chipping with hand-held
chipping hammers or special equipment known as "peelers", and scarfing can be done
with hand torches, or with automated equipment that has recently become available.
Slabs are generally scarfed automatically before they enter the hot-strip mill.
The conditioned billets, blooms, or slabs are reheated to the required rolling
temperatures in special furnaces that are fired with fuel gas. After being reheated to
the desired temperature, the billets, blooms, or slabs are transferred out of the fur-
nace and transported to the hot mills where the rolling is done.
Billets and blooms are processed into bar, wire products, and structurals of vari-
ous weights and sizes. Air-borne emissions are not a problem in the production of
these products. Iron-oxide scale is formed during the time the semifinished products
are reheated for rolling, and is broken off by high-pressure water sprays as the steel
enters the first rolling stand of the mill. This scale is collected in scale pits and sent
to a reclamation plant for recycling (usually) to the blast furnace or sinter plant.
The processes that are of interest to this project are the surface-treatment opera-
tions such as acid-pickling lines, blast descaling, tin lines, galvanizing lines, plastic-
coating lines, and other coating operations. Thes e proces s eS are carried out continu-
ously with provision for the supply of steel via a "pay-off" coil at the process input end,
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
A-28
and a similar provision for removing the finished, coiled product from the "re- COil"
end. A horizontal continuous processing line that is typical of older plants is illustrated
in Figure A-24, and the vertical type of processing line that is representative of newer
processing lines is illustrated in Figure A-25(20). Equipment designs vary with the
process, space available in the steel plant, and preferences of steel-plant personnel.
.350'
RECOIL
FIGURE A-24.
NATURE OF HORIZONTAL PROCESSING LINE
120'
PAYOFF
,;
PROCESS
" "'~'\~"'~ "'~~~~~'" "''''''''''~~~~~'''''''''~~~~~'''~~~~~~'''~~'''~~'''~'
FIGURE A-25.
NATURE OF VERTICAL PROCESSING LINE
Auxiliary Operations
Two operations that usually are not considered as process segments in the making
or iron and steel, but nevertheless incidental to the manufacture of steel, are the found-
ries and incineration facilities associated with iron and steel plants.
Foundries. The foundry installations associated with iron and steel plants are pri-
marily used as a means of supplying castings for maintenance purposes, and in this
capacity are usually under the management of the maintenance department. In the large
majority of steelworks installations, the captive foundries have modest facilities for the
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
A-29
melting of nonferrous alloys such as aluminum, brass, and bronze. These facilities are
usually well ventilated and their emissions collected. Molten steel or iron required for
large castings sometimes is obtained from the existing facilities located at the steel
plant. Iron can be obtained directly from the blast furnace. Additions of steel or alloy-
ing elements are made to obtain certain metallurgical properties. Molten steel is trans-
ported to the foundry for pouring into molds prepared for the purpose. Some steelworks
foundries have cupolas or electric furnaces for melting ferrous metals for castings.
Two major foundry items required for the production of steel mayor may not be made
within the steelmaking complex. These are (1) ingot molds required for the transforma-
tion of molten steel into a solid product for additionai processing, and (2) iron and steel
rolls necessary for rolling ingots into finished products.
Most steel companies purchase ingot molds -from foundries that are in the business
of supplying this specialized item to the steel industry. In some cases, the ingot-mold
companie s have facilities adjacent to the steel plant and obtain the iron for casting the
molds from the steel-plant blast furnaces. If a s.ingle steel plant is large enough, or if
several steel plants of one company are centrally located, and in the final case, if a steel
plant is in an isolated location with respect to available outside ingot-mold sources, the
company may have its own in-plant facilities for making ingot molds.
Iron and steel rolls used in the rolling of ingots to semifinished and finished prod-
ucts are continually replaced to maintain desirable quality standards. The manufacture
of rolls is a much more specialized operation, technically and process-wise, than the
making of ingot molds. Because of this highly specialized requirement, few steel com-
panies make their own rolls, preferring to purchase them from companies specializing
in this item.
Incineration Facilities
The making of iron and steel requires the use of many materials that are not a part
of the ironmaking and steelmaking processes, but are necessary adjuncts. These ma-
terials include (1) wood from pallets used to ship refractory brick into a steel plant,
(2) paper or plastic bags used in the shipping, storing, and handling of various required
materials, (3) paper scraps that result from the various packaging and shipping opera-
tions, and (4) various other solid-waste materials that are generated in the iron and
steel plants. Considering the large amounts of such materials that are used, the prob-
lem of disposal of solid waste is considerable in the iron and steel industry. Therefore,
incineration is commonly practiced in most steelworks.
MAJOR REFERENCES FOR APPENDIX A
(1 )
Schueneman, J. J., et al., "Ai.r Pollution Aspects of the Iron and Steel Industry",
Public Health Service, Cincinnati, Ohio, pp 10-27, June 1963. PB 168 867, U. S.
Department of Commerce, Clearinghouse for Federal Scientific and Technical
Info rmation.
(2 )
"The Making, Shaping, and Treating of Steel", Eighth Edition, 1964, United States
Steel Corporation, Pittsburgh, Pennsylvania.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
A-30
(3) "Annual Statistical Report, American Iron and Steel Institute, 1968", American
Iron and Steel Institute, New York, N. Y., 1968, p. 80.
(4) Communication from the American Iron Ore Association, September 12, 1968.
(5) "Empire Mine is Dedicated - Production Exceeds 1,200,000 Tons of Pellets",
Iron and Steel Engineer, !!. (6), 179 (June 1964).
(6) "Lime for Steelmaking: Tailored to Fit New Demands", 33/The Magazine of
Metals Producing, ~ (1), 58-93 (January 1967).
(7) Ischebeck, P., et al., "Injection of Heavy Oil Into a Blast Furnace at High Blast
Temperatures and With Fully Beneficiated Burden", Stahl und Eisen, 83 (24),
1541-1546 (November 21, 1963). . -
(8)
Bell, S. A., et aI., "Coal Injection-Bellfonte Furnace", Journal of Metals, 20 (4),
85- 88 (April 1968).
(9)
"Bessemer Converters Fading Away - Almost", American Metal Market, May 28,
1968.
(10) Kalling, B., and Johansson, F., 'IStorals Kal-do Rotary Oxygen Steelmaking
Process", Blast Furnace and Steel Plant, 45 (2), 200-203 (February 1957).
(11 )
(12 )
(13 )
(14)
(15 )
(16 )
(17)
"Electric Furnace Round-Up", 33/The Magazine of Metals Producing, ~ (6), 71-80
(June 1966).
"Everybody Is Getting Into the 'High Power' Act", The Iron Age, 202 (2), 22-23
(July 11, 1968).
Metcalf, A., "Oxy-Fuel Steelmaking for the Fuel Oxygen Scrap (FOS) Process",
Steel and Coal, 1266-1268 (December 27, 1963).
Trilli, L. J., "Oxygen-Fuel Roof Burner Design and Operation-III at Inland Steel",
Journal of Metals, ~ (9), 1059-1060 (September 1966).
"Pittsburgh Ups the Ante on BOF Scrap",. 33/ The Magazine of Metals Producing,
~ (3), 69-82 (March 1968).
Gockstetter, G. J., "Oxygen-Fuel Roof Burner Design and Operation-I At Republic
Steel", Journal of Metals, ~ (9), 1055-1057 (September 1966).
Groen, R. G., "Scrap Preheating in The Basic Oxygen Furnace at Wisconsin Steel
Works", Journal of Metals, ~ (4), 478-483 (April 1966).
(18) Jaicks, F. G., et a!., "Review of Paper on Continuous Casting of Three Types of
Low-Carbon Steel", AIME Open Hearth Proceedings, 40, 67- 84 (1957).
(19) Shah, R., "Curved Mold Lowers Silhouette of Continuous Casting Line", The Iron
Age, 192, 58-59 (August 1, 1963). --------
(20) Foreman, A. R., "New Shapes in Processing Lines", Iron and Steel Engineer, 42
(9), 181-r83 (September 1965).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
B-1
APPENDIX B
GENERAL DESCRIPTION OF
AIR-POLLUTION CONTROL EQUIPMENT
Many processes involved in the making of iron and steel create particulate and
gaseous emissions that result in air-pollution problems of varying degrees. Some of
the emissions are nothing more than simple dusts, occurring in rather small amounts,
that can be removed by equipment of somewhat simple design. Other particulate emis-
sions are more complicated in character and require more complicated equipment to
achieve desired levels of discharge to the atmosphere. Capture of different particulate
emissions generated in the making of iron and steel also requires different amounts of
energy (and involves different operating costs) to achieve acceptable dust loadings to the
atmosphere. One example of typical variations in energy requirements for various exit-
dust loadings is shown in Figure B-l. (l)'"
60
Q) 50
CI
A::J
0.0
001 40
~
c~
Q)
-
~o 30
::J ~
~o
Q) 20
~cn
Q..Q)
..s:::
u
c: 10
0 .05 .10 .15
FIGURE B-1.
Exit Dust Loading t grains per standard cubic foot
EXAMPLE OF COMPARATIVE ENERGY LEVELS TO
MAINTAIN TYPICAL EXIT-DUST LEVELS
In addition to particulate emissions, some processes in the steel industry. produce
gaseous emissions that require the application of chemical methods if they are to be
captured.
One of the outstanding methods for control of emis sions, e specially for control
of gaseous emissions, involves the installation and application of automatic cont.rols
on combustion equipment. This method attacks the problem by inhibiting the forma-
tion of some undesirable components rather than by collecting them after they are
formed.
.References cited in this appendix are listed at the end of Appendix B.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
B-2
Although the design and nature of emission-control equipment is vitally important
to the effectiveness of collection, the behavior and efficiency of even the best and most
suitable equipment in a particular situation is affected drastically by factors such as the
age and condition of the equipment, the skill and attitudes of its operator s, and the
degree of attention given to regular and sufficient maintenance.
Particulate emissions in the integrated iron and steel industry can be classified
in three forms(2):
(l) Emis sions containing particulate matter in the form of relatively
coarse particles, a substantial portion in sizes above 10 microns.'~
This includes grit from combustion processes, kilns and calciners,
grinding and screening operations, and from driers.
(2) Emissions in which the majority of the particles are between 1 micron
and 10 microns, and arise from steelmaking processes. The fine air-
borne particles found in industrial atmospheres are in this size range.
(3) The third size range of emissions includes true fumes, which are pre-
dominantly below 1 micron. They may produce industrial haze, either
alone, in combination with normal atmospheric fogs, or after interac-
tion with other contaminants in the atmosphere. Typical of these
are fumes from oxygen- blown open hearths, electric-arc furnaces,
and basic-oxgyen (BOF) furnaces.
Equipment that can be used to control air pollution in the integrated iron and steel
industry can be classified into four general groups which are: (1) cyclone dust col-
lectors, (2) electrostatic precipitators, (3) bag filters, and (4) wet scrubbers, including
spray scrubbers. (2, 3)
Cyclone Dust Collectors
The principle of operation of cyclone separators is the imposition of a centrifugal
acceleration on gas-borne particles. This is usually achieved by admitting the dust-
laden gas tangentially to the periphery of a cylindrical vessel, resulting in a spiraling
flow pattern that causes the solid particles to be thrown outwards to the wall of the
ves sel, where they fall to a conical discharge pipe. The clean-gas exit is located on
the axis of the vessel. The configuration and relative size of the gas inlet and outlet
pipes is governed by the required characteristics of the unit. High-throughput units
differ from high-efficiency units by having larger inlet and exit areas for a vessel of a
given diameter. High-efficiency and high-throughput cyclones are illustrated in
Figure B-2. In theory, small-diameter cyclones will have an efficiency~":' superior to
larger units of similar proportions. High throughput combined with high efficiency is
sometimes achieved by nesting a n.umber of small cyclones into a single unit as shown
in Figure B-3. However, there is a hazard that units containing small nested cyclones
can develop poor performance characteristics because of blocking of the small solids-
discharge pipes, which results in uneven distribution of gas.
. 1 micron = 0.001 mm.
.. Efficiency refers to the amount of particulate matter removed by the control system, expressed as a percentage of the amount
of particulate matter in the entering gas.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
B-3
+EXIT
a.
High-Efficiency Cyclone
High-Throughput Cyclone
b.
FIGURE B-2.
TYPICAL CYCLONE CONFIGURATIONS
~
DUST lLADEN GAS
FIGURE B-3.
NESTED TUBULAR CYCLONES
-""/ TUBE PLATE \
CLEANED
GAS
~
FIGURE B-4. CELLULAR CYCLONE
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
B-4
. . The efficiency of a cyclone can be increased by introducing water into the system.
ThIs IS usually done with a ring main located just below the top cover plate of the
cyclone.
Medium-efficiency cyclones (such as the cellular cyclone illustrated in Figure B-4)
are. soz:ne~imes used to reduce the load on subsequent dust-removal equipment. Swirling
achon IS mduced by passing the gas through pitched vanes.
Electrostatic Precipitators
The action of electrostatic precipitators is based on the passage of a dust-laden
gas through an intense electrostatic field between electrodes of opposite polarity. The
particles take up an electrical charge from the discharge. electrodes, and are acceler-
ated towards the grounded electrodes or collector plates, which are at opposite polarity.
The dust particles give up their charge and are deposited as a layer on the collector
plates. The layers of dust are usually removed from the collector plates by periodic
rapping that causes the dust to fall into two collecting hoppers.
Two general types of electrodes are in general use: (I) the wire-in-tube system
as shown in Figure B-S, and (2) the wire-and-plate type shown in Figure B-6. There
are many refinements of electrode configuration employed by equipment manufacturers.
A schematic illustration of a full-size wire-and-plate electrostatic precipitator is shown
in Figure B-7.
GROUNDED
~ COLLECTOR
TUBES
GAS
FLOW
GAS FLOW t
DISCHARGE
ELECTRODES
FIGURE B-S. WIRE-IN-TUBE
ELECTROSTATIC PRECIPITATOR
FIGURE B-6. WIRE-AND-PLATE
ELECTROSTATIC PRECIPITATOR
When designing electrostatic precipitators, a great deal of attention must be paid
to the quantities of dust that must be handled, and large-scale pilot-tests are often re-
quired to obtain accurate assessment of precipitator performance. Lack of attention to
such items as the resistivity and stickiness of dusts has led to serious malfunctioning
of precipitator installations, because of excessive buildup of the voltage gradient across
the collected dust. This results in local areas of intense electrical discharge and leads
to the effect known as "back ionization". This back ionization results in current re-
quirements in exces s of the capacity of the electrical equipment, which causes a drop
in voltage and results in poor collection efficiency.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
B-5
FIGURE B-7.
INTERNAL CONSTRUCTION OF WIRE-AND-PLATE
ELECTROSTATIC PRECIPITATOR
Bag Filters
The filter cloth of bag filters consists of threads about 500 microns in diameter
which are spaced 100 to 200 microns apart, thereby forming a sieve with large openings.
However, the openings are criss-crossed by fine fibers about 5 to 10 microns in diam-
eter, which are the individual fabric fibers. The fine fibers form effective impingement
targets and can remove a high portion of submicron particles. (4) A diagram of a typical
filter fabric is illustrated in Figure B-3.
During the passage of a dusty gas through the fabric, particles will impinge
upon, and be retained by, these fine fibers and cause a buildup of a layer of solid
material on the fabric. If the gas velocity through the fabric is low enough, this solid
accumulation will be in the form of a loose floc that will effectively trap even submicron-
sized particles. The progressive accumulation of solid material on the fabric eventually
leads to an excessively high pressure drop in the system, or to a local breakdown of the
filter bed (i. e., the accumulated dust). The necessity for the periodic cleaning of the
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
B-6
filter fabric results in a tendency for the efficiency to vary in a cycle-like manner, as
the floc filter bed progres sively builds up, and then cyclically is removed. The opera-
tion of multiple- bag units enables collection efficiency approaching 100 percent to be
obtained. Care must be taken not to cool the entering gases to below their dew point.
The condensation of moisture on the bags leads to rapid blinding of the fabric and
usually forces shutting down the unit for cleaning.
\
\
t
Main strands, 500
Fine fibers, 5 to 10 microns
in diameter
microns in diameter
FIGURE B-8.
DIAGRAM OF A TYPICAL FILTER FABRIC
Bag-filter fabrics must have properties that will permit them to operate in vari-
ous atmospheres and at various temperatures. Operating atmospheres and temperature
limitations for some representative bag-house fabrics are listed in Table B-l(5J. The
service life of fiberglas s bags has been extended somewhat by treating the fabric with
silicone compounds.
TABLE B-1.
OPERATING CONDITIONS FOR TYPICAL BAGHOUSE FABRICS
Fabric
General Use
Maximum
Temperature, F
Cotton
Noncorrosive or mildly alkaline dusts and gases
180
Wool
Mildly acid conditions
200-215
Nylon
Alkaline dusts or gases unsuitable for acid conditions
200-215
Dynel
More acid- resistant than wool or Orlon
200-215
Orlon
Widely used for corrOSlve gases
250-275
Terylene
Widely used for corrosive gases
275-300
Fibe rglas s
Resistant to most gases except hydrogen fluoride
500-650
Bag filters are capable of handling gases with medium to high dust concentrations.
Three general types of bag filters are generally used: (1) low-velocity filters, (2)
(2) shaker-type bag filters, and (3) the reverse-jet filters.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
B-7
Low-Velocity Bag Filter
This type of bag filter consists of multiple fabric bags suspended vertically in a
box housing as shown in Figure B-9. A relatively simple bag-shaking gear is used. A
major disadvantage of the low-velocity bag filter is that, to prevent blinding of the fabric,
gas velocities are low (about 3 feet per minute). Higher velocities tend to drive particles
into the fabric making it difficult to remove the dust by simple shaking.
~' '. DUST-LADEN
, ,
, GAS
FIGURE B-9.
LOW-VELOCITY BAG FILTER
Shaker-Type Bag Filter
The shaker-type filter is similar in configuration to the low-velocity bag filter
except that it is equipped with an automatic rapping gear which may be actuated on a
predetermined cycle, or on reaching a certain pressure drop across the filter. The
.3haker-type filter typically is able to tolerate a face velocity of about 6 feet per minute
when dust loadings are low because of the more efficient method of bag cleaning. The
more cornman face velocities for bag collectors used in the steel industry are 3 feet
per minute or less. The shaker-type bag filter is illustrated in Figure B-lO.
Reverse-Jet Filter
In order for normal bag filters to operate at high efficiency, it is necessary to
provide large areas for filtration and avoid frequent cleaning, which in turn limits the
maximum dust concentrations that can be handled. The reverse-jet filter has design
and operating characteristics intended to overcome these limitations of normal bag fil-
ters. These characteristics of reverse-jet filters are as follows:
(1) A felt of compressed wool is used as the filter fabric and has a suf-
ficiently close texture that it can act as an effective filter without the
requirement that a floc must build up to aid in dust-retention capability.
(2) The deposited dust is removed by a reverse current of air from an
external blow ring which traverses the length of the bag, usually con-
tinuously. The construction of a typical reverse-jet filter is illustrated
in Figure B-ll.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
B-8
FIL TER BAGS
FIGURE B-10,
SHAKER-TYPE BAG FILTER
1 DUST-lADEN
, GAS
INLET MANIFOLD
FEl T Fll TER
TUBES
TRA VERSING
BLOW-RING
FRAME
REVERSE JET
AIR HOSE
CLEANED GAS
FIGURE B-ll.
REVERSE-JET BAG FILTER
Limitations on operating velocities are less than for normal bag filters, because
the requirement for deposited dust to improve efficiency is eliminated. Face velocities
of 10 feet per minute are common, and in some cases the face velocities can be as high
as 15 to 20 feet per minute. Efficiencies are close to 100 percent for particles down to
about 1 micron in size.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
B-9
A modification that has been developed is called a "pulse-jet filter". Bag-cleaning
is achieved by blowing strong jets of air into the bags from the "clean" side of the fabric,
thus causing a reversal in air flow and agitation of the bags. The bags may be cleaned
in sequence so as to maintain consistent collection efficiency. However, difficulty may
be encountered in the removal of material from the fabric, and a progressive buildup of
dust may result in an increased pressure drop in the system and a requirement for
periodic removal of bags for cleaning.
Wet Scrubbers
Wet scrubbers can be divided into two categories, (1) wet-impingement scrubbers
and (2) spray scrubbers. The principle of operation in these scrubbers is that when
dust-laden gas impinges on a liquid body, the gas will be deflected around or away from
the liquid body, but the dust particles (having greater inertia) will tend to collide with
the surface of the liquid and be subject to a retaining force.
Wet-Impingement Scrubbers
Wet-impingement scrubbers are dependent on a layer of water as the entraining
medium for dust particles. Irrigated-target scrubbers, orifice-plate scrubbers, and
disintegrators fall in this category.
Irrigated-Target Scrubbers. This scrubber functions by passing a gas upward
through a flooded perforated plate so that the liquid on the top of the plate is atomized
at the edges of the orifices. This atomization creates a dust-trapping spray directed
at targets located above the orifices, as illustrated in Figure B -12. An important fea-
ture of this design is its apparent freedom from choking of the holes in the orifice plate.
Scrubbers of this type may incorporate several orifice plates in series, two or three
being usual. Each plate imposes a pressure drop of about 3 inches of water gage.
TARGET
PLATE
WATER DROPLETS
ATOMIZED AT EDGES
OF ORIFICES
FIGURE B-12.
IRRlGATED-TARGET SCRUBBER
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
B-lO
Orifice-Plate Scrubbers. Scrubbers of this type operate on essentially the same
principle as the irrigated-target scrubber, except for the omission of the target above
each orifice. Scrubbers of this type have been used in the cleaning of blast-furnace
gas. (6)
Disintegrators. Disintegrator scrubbers have found use in the past in cleaning
blast-furnace gases. (7) In scrubbers in which a spray is generated, the energy re-
quired to atomize the liquid is obtained at the expense of a pres sure drop or by pumping
the liquid through nozzles. Because the collection efficiency increases with the in-
creased relative velocity between the liquid droplet and particle, higher efficiencies
are attained by increasing the energy input to the system. In the case of the disin-
tegrator, this increased energy is obtained by passing both the dirty gas and the
scrubbing liquid into the intermeshing vanes of a stator and high-speed rotor as shown
in Figure B-l3. Disintegrators are relatively inexpensive when their high performance
is considered. However, their energy and water consumptions are high.
DUST-LADEN GAS
FIGURE B-13.
DISINTEGRA TOR
Spray Scrubbers
Spray scrubbers are those in which the scrubbing liquid is broken into a spray to
form a large number of collection sites. The efficiency of spray scrubbers is improved
by increasing the relative velocity between the spray droplets and dust particles, thus
raising the collision rate between particles and droplets. Spray towers, venturi
scrubbers, and flooded-disk scrubbers fall into this class.
Spray Towers. The spray tower (as illustrated in Figure B-l4) has become some-
what obsolete because of its relatively high cost. However, where the structures exist
as a part of original blast-furnace installation, the spray towers sometimes are used as
precoolers for the large quantities of gas involved. Spray towers have the advantage that
no very close clearances are involved, and as a result, the unit can handle relatively
high dust concentrations without suffering from choking problems. In addition, because
very fine spray is not involved, the spray generators do not require fine jets and reli-
ability is improved. Also, the spray water often can be recirculated until it contains
quite a high concentration of solids.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
B-ll
FIGURE B-14.
SPRAY TOWER
Venturi Scrubbers. Venturi scrubbers are characterized by their high collection
efficiency over a large range of particle sizes and their correspondingly high pressure
drop.
Water is admitted from jets to the throat of a constriction in the duct carrying the
dusty gas. The high gas velocity atomizes the water, and the rapid acceleration of the
droplets of water leads to a very high particle-droplet collision rate. From the venturi,
the gas is then passed to a cyclone where the agglomerated particles that are now
relatively large can be separated easily. A simple venturi scrubber is illustrated in
Figure B-15. Optimizing of the performance of high-energy scrubbers (such as venturi
scrubbers) is important, because the power consumption of such scrubbers can be quite
high. Optimizing of venturi- scrubber performance has led to the development of
scrubbers whose pressure drop and performance can be controlled while the scrubber
is operating, either to deal with varying gas rates or to maintain a given efficiency
during a particular period of operation. A variable-throat venturi scrubber that oper-
ates in this manner is illustrated in Figure B -16. A sc rubber of this type can be oper-
ated for short periods of time when the proces s requirements demand it, at a pressure
drop that would be impractical or uneconomical for continuous operation.
PARALLEL THROAT
\ \l..J WATER TO RADIAL NOZZLES
\... DUST LADEN GAS INLET
TO CYCLONE
SEPARATOR ~
.AI"
FIGURE B-15.
VENTURI SCRUBBER
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
B-12
FIGURE B-16.
VARIABLE-THROAT VENTURI SCRUBBER
Another modification of a venturi scrubber is shown in Figure B-17. Fine nozzles
and jets are omitted, and slurries can be used as the cleaning medium. The liquid flows
down the sides of the scrubber toward the venturi throat where it cascades and forms an
atomized spray to perform the entrapping function on the particles.
CLEANED
GAS
FIGURE B-17.
VENTURI SCRUBBER USING SLURRIES
Flooded-Disk Scrubber. In the flooded-disk scrubber, an atomized spray is ob-
tained by positioning a rotating disk in the path of a dusty gas and flooding the surface
of the disk with water, as shown in Figure B-lB. An atomized spray generated in this
way has good particle-collection characteristics. Also it is possible to vary the posi-
tion of the disk in the tapered throat and allow for fluctuations in gas throughput.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
B-13 and B-l4
/ TO CYCLONE
SEPARATOR
AUTOMATIC
DISC POSITIONER
FIGURE B-l8.
FLOODED-DISK SCRUBBER
MAJOR REFERENCES FOR APPENDIX B
(1)
Borenstein, M., "Air Pollution Control for the Iron and Steel Making Processes",
Industrial Heating, 34(9), 1646-1648 (September 1967).
(2)
Stairmand, C. J., "Removal of Grit, Dust, and Fume From Exhaust Gases From
Chemical Engineering Processes", Chemical Engineer, No. 194, pp. CE310-
CE324 (December 1965).
(3)
Fox, M. R., "Dust Arrestment", W. S. Atkins Bulletin, No. 10, 29-40
(Summer 1966).
(4)
Stairmand, C. J., "The Design and Performance of Modern Gas-Cleaning Equip-
ment", Journal of the Institute of Fuel, 29 (181), 58-76 (February 1956).
(5)
Squires, B. J., "Fabric Filter Dust Collectors", Chemical and Process
Engineering, 43, 156-159 (April 1962).
(6)
Lowe, J. R., "An Orifice Gas Washer", AIME Blast Furnace, Coke Oven and
Raw Materials Proceedings, ~, 28-30 (1957).
(7)
Reid, G. E., "Experience in Cleaning Blast Furnace Gas with the Orifice
Washer", Iron and Steel Engineer, ~ (8), 134-13 7 (August 1960).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-l
APPENDIX C
CHARACTERISTICS OF EMISSIONS BY THE
INTEGRATED IRON AND STEEL INDUSTRY
The manufacture of iron and steel involves many different processes until semi-
finished or finished products are available for sale or further use. Some of the pro-
cesses can produce large quanitities of particulate and gaseous emissions, while other
processes are relatively free of air-pollution problems. Sources of emissions, their
characteristics, amounts generated, and types of equipment used to control emissions
from the various processes are discussed in this Appendix. Processes are considered
in the same order as described in Appendix A; "Processes in the Integrated Iron and
Steel Industry". However, as a guide to orientation, major sources of air pollution by
the integrated iron and steel industry include coke -oven plants, sintering plants, blast-
furnace operations, steelmaking furnaces (especially those using large amounts of
oxygen for steelmaking), and boiler plants.
Preparation of Raw Materials
Raw-material preparation within integrated steel plants includes the receipt,
stocking and de-stocking, sizing, and agglomeration of iron ores, fluxes, and miscel-
laneous charge materials. For the purposes of this study, the preparation of raw
materials is considered under the following operations: (1) receiving of raw materials,
(2) preparation of iron ore, (3) making sinter, (4) making pellets, (5) preparation of
limestone and lime, and (6) making coke.
Receiving of Raw Materials
Raw materials may arrive at an integrated steel plant via road, rail, or water,
and most are unloaded to open stockpiles. In the case of water transport, boats or
barges typically are unloaded to a receiving trough and then transferred to stock by a
gantry crane. The use of long-span gantry cranes with large clam- shell buckets (known
as ore bridges) is almost universal in the United States for handling bulk raw materials
at large steel plants.
Except for scrap iron and steel, coarse materials are generally broken to a maxi-
mum size of 2 to 4 inches. There is a trend within the industry to smaller sizes, and
also a trend toward performing the crushing and sizing operations at the mines.
Materials containing appreciable fines are screened, usually to remove sizes smaller
than 3/8 or 1/2 inch. This operation is also being done more at the mine, with a
resultant product shipped in a closely held siz.e range. Emissions from raw-material
receiving operations consist mostly of dust from handling th~ ores and fluxes.
The chemical nature of iron ores and fluxes is such that particulate dust from these
materials is mainly a nuisance, rather than a health hazard. Dusts from iron ores are
iron oxides combined with oxides of silicon, calcium, and magnesium. Dusts from
fluxes such as limestone and dolomite contain calcium and magnesium carbonates,
aluminum oxide, and silica. Fluorspar (another flux used in ironmaking and steelmaking)
contains calcium fluoride, calcium carbonate, and oxides of iron, aluminum, and
silicon.
BATTELLE MEMORIAL. INSTITUTE - COLUMBUS LABORATORIES
-------
C-2
Dust arises mainly from the free -fall handling of ores during unloading, transfer'
to stock, and transfer from stock to processing or to the blast furnace. Most in-plant
handling is via the overhead clam-bucket gantry (the ore bridge), which removes ore
from the receiving area to the storage piles, and from the piles to bottom-dump trans-
fer cars operating usually on an elevated section of railway. Because of its large scale,
the operation is virtually unshroudable, and often is elevated and exposed to wind. Good
control over dust as a nuisance depends upon careful ore-bridge operation and attention
to weather. One reason why ore manipulation gives little trouble is that the huge ore
bridges must be partly immobilized (for reasons of mechanical safety) when the wind
velocity rises to about 35 knots.
Storage piles are usually accumulated between the high piers of the ore-bridge
rails, hence accumulated dust tends to stay in place even in a stiff breeze. Two areas
which normally require special attention to housekeeping are the "high-line" (where
the transfer car runs) and the dock or receiving area. These places, if allowed to
accumulate fine dust, can be sources of particulate emission during windy weather.
Some materials are worse than others in terms of their natural tendency to form
dust. In the United States, there is a strong trend to the use of mas sive pre sized lump
ores and to artificial marble-size agglomerates of iron oXide in the form of pellets.
These materials seldom contain much dust. Those ores most likely to contain
appreciable dust (such as sinter fines) are sometimes wetted deliberately to obtain
some control over dusting. Spraying piles of ore and coal with oil or with plastics and
other coatings to minimize dust problems has been cited by some as an inadequate means
for control( 1)':', while other plants have found the method helpful.
Preparation of Iron Ore
Ores and fluxes are now usually crushed and screened at the mines before shipment
to the steel plant. This is a strong general trend in the American industry, but it has
recently been demonstrated that there often is a technological advantage to re-screening
of raw materials just prior to charging them to the blast furnace(2-7). A ~umber of
blast-furnace plants have installed final screening systems. Whether these final screen-
ing systems are located in the stock yard or within the charging stockhouses, they are
considered a part of ore preparation, and are, of course, potential sources of dust
emission. However, these dusts usually are confined to the screening installations by
the use of shrouding on the screens, closed reception and conveyance of fines, and
exhaust systems where they may be appropriate.
Making Sinte r
Originally, sintering--was used as a means of recovering fines (flue dust) produced
during the making of iron in blast furnaces. However, it is now recognized that fine ore
bought at lower cost than lump ores, then combined with flux and sintered, can be con-
verted into a useful and economical burden material for blast furnaces. As shown in
in Figure C-1, the use of sinter in the blast furnace burden in the United States has
increased over the years, but in 1964 the total usage started to decrease, mostly because
of increased use of pellets.
"References for Appendix C are given at the end of Appendix C.
BATTELLE MEMORIAL 'INSTITUTE ~ COLUMBUS LABORATORIES
-------
~
o
Q)
>-
a> 50
Q.
If)
C
o
:: 40
Q)
c
'+-
o
If)
c
o
E 20
c
.Q
-
Q. 10
E
:J
If)
C
0
U
FIGURE C-l.
C-3
60
30
/--
/
/
/"
/
I
I
",,__I Pellets
.,""
--
/
'"
_J
o
1958 1959 1960 1961 1962 196319641965/966 1967 1968
Year
CONSUMPTION OF SINTER AND PELLETS IN BLAST
FURNACES IN THE UNITED STATES
The main feed to most sintering machines is iron ore that is too fine in particle
size to be used directly in the blast furnace. The fuel usually is coke "breeze" that also
is too fine to be used directly in the blast furnace, although sometimes coarser coke is
crushed down to sinter-plant size if the plant has a deficiency of fine-size coke. The
foregoing is somewhat of an oversimplification, however, because many sintering plants
in integrated steel plants are fed with a wide variety of iron-bearing and carbon-bearing
materials useless for other purposes. Among materials of this type are mill scale
from rolling operations and certain types of dusts from steelmaking furnaces. In order
to be acceptable as feed to a sintering machine, dust from a steelmaking operation must
be low in zinc content. If the charge to the steelmaking furnace contains a substantial
amount of galvanized scrap, the dust will contain zinc. If the level of zinc content in the
dust is appreciable, the dust or sinter made from it would have an adverse effect on the
refractories in a blast furnace. In summary, .sintering machines generally accept and
proces s a wide variety of feeds, differing from plant to plant and sometimes from week
to week in each plant, and produce a considerable quantity of emissions of uncertain and
variable quantity and nature.
In modern, 'high-production sintering operations, the effective control and capture
of proces s dust is all but imperative even for material maintenance demands. Abrasive
dust has, in many instances, caused a major problem in maintaining operation of the
fans in the sintering machines(8, 9,10).
Identification of Emissions. Emissions from sinter-plant operations usually con-
sist of the following: (1) minor amounts of dust in the handling and grinding of raw
materials, (2) dust that is sucked through the grate bars, (3) combustion gases from
ignition and firing, and (4) dust generated in the cooling and screening operations. The
points of emission are designated in the flow sheet for a sinter plant in Figure C-2.
Circled numbers on the flow sheet are indexes to the circled numbers in the following
discussion.
BATTELLE. MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
Sinter fines
Steelmaking
dust
Coarse ore
Fine ore
Air
C-4
Limestone
@
MIXING
DRUM
Coke
@
Blast-furnace
dust
Sinter mix
SINTER MACHINE
Hot sinter
SINTER COOLER
BLAST FURNACE
Ignition fuel
@
G)
FIGURE C-2. TYPICAL FLOW SHEET FOR A SINTERING PLANT
Circled num bers refer to emission characteristics
tagged with the same num ber in the accompanying text.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-5
Enrissions from Crushing. The ores and limestone used in the making of sinter
are naturally moist or are w.etted before and during the crushing and grinding operation.
Possible emissions are identified as follows:
@~)0
CD
Iron Ore Dust - particles are rounded to elongated in shape and can have
a size as small as 2 nricrons. Larger particles are opaque, and red-orange
in top light. Individual small grains are transparent and blood red. (11)
Hardness:
5 (Mohs)
Specific gravity:
5.2
Chemistry: Usually mostly Fe203 or Fe304; some silica and limestone,
mostly soluble in HCl
@
Limestone Dust - nrineral name calcite.- It is c;olorless, with light-
transmitting characteristics varying from transparent to translucent.
Particles generally occur as rhombohedra because of their good
cleavage. Fragments may also occur as prisms. (11)
Hardness:
Specific gravity:
Chemistry:
3
2. 7
mostly CaC03
0)
Coke Dust - Particles are opaque, irregularly shaped, quite porous and
rough with some straight, sharp edges. They are gray-black in reflected
light. ( 11 )
Chemistry: About 90 percent carbon.
Q)
Combusion Products - The gases leaving a sintering strand are a result
of the combustion of the coke in the sinter mix and of the fuels used to
ignite the sinter mix. Fuel for ignition is frequently coke-oven gas.
Because an excess of air is used during the making of sinter to provide
an oxidizing atmosphere, it is unlikely that any unburned hydrocarbons
exist in the combustion products. However, coke-oven gas does contain
sulfur compounds that combine with sulfur in the sinter-mix coke and as a
result contributes to the presence of sulfur dioxide in the combustion
products.
Sinter Dust - Dust may contain particles of iron oxides, calcite, iron-
calcium silicates, and quartz. Iron oxide can be opaque, black, rounded
particles of magnetite (Fe304) with granular faces, and/or dense, rounded
elongated, and nearly spherical agglomerates of hematite (Fe203)' Calcite
occurs as smooth, rounded particles, and quartz as a transparent, rounded
particle. The iron-calcium silicates are transparent, vitreous, colorless
to yellow to green. Particles are irregularly rounded with smooth
surfaces. (11)
@
No detailed information is yet available on the compositions of dusts going
to the atmosphere, but they undoubtedly have the same characteristics -
as above ("sinter dust"), with the exception that the particles are finer.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-6
Typical size distributions for fine, medium, and coarse dusts from sinter plants
are shown in Figure C -3. (12) The samples were taken from the wind box of a sintering
24
22
20
C 18
Q)
~ 16
Q)
a. 14
"0 12
~ 10
~ 8
0:: 6
-
-a, 4
~ 2
o
. Fine dust
~ Medium dust
o Coarse dust
16
20
70
100
Size I microns
140 200 zro 325
40
100
Equivalent U.S. Screen Series
FIGURE C-3.
SIZE DISTRIBUTIONS OF VARIOUS SINTER-PLANT DUSTS
machine. Reported results on sampling tests of a sintering machine operating in the
United States provided data on amounts of dust generated and the average size distribu-
tion of the particulates. (13) These results are given in Tables C-l and C-2.
TABLE C-l. SINTERING-MACHINE STACK-EMISSION TEST DATA
ON EXHAUST GAS DUST LOADING AT IGNITION END
Test Numbers
Conditions 1 2 3
Standard Cu. Ft. of Gas/Minute 195,500 198,800 221,700
Actual Dust Loadings
Pounds of dust/1000 pounds of gas 0.230 0.236 0.295
Grains/standard cu. ft. of gas (32 F) 0.126 0.130 0.166
Maximum Stack Gas-Emission Rates
Pounds/min 3.39 3.49 4.98
Pounds/hr 203.4 209.4 268.8
TABLE C-2. SCREEN ANALYSIS OF PARTICULATE EMISSION FROM A SINTERING MACHINE
Screen Size,
microns
Weight Retained,
percent
Cumulative Weight,
percent
5
10
20
30
44
25.1
47.6
14.6
5.8
5.0
25.1
72.7
87.3
93.1
98.1
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-7
Shown in Figure C-4 are the dust emissions, gas flow, and dust concentrations
from a sintering machine 92 feet long operating in the United Kingdom. (12)
~
e
o
'+= >.
E-5 10
co.:
~r-: 8
52 6
u+-
t;~ 4
~~ 2
001
'?O 0
~ x 15
~ ~ >.
o '~.u 10
LLEo.:
If),.,' I-' 5
o +- '
<.9......2 0
~
e
,Q 10
If)
,5Q
E 5
we
U)'E
~.......
o:e 0
FIGURE C-4.
1 3 5 7 9 II 13 15 17 19 21 23252728
FEED '
END --'--WI ndbox Leg Number
DUST-FLOW DISTRIBUTION ALONG A
92-FOOT SINTERING MACHINE
Sinter -Plant Emission-Control Equipment. There are about 48 operating sintering
plants in the United States. This total is made up approximately of 38 single-strand
machines, 7 two-strand machines, 1 three-strand machine, 1 four-strand machine, and
1 six- strand machine. Emission-control equipment varies from plant to plant and is
determined by local regulations with respect to allowable emissions, and with individual
company experience with the various types of emission-control equipment. Mechanical
precipitators, single cyclones, multiple cyclones, and, in some cases, simple dust
catchers have been used in the past to collect particulate emis sions from sintering
machines. (14) However as the need for lower allowable particulate emissions became
apparent, more efficient equipment has been installed, such as venturi scrubbers(9),
bag house s( 15), multicyclone -electrostatic precipitator combinations( 16), and a system
incorporating multicyclones, mechanical collectors, bag house, and electrostatic
precipitators. (17) Details of some of these representative systems are given in
Table C-3.
An additional practice that is used to maintain low dust levels in the handling of
sinter and sinter fines is to use water to wet the materials at the various transfer points
in the sinter-plant materials-handling system.(16, 17) Special wetting agents and
equipment to handle them are used as pad of such dust-suppression installations. This
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-8
technique appears to work well in some cases in
transfer points, and also permits more effective
control equipment by lowering the load on it.
maintaining low dust levels at the
operation of associated emission-
TABLE C-3. CHARACTERISTICS OF SOME SINTER-PLANT EMISSION -CONTROL SYSTEMS
Dust Loading,
Gas Volume, Inlet grains per cu. ft.
Company Plant Equipment cfm Temp, F Inlet Discharge
Bethlehem Bethlehem, Pa. Bag house 225,000 200-400 n.a. n.a.
Inland E. Chicago, Ind. Muiticyclone and 457,000 375 2.5 0.038
Electrostatic
McLouth Trenton, Mich. Flooded -disk 100, 000 100-150 n.a. 0.01
scrubber
United States Steel McKeesport, Pa . Venturi n.a. n. a. n. a. n.a.
Scrubber
Gary, Ind. MuIticycIone and n.a. n.a. n. a. n. a.
electrostatic
Muiticyclones n. a. n.a. n. a. n.a.
Bag house 172,000 175-300 n. a. n. a.
Note: n.a., not available.
Incorporating flux materials with the sinter to produce high-basicity (often- called
self-fluxed and super-fluxed) sinters has created some additional problems in the control
of particulate emissions at sinter plants. With reference to electrostatic precipitator
performance when operating on high-basicity sinter, the following has been stated. (17)
"The secondary gas cleaner is an electrostatic dry precipitator between the
end of the main and the induced draft fan. The entire system does an adequate
job of cleaning the exhaust gases, with routine maintenance of electrodes,
rectifiers, rappers, and wear areas, to protect the fan and to prevent a
community problem. "
"Very little more needs be said about this system, unless it would be that the
higher the sinter basicity, the less efficient the precipitator. It appears that
with super-fluxed sinter, larger precipitators may be needed to maintain a
clean discharge stack. II
The underlining has been added by the present authors. Other attempts to use electro-
static precipitators on sinter plants used to make self-fluxing sinter resulted in such
lack of precipitator stability that its use was abandoned, and efforts were directed
toward the application of bag houses and wet scrubbers. From this, it appears that
any steel company wishing to take advantage of high-basicity sinter to increase the
productivity of its blast furnace s (to remain competitive in dome stic markets and
foreign competition in those markets) will find that manufacture of high-basicity sinter
will involve air-pollution-control costs higher than thoseassociated with the manufac-
ture of conventional sinter.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
Making Pellets
C-9
The use of iron-ore pellets in blast furnaces for the making of pig iron has become
a significant mode of operation in the American integrated iron and steel industry, The
increased use of pellets from 1958 through 1967 is shown in Figure C-l. Pellet-making
plants in the United States and Canada are listed in Tables C-4 and C-S. (18) Canadian
plants are listed because many United States steel companies have interests in these
plants, In contrast to sintering plants that usually are located near the blast-furnace
plant (because sinter does not withstand shipment without degrading), pellet plants
usually are located near the are mine (or within several hundred miles of the mine),
The pellets, which are strong, are often shipped hundreds or thousands of miles to blast-
furnac e plants,
TABLE C-4. IRON ORE PELLET PLANTS IN THE UNITED STATES
Company
Location
Annual Capacity,
gross tons
Bethlehem Steel Corporation
Cornwall
Grace
The Cleveland-Cliffs Iron Company
Em pire Iron Mining Co. .
Humboldt Mining Co.
Marquette Iron Mining Co.
Marquette Iron Mining Co.
Pioneer Pellet Plant
The Hanna Mining Company
Butler Taconite
Groveland
National Steel Pellet Co.
Pilot Knob Pellet Co.
Inland Steel Company
Jackson County Iron Co.
Kaiser Steel Corporation
Eagle Mountain
Meramec Mining Company
Pea Ridge
Oglebay Norton Company
Eveleth Taconite Company
Pickands Mather & Co.
Erie Mining Company
Reserve Mining Company
E. W. Davis Works.
United States Steel Corporation
Atlantic City are
Minntac
Cornwall, Pennsylvania
Morgantown, Pennsylvania
Palmer, Michigan
Humboldt, Michigan
Eagle Mills, Michigan
Republic, Michigan
Eagle Mills, Michigan
N ashw auk, Minnesota
Iron Mountain, Michigan
Keewatin, Minnesota
lronton, Missouri
Black River Falls, Wisconsin
Eagle Mountain, California
Sullivan, Missouri
Eveleth, Minnesota
Hoyt Lakes, Minnesota
Silver Bay, Minnesota
Atlantic City, Wyoming
Mountain Iron, Minnesota
Total United States Annual Capacity
700,000
1,500,000
3,200,000
800,000
800,000
2,000,000
1,200,000
2,000,000
2,100,000
2,400,000
1, 000, 000
750,000
2,000,000
2, 000, 000
1, 600, 000
10,300,000
10,700,000
1,500,000
4,500,000
51, 050, 000
BATTELLE MEMORIAL INSTITUTE -COLUMBUS LABORATORIES
-------
C-IO
TABLE C- 5. IRON -ORE PELLET PLANTS IN CANADA
Company
Location
Annual Capacity
gross tons
Bethlehem Steel Corporation
Marmoraton Mining Co., Ltd.
Marmora, Ontario
450,000
The Cleveland-Cliffs Iron Company
Sherman Mine
Timagami, Ontario
I, 000, 000
The Hanna Mining Company
Carol Pellet Company
National Steel Cor. of Canada
Labrador City, Newfoundland
Capreol, Ontario
10,000,000
625,000
Inland Steel Company
Caland Ore Co.
Atikokan, Ontario
I, 000, 000
The International Nickel Company
of Canada, Limited
Sudbury Mine
Copper Cliff, Ontario
1,170,000
Jones & Laughlin Steel Corporation
Adams Mine
Kirkland Lake, Ontario
1,250, 000
Pickands Mather & Co.
Hilton Mines, Ltd.
Wabush Mines
Griffith Mine
Shawville, Quebec
Point Noire, Quebec
Iron Bay, Ontario
900, 000
6,000,000
I, 500, 000
Steep Rock Iron Mines Limited
Steep Rock Mine
Steep Rock Lake, Ontario
Total Canadian Annual Capacity
1,350,000
25,245,000
Identification of Emissions. Materials received at pellet plants include iron-ore
concentrates and a binder material, usually bentonite. The concentrates are received
usually in a moist condition, and the generation of dust during receiving is not generally
considered a problem. Bentonite is received in covered hopper cars and unloaded -into
special bins for metering into the pelletizing operation. The pelletizing operation is
conducted with some moisture in the ore-binder mixture to achieve the necessary
pellet formation in rotating drums or on rotating disks. The final stage in pelletizing
of the concentrates is indurating (heat hardening) to produce the required hardness that
will permit subsequent handling of the pellets during shipment and charging to a blast
furnace.
The concentrates as delivered to the pellet plants are magnetite, except for two
plants that are operating with hematite concentrates. During the indurating treatment
of the magnetite pellets, the magnetite is transformed (by oxidation) to hematite. When
particulate emi s sions are generated to any extent in the production of pellets, the
particulates are magnetite, hematite, or bentonite. The se particulates have the follow-
ing characteristics:
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-11
Magnetite Dust. All partiCles of this isometric mineral are opaque,
even those with dimensions less than 1 micron. The crystals fracture
unevenly and rarely show cleavage. The crushed fragments are attracted
to one another and cohere in a fluid mount because they are magnetized. (11)
Chemistry: mainly Fe304, with some gangue, mostly silica.
Hematite Dust. Particles are rounded to elongated and can be as small as
2 microns. Larger particles are opaque and red-orange in top light. Individual
small grains are transparent and blood-red. (11)
Hardnes s :
Specific gravity:
Chemistry:
5 (Mohs)
5.2
Mostly Fe203,
mostly silica.
Soluble in HCl.
Minor amount of gangue,
Bentonite. Transparent, colorless, apparently rounded particles that can
occur as small agglomerates. Because the particles are so small, their
agglomerates look almost fibrous. (11) Bentonite is a fine-grained clay,
usually of the montmorillonite type.
Pellet-Plant Emission-Control Equipment. The minor amounts of dust generated
in pelletizing plants are usually handled by simple cyclones. The indurating operations
are conducted under rather low air-flow rates. High air-flow rates, such as encountered
in sinter plants or blast furnaces, are not generally present, thus in-process formation
of particulate emissions is usually not substantial.
Preparation of Limestone,
Dolomite, and Lime
The use of limestone (mostly CaC03), dolomite (mostly CaC03' MgC03)' and
lime (mostly CaO) in the making of iron and steel is necessary to remove undesirable
elements from the iron ores (mainly silica, alumina, and sulfur, in the case of the blast
furnace) and from pig iron (mainly sulfur, phosphorus, and silicon in the case of steel-
making). The amounts of limestone ':' and lime consumed in the American integrated iron
and steel industry are shown in Figure C-5. (8) The rapid increase in lime consumption
from 1962 to date has been due to its use in BOF steelmaking. Limestone consumption
is influenced by the total production of pig iron as well as by improved blast-furnace
technology, which has resulted in a decrease in the amount of limestone consumed in
the production of each ton of pig. The trend in limestone consumption per net ton of pig
iron from 1958 through 1967 is shown in Figure C-6. (8)
In BOF steelmaking, lime is used instead of limestone because of the faster reac-
tion rates obtainable with the lime, and because the use of lime conserves heat in the
..... OF furnace. It has been determined that the characteristics of the lime itself have an
important effect on the reaction rates. (19,20) A soft lime, which is calcined at a lower
temperature and for a shorter time than hard lime, is the preferred material. Both
types have the same chemical composition, but the soft lime is less dense. (20)
Increasing demands for lime for BOF steelmaking have necessitated an increase in
calcining facilities.
*In the U. S. steel industry, statistics on "limestone" include dolomite.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
If)
c 30
o
+-
+-
Q)
C 25
-
o
If)
c
.2 20
- 5
E
c 4
o
+-
a.
E 3
:J
If)
c 2
o
u
0
:J
C
c
-------
C-13
Estimates have been made of the lime production that is captive to the integrated
iron and steel industry. (21) These estimates since 1965 are given in Table C-6. It
also has been estimated that by 1970 the captive lime production will be less than
20 percent of the 5.6 million net tons required annually by the steel industry.
TABLE C-6.
ESTIMATED LIME PRODUCTION
CAPTIVE TO THE INTEGRATED
IRON AND STEEL INDUSTRY IN
THE UNITED STATES
Year.
Net Tons
1965
1966
1967
1968
1970
158, 000
358, 000
600, 000
860,000
1,000,000
Identification of Emissions. Compositions of particulate emissions that may be
generated during the preparation of lime stone, dolomite, and lime are characteristic of
the minerals. Compositions vary from about 95 percent calcium carbonate and less
than 1 percent magnesium carbonate in high-quality limestones, to about 55 percent
calcium carbonate and 40 percent magnesium carbonate in dolomites. Dusts produced
during crushing operations are the main particulates generated during the preparation
of limestone and dolomites. However, little is actually emitted to the atmosphere.
Primary crushing is usually done with the minerals dry or at their natural moisture
content, and during this stage of preparation the generation of air -borne particulates is
small. In the stages of finer crushing, where dusts could be a problem, processing
usually is carried out wet. Quoting from a recent report on a new plant for the prepara-
tion of flux and BOF lime, the following is descriptive of these operations. (22)
11 Bin No. 2 provides 1, 100 tons of minus 4-inch feed to the second
principal proces sing section. This' section consists of four parallel identical
circuits, each having a 48-inch x 72-inch vibrating feeder, a single-deck
5-foot x 14-foot screen, an 8-foot x 20-foot rotary scrubber; two 5-foot x
14 -foot double-deck screens and a 460 Hydrocone tertiary crusher. 11
11 The feeders draw stone from the bin and distribute it to the single-
deck screens. This circuit is entirely wet processing, with all screens
equipped with spray bars. . ."
(Underlining has been added for the purposes of this present report. )
Emissions from lime kilns can include lime dust generated at the discharge end of
the kilns, carbon dioxide from the calcining of the limestone, and the products of com-
bustion from the fuels used to heat the kiln. Fuels used are usually IQ.w-sulfur fuel oils
and natural gas. Because lime is used to accolnplish desulfurizatton during the steel-
making process, it would be to the disadvantage of lime producers to use fuels that
would result in increased sulfur content in the lime.
Emissions from limestone preparation and calcination of limestone can be
identified as follows:
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-14
( 1)
Limestone Dust - Mineral name calcite. Colorless, with light-transmitting
characteristics varying from transparent to translucent. Particles
generally occur as rhombohedra because of their perfect rhombohedral
cleavage. Fragments may also occur as prisms. (11)
Hardness:
Specific gravity:
Chemistry:
3 (Mohs)
2. 7
Mostly CaC03
(2)
Dolomite Dust - Mineral name dolomite. In the pure state, dolomites
are colorless and the fine particles are generally transparent to translucent.
Electron-microscope studies of finely ground dolomite have shown the
presence of few good cleavage planes and an absence of perfect rhombs. (23)
Specific gravity: 2.85
Chemistry: Mainly CaC03' MgC03 in various combinations.
(3)
Lime Dust - Lime is usually white in color of varying intensities, but
some may have a light cream, buff, or gray cast depending on the
nature of the impurities in the lime. (24) . '
Specific gravity: 3.3
Chemistry: Mainly CaO
Lime-Preparation Emission-Control Equipment. As stated earlier in this Appendix,
the generation of air -borne emissions during the preparation of limestone and dolomite
is not a substantial problem because much of the processing is carried out wet. This is
not the case in the making of lime. .
Several types of equipment are used to collect dust from the discharge and cooling
locations of calciners. A scrubbing tower is reported to remove all dust and smoke
from a commercial lime plant operating with a rotary kiln. (25) Cyclones are used with
some circular-hearth calciners(26), and electrostatic precipitators are included in a
new lime plant placed into ope,ration by the Republic Steel Corporation in 1968. (27) A
recent report on emission control equipment placed into operation in 1968 at the Pueblo,
Colorado, plant of CF & I Steel Corporation, describes the use of multicyclones and a
bag house to recover dust from a vertical lime kiln. (28) No information has been found
yet in the published literature pertaining to the amounts or sizes of particulates that
may escape to the atmosphere from lime-burning equipment.
Making Coke
Metallurgical coke is the major fuel and reducing agent used in the production of
blast-furnace hot metal, and will probably be the major fuel and reductant for many
years in the future. Technological developments in the making of hot metal have
resulted in a decrease in the amount of coke ne'eded to make 1 net ton of hot metal.
(The amount of coke required to make 1 net ton of hot metal is referred to as the "coke
rate".) Even though the coke rate has decreased, the'total consumption of coke has
increased because of the increased production of hot metal.'. The trends in coke rate
and in total coke consumption in American blast furnaces from 1958 through 1967 are
shown in Figure C-7. (8) Much of the lowering of coke rates has been the result of
BATTELLE MEMORIAL INSTITUTE -COLUMBUS LABORATORIES
-------
C-15
improving the composition and physical form of the iron ore. The decrease in coke rate
that is possible with sinter and pellets in the blast-furnace burden is illustrated in
Figure C-8. (29) It is possible to reduce the coke rate to about 840 pounds (ideally) per
net ton of hot metal by the use of improved burden materials, injection of auxiliary fuels
(such as oil and natural gas), and the use of very high blast temperatures. (30, 31)
1600
Q)
..:JC.
o
<.)
-
o
1550
1500
IJ)
1:J
C
:J
o
Q.C
~O
Q)'~ 1400
+-en
0.-
0::Q.
Q) 15 1350
..:JC.
o c
U.8 1300
Q)+-
enQ)
e c 1250
Q)...
>Q)
<[Q.
1450
/1\
/ \
/ \
\
/-""\
/ \
I \
I--
I Total coke
I consumed
annually
I
I
/
/
62 U>
a
60 CD
c
58 ~
OIJ)
:J C
co
56 C +-
<[Ci)
1:JC
54 Q)-
EO
:JIJ)
52 g? 5
0=
u.-
50 Q)E
..:JC. ~
o~
48 U g
oc
+- ...
~~
1200 46
1958 1959 19601961 1962196319641965 196619671968
Year
TRENDS IN THE CONSUMPTION OF COKE AND IN COKE RATE
IN THE PRODUCTION OF HOT METAL IN AMERICAN BLAST
FURNACES
FIGURE C-7.
/700
...
Q)
Q.
IJ)
1:J
C
:J
o
Q.
~O
~ Q) 1500
°E
0::
O)+-
C5 2 1400
U-
Q)0
enC
e.8 1300
Q)+-
>Q)
<[C
1600
1200
o
All pellets
1000 2000 3000
Sinter and Pellets Charged to Blast Furnace,
pounds per net ton of hot metal
FIGURE C-8.
EFFECT OF AMOUNT OF SINTER AND PELLETS IN
BLAST FURNACE BURDEN ON COKE RATE
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LApORATORIES
-------
C-16
The manufacture of coke as discussed in this Appendix is concerned with by-product
coke ovens. Beehive coke is still made to a limited extent. However, during the last
10 years, production of beehive coke has not exceeded 1 percent of the total coke pro-
duction in the American integrated iron and steel industry. From 1965 through 1967,
this percentage has decreased to O. 7 percent. A beehive coke oven is difficult to adapt
to air -pollution control, and in all probability the industry will phase the beehive coke
ovens out of operation as rapidly as the required additional capacity can be made
a vailable through the construction of by-product coke ovens.
By-product coking is a process combining agglomeration and thermochemical
conversion. Bituminous coals that are too finely sized and too friable for direct use as
the major blast-furnace fuel are blended and baked at high temperature. The baking
process decomposes the long-chain organic polymers of the coal. Aromatic tars and
oils are vaporized and driven off along with various free radicals, and the carbon resi-
due reforms in massive chunks of strong coke suitable for use in smelting. Fuel gas,
tar, aromatic oils, and ammonia may be recovered as by-products from the crude gas.
Conventional coke plants emit both particulate materials and offensive gases in
the normal course of operation. (32) The particulates are mainly coal dust and coke
dust that become airborne during handling. The gases are mainly mixtures of ammonia
and aromatic vapors escaping from the ovens and from sumps, seals, and vents in the
by-product system. Because these emissions are considered to be particularly unplea-
sant, coke plants deserve very serious attention in planning for environmental control.
Identification of Emis sions. Emis sions are generated in several locations and
operations during the manufacture of coke and during the processing of coke by-products.
These are discussed under the appropriate headings as follows: (1) coal handling; (2) oven
charging; (3) oven operation, pushing and quenching; (4) coke handling; and (5) by-product
processing. The points of emission are designated in the flow sheet for a by-product
coke plant shown in Figure C-9. The circled numbers on the flow sheet are indexes to
similar numbers in the following text where emissions are discussed.
Coal Handling. Coking coals are received usually by hopper car, river barge,
or lake freighter. Because these coals are highly beneficiated at the mines, they
average a much finer size than steam coals. Shipment of high-grade coking coals at a
top size of 1/2 inch is a common practice. For this reason, coking coals are dusty
(unless wet), and each transfer point in the unloading and handling steps is a potential
emission site.
Although coking coals are often wetted before shipment, they are seldom wetted
upon receipt. The problem is that fine coal is hard to wet uniformly, and nonuniform
moisture interferes with both handling and blending. Overwetted coal will stick to
belts and foul the transfer points, while underwetted coal continues to dust. Blending is
based on dry weights and presumes uniformity of moisture in each individual coal used.
Fuel oil is sometimes sprayed on coal just prior to pulverizing or blending. The purpose
of the oil is to control bulk density of the blend - it also controls dusting.
Typical coal-handling operations include unloading, transfer to and from stock,
pulverizing, blending, and transfer to the coke ovens. Most handling is with rubber
belts for upward movement and by gravity chute for downward movement. There are
BATTELLE" MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-17
Cleaned Coal
3100 pounds
Air
@@@
(OO)@@
COKE OVENS
3085 ounds
Hot Coke
2220 pounds
@
Other plant use
Water +
Spent Liquor
FIGURE C-9. TYPICAL FLOW SHEET FOR A BY -PRODUCT COKE PLANT
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-18
many transfer points in most conveyor systems, and a typical particle may free-fall
as many as twenty to thirty times before it arrives at the coke ovens. Each occasion of
free-fall provides opportunity for fines to escape as dust. However, coking-coal trans-
fer points can be controlled by utilizing water-spray systems, plus detergent. A re-
cent report on an investigation of settled particles at a coke oven plant contains the
following statement(33): -
"On the basis of the data obtained during a 6-month study near a coke-making
operation it is concluded that material handling and stockpiling operations are
major contributors to settled particulate deposition, while coke oven charging
was not a major source. "
The extent of the coal-storage problem can be visualized by realizing that as much
as 3 million tons of coal is stored at times in one storage area. To some degree, the
problem of fugitive dusting can be minimized by limiting storage of coal to smaller
quantities. Deliveries of coal by unit trains directly from mine to coke-oven area lowers
the amount of coal that must be stored within the steel plant.
Oven Charging. The last step in handling of the coal is charging of the blended
coal fines (now pulverized to a top size of 1/8 inch) through ports in the top of the coke
oven. The oven itself is a slot-like retort about 10 to 18 feet in height, 30 to 60 feet in
length, and 15 to 20 inches in internal width. The sidewalls of the oven are at an
incandescent temperature at the time of charging. There are usually 4 or 5 coal-charging
ports, each about 10 to 14 inches in diameter. The charging vehicle is called a larry
car. It receives a weighed charge into separate hoppers (one for each port) and dis-
charges from the hopper bottoms.
The oven slot is vented to the by-product system at one or both ends via vertical
ascension pipes leading to a manifold called the collector main. Coke ovens are oper-
ated under slight negative draft in this main, but the oven pre ssure is quickly equalized
when the charging ports are opened. Modern plants have steam jets in the raw-gas
ascension pipes so that the draft on the opened oven can be increased during charging.
Even so, the charging operation is characterized by particulate and gaseous emissions.(34)
When the cold, damp coal falls against the hot walls of the coke oven, the moisture
is quickly changed to steam, and the breakdown of polymeric materials in the heated
coal proceeds very rapidly. The volume of the mixture of steam and raw-coal gases far
exceeds the capacity of the aspirated ascension pipe; and so it rushes out of the charging
ports as smoke. The escaping coal gas is usually ignited by incandescent coal fines,
which are also blown out of the oven by the reaction. Of course, this rush of gases
interferes with charging, and some fines are blown out of the descending coal. At the
end of charging, the steam, flame, and smoke puff upward through the ports of the larry-
car hoppers. .
A new development is being used in Europe to minin.ize emis sions during oven
charging. The larry car mechanically remeve-s the oven lid, drops coal through a
sleeve into the oven, collects the emissions produced during .charging, and treats them
in a wet scrubber. Adaptation of this system to existing American ovens presents major
problems because the oven structure usually would not necessarily be strong enough to
support the additional weight. Also, the larry car usually would be so high in the new
system that the car would not fit under existing coal bunkers. The new system, if it
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-19
works out consistently in practice has potential, for use on new American batteries.
There is some hope that work being done in the United States on pneumatic charging
may be succes sful enough to warrant inclusion in future American practice.
Oven Operation, Pushing, and Quenching. Coke ovens are heated by complex flue
systems between the slot retorts. The fuel is usually coke -oven gas or blast-furnace
gas, depending on the energy balance within the steel plant. Combustion in coke-oven
flues is accomplished with an excess of preheated air in such a way as to preclude the
presence of unburned fuel in the products of combustion. Accordingly, the emissions
from the coke-oven exhaust stacks are quite ordinary, unless coal gas from the retorts
leaks into the flues. This sometimes occurs in older coke plants.
After an oven is charged, the release of raw-coal gases continues at a high rate
for a considerable period of time, at least until the entire charge has been dehydrated
and steam emission stops. During this time, the oven is under negative pressure, but
localized pressures are generated within the coking bed and raw gas may leak out
around the ports and the end doors. This emission is normally a mere wisp when com-
pared to the gas released during charging, and no direct attempts have been made to
suppress or collect it. It is to the best interests of the operator to minimize door leak-
. age, because the cracks that allow gas to leak out in the first part of the cycle will allow
air to leak in (to consume coke) in the later stages. In any particular coke-oven bat-
tery, the amount of leakage is generally related to the age of the ovens, the level of
maintenance applied, and the skill and motivation of the operating crew. The idea of
collecting oven emissions by building an enclosure over the whole battery represents a
prodigous undertaking, but has the potential for great improvement of emission control.
During the actual coking operation (which typically requires 16 hours), usually
little gas or particulate matter is evolved into the atmosphere. Some escapes because
of ineffective sealing of the charging ports and end doors. When an oven is pushed
(emptied) at the completion of the coking cycle, the end doors are removed and a
mechanical pushrod is forced all the way through the oven to eject the coke. The open-
air release of 10 to 15 tons of incandescent coke produces a considerable induced draft
in the immediate surroundings, and fine particulates are blown high into the atmosphere.
However, at this point. the coke is both strong and massive, and only two kinds of fines
are likely to be present. Some of the air-borne particulates may be those that have
settled in the area because of prior coal handling, but the rest comes from incompletely
coked coal (adjacent to the cool coke -~ven doors) and from abrasion of the coke as it is
pushed from the oven. Particulate emissions are not great during the pushing operation
unless the amount of uncoked coal is considerable, under which circumstances coal
smoke may also be released for a brief period. With proper coking cycles and ove1I1-
heating practices, emissions generated from pushing "green coke" (coke with incom-
pletely coked coal) can be minimal.
The pushed coke is received into an open hopper car of special design, with a
sloped bottom and side gates made of grating. This car may be self-propelled or moved
by locomotive to a large brick chimney that fits over the open top of the car. Sprays in
the chimney deluge the hot coke with water to cool and quench it. During the quenching
operation, particulates are lifted into the atmosphere by the chimney effect created by
the rising cloud of steam. The particulates tend to fall out locally in the vicinity of the
quench tower, and usually are not carried great distances as a suspended dust plume.
The quenching time is about 2 minutes (depending on the practice in any particular plant),
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-20
and 87 percent of the particulate emission from quenching occurs during the first min-
ute of the quenching cycle, as shown in Figure C.,...lO. (35)
100
'+-
o
-c
Co
B 0 in 80
... I/)
If °E
_w' 60
'-g,~
°w.Q
3: B 40
Q) '';::
> ...
'+= 0
o a.. ° 20
:J-
E.E
8~
o
o
20 '40 60 80 100
Quenching Time, seconds
120
FIGURE C-lO °
PARTICULATE EMISSION DURING COKE QUENCHING
Coke Handling. The quenched coke is dumped from the quench car onto a sloping
brick wharf from which it is carried on conveyors to sizing and screening operations.
Dusting is not so much of a problem in the handling of coke as it is with coal handling.
Usually only a small amount of dust is generated in the handling operations. Transfer
points, screens, and crushing machinery are protected by steel and rubber shrouding to
minimize dust losses and reduce housekeeping maintenance. Also, the handling equip-
ment often is enclosed by galleries and buildings. Aside from these factors, the coke
itself is a massive and stable material, and fines generated during crushing and handling
are removed before the mainbody of coke is transferred to stock or to the blast furnace.
The fines usually are dumped into open hopper cars via chutes from the screening
plant. The fines have a higher moisture content than the massive, coke and typically
contain over 8 percent moisture. Dusting of this "breeze" coke usually is not a serious
problem except in windy weather. The small amount of volatile organic material
remaining in the coke is stable and is known to persist to temperatures of 2300 F or
higher within the blast furnace. Therefore, it is not emitted as an air contaminant
during the storage of coke.
By":Product Processing. The mixed gases driven off of the coking coals during
conversion to coke enter the by-product system via the ascension 'pipes at one or both
ends of each retort oven. These ducts lead into a collector main (or two) and are
maintained under negative draft by exhauster fans far downstream in the system.
The mixed gases contain organic (usually aromatic) compounds released by the
destruction of long-chain polymers in the coal,plus carbon monoxide, hydrogen,
methane, hydrogen sulfide, ammonia, and nitrogen from tramp air drawn into and
through the ovens. Among the prominent organic groups one may number anthracene
and other tarry compounds, benzene, toluene, xylene, naphthalene, phenols, and pitch.
The purpose of by-product processing is to condense, separate, absorb, distill, clarify,
and otherwise segment and recover the valuable substances in the raw gas. With the
BATTELLE MEMORIAL INSTI.TUTE - COLUMBUS LABORATORIES
-------
C-21
advent of high-volume petrochemical processing in the 1940's, the market prices of
many organic chemicals were lowered drastically, and by-product recovery from coke-
oven gas is not now,so economically attractive to steel companies as in the past,
especially in the case of small-capacity coke-oven plants. However, most American
coke plants have by-product systems, and these are still used to obtain credits to the
coking process. A flow sheet for a typical by-product plant is given in Figure C-ll.
Raw-coal gas (which includes steam) rises into the coke-oven ascension pipes and
passes into the collector main. Some ovens have mains and ascension pipes at both
ends of the ovens. The beginning of cooling and condensation is forced almo st
immediately by sprays of recycled water in the ascension pipes and along the mains, but
the final temperature at the end of the main is still well above the boiling point of water.
The mechanism of cooling is the evaporation of the sprayed water (called flushing
liquor); thus the raw gas is consider"ably diluted by steam. Shortly after emerging from
the collector mains, the gas temperature drops below 212 F and water condense s
rapidly from it.
With the condensation of water, droplets of tar are also condensed. This process
is hastened in primary coolers cooled normally with river water, and the mixed
tar and condensed liquor are separated in decanters. Some of the liquor is recycled to
the flushing-cooling system, and some is drawn off for recovery of ammonia values by
distillation and absorption in sulfuric acid. The partly cleaned gas goes through the
exhauster fans into the rest of the by-product system.
Emissions from the primary end of the system usually are minor in amount be-
cause the system is under negative pressure. There is some odor indicative of free
vapor at the tar collectors and decanters and wherever the liquor runs in lines that are
not fully enclosed. In particular, ammonia and organic fumes are strong at the sumps
where decanted liquor and other flush liquor is collected for recycling to the collector-
mains sprays., It is arguable that the flush liquor should be handled in closed, unvented
ducts. To minimize local nuisance, the flushing liquor sewers are usually fairly well
capped and covered, but there often is no sealing because these ducts become fouled
with tar and other "goop". Access for steam cleaning is essential, but is not always
used frequently.
Downstream of the exhausters, the detarred gas is still quite rich in ammonia,
both free and in combination. The liquor decanted from the tar at primary cooling is
also quite rich, and because of the moisture vaporized from wet coal this liquor exceeds
the requirements of the flushing-liquor system. Both the excess liquor and the gas are
stripped of ammonia, which is recovered usually as ammonium sulfate. (America.! s
largest coke plant is big enough to recover the ammonia on a commercial basis as
anhydrous ammonia), In the conventional sulfate process, the gas is reheated (to hold its
moisture in the vapor phase) and passed through sulfuric acid where ammonium sulfate
is precipitated. Ammonia vapor distilled from the surplus flushing liquor is also passed
through this precipitator. The residue after distillation, known as weak liquor, may be
processed further or disposed of via sewers or to the coke quench. The residue still
contains ammonia plus a number of soluble organics such as phenols.
Emission to the air from the ammonia system generally is quite minor, because
the ducts are closed piping for the most part, and leaks usually are promptly detected
and repaired. One activity that can cause trouble in the older plants is the addition of
strong sulfuric acid to the precipitator tank, an operation usually accompanied by
considerable fuming of the acid.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-22
COKE OVENS
Ammonia
Further Processing.
SCRUBBER.
Steel Plant Use
FIGURE C -11. BY -PRODUCT PLANT FLOW SHEET
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
G-23
In large coke plants, the ammonia-recovery system may be augmented by systems
for recovery of phenol (carbolic acid) from the weak liquor after ammonia distillation
and for recovery of pyridine bases which dissolve in the ammonium sulfate precipitator
tank from the gas. In smaller coke plants, one or both of these activities may be by-
passed where volume does not justify capital investment for a recovery system. Phenol
and pyridine systems are closed except for tank vents, and usually have no substantial
problems with regard to emission to the atmosphere.
The crude tar collected in the collector mains and in the tar decanters is aug-
mented by minor amounts of tar precipitating at the secondary coolers or other points in
the gas system. From this crude tar, pitch sludge settles to the bottom of the decanter
and is mechanically raked out for disposal. Settled crude tar is sent to a separate plant
for secondary processing by distillation.
Conventional tar proces sing yields pitch tar, creosote, and two or more weights
of tar oil that may be used in roadbuilding. Crystals of naphthalene are a by-product
(often discarded), and in larger plants the tar oils may be further refined to yield
salable fractions.
Tars are heavily loaded with polycyclic aromatic hydrocarbons (PAH), and are
generally considered hazardous. Steelmakers have recently taken steps to lead tar-
processing tank vents and storage tank vents through scrubbers that absorb or destroy
the fumes. The work has not met with great success yet because tars tend to condense
upon and foul the equipment. When the scrubbers get fouled, they will discharge
unscrubbed vent vapor.
Gas leaving the ammonia precipitation tank is still warm and contains a number of
light aromatics boiling between 200 and 400 F. These light oils are condensed from
vapor by cooling the gas to about ambient temperature in a final cooler. The condensate
is then scrubbed with a high-boiling wash oil (derived from petroleum) to dissolve the
aromatics, and the wash oil may be steam distilled to recover the principal aromatics
benzene, toluene, and xylene in commercial form. The gas is now completely stripped
and enters the plant fuel system.
The vapors of the light oils are considered hazardous both because of toxicity and
because of flammability. Nevertheless, the condensers in the distilling system and
some of the process tanks are vented. The vapors issuing from these vents often pervade
a large area with the sweet, almost pleasant smell characteristic of aromatic vapors.
Abnormalities (such as fire or major leakage) are rare in by-product systems
because the high hazard level prompts strong preventative measures. However, in the
ordinary course of pumping, straining, dewatering, and otherwise treating coke -oven by-
products, leakage and vapor loss is inevitable. There may be as many as 40 pumps in
the liquor and stitl systems; usually some of these have imperfect seals. Forced
ventilation in the pumphouse is a necessity.
Identification of particulate emissions from coke-plant operations has been limited
in the past, but as a result of research carried out by the integrated iron and steel
industry since 1950(36-43), methods have been developed that permit the identification
of particulate emissions generated in coke-plant operation. While this research has
been directed primarily toward obtaining a better understanding of the coking proces s,
and toward improvements in the properties of coke required by the advancing blast-
furnace technology, it has also provided a means of identifying air-borne particulates.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-24
Optical reflectance has been found to be the best method for identifying different types
of coal and their particulates. Figures C-l2 and C-13 illustrate some relationships
between reflectance, ultimate carbon in the coal, and volatile matter in the coal(36)j the
latter two materials are an indication of the "rank" or classification of the coals.
100
.
.. .
-
c:
~ 40
~
Q)
a.
..: 30
Q)
::.
~ 20
-
c:
~ 90
~
Q)
a.
c: 80
° .
.Q
~
8 70
~~.~ .
,
./
-
I
.
Q)
-
o
.S 60
-
::J
50
~
-
.2
g
50
o
1.0 2.0 3.0
Reflectance (Ro)' percent
10
o
o
1.0 2.0 3.0
Reflectance (Ro)' percent
FIGURE C-13.
RELA TION OF
REFLECTANCE TO
ULTIMATE CARBON
FIGURE C-l2..
RELA TION OF
REFLECTANCE TO
VOLA TILE MATTER
The research on reflectance properties of coals was extended to coke, and a relation-
ship was established between the reflectance of coal and the reflectance of the coke wall
(referring to the cell structure of coke) in the coke made from,that coal. This relation-
ship is illustrated in Figure C-14. (40) For such relationship to be valid, the final
coking temperature must be known.
c: 1.5
Q)
u
Cii 1.4
a.
rr.0 1.3
Q)
u
c:
o
-
u
~
.....
Q)
rr.
o
o
u 0.9
0.8
0.7
6
1.7
14
FIGURE C-l4. RELATION OF COAL REFLECTANCE TO
COKE-WALL REFLECTANCE
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
1.6
1.2
1.1
1.0
7 8 9 10 11 12 13
Coke Reflectance,Ro' percent
-------
C"725
The techniques developed extensively for the petrographic examination of coal
and coal have been applied to the identification of particulate emissions in a coke
plant. (33) Particulate emissions generated in a by-product coke plant are identified
as follows:':'
CD
Coal dust - Bituminous, or soft coal, is translucent in thin areas; it is
reddish-brown by transmitted light, and brownish-black with dull to mod-
erately high reflectivity in reflected light. The surfaces are slightly rough
with occasional indications of the original fibrous structure. These irreg-
ular chips have sharp edges, and in places show conchoidal surface
fractures. (11)
@
Coke balls - Identified as oval in shape with an unusual network-like internal
structure. It is suggested that coke balls are produced during the thermal-
drying stages of coal proces sing and are inherent in products leaving coal-
processing plants. Similar conditions occur during charging of by-product
coke ovens, where some coal fines are carried through the hot zone and
out adjacent, open charging holes. (33)
Q)
Char - Partially devolatilized coal particles that exhibit optical properties
between those of coal and coke. The partial devolatilization of coal particles
suggests that they have not been subjected to temperatures high enough
or for periods long enough to complete the coking process. (33)
o
Pyrolytic carbon - The tarry residue from the volatile organic portion of
coal. Two forms of pyrolytic carbon are identified. The first is an
aggregate of minute oval grains; each grain is relatively uniform in size,
extremely smooth in appearance, and exhibits extreme anisotropy in
polarized light. The second normally occurs as a crenulated band of
varying width and length, smooth in appearance, and strongly anisotropic
in polarized light. The size of these materials is extremely variable. (33)
@
By-product coke - The optical characteristics of particles of by-product
coke are controlled by the rank (reflectance) of the coal which is
carbonized (as illustrated in Figure C-14). Because coals of different
rank are usually blended to make an optimum mix, particles of coke
produced from these mixes have complex and highly variable optical
properties. Particulates of coke made from high-volatile and medium-
volatile coals may be granular in appearance, have thick coke walls, and
have few internal pores. Particulates from coke made with low-volatile
coals have distinctive ribbon-like graphitic textures, have thin coke walls,
and comparatively large internal pores. (33)
Results of a 6-month study have shown that the origin of particulates in a coke
plant can be divided on a weight basis as follows: 40 percent from coke, 30 percent
from coal, and 30 percent f~0m other sources, where other sources are classified as
road dust and other mineral dusts normally found in a steel plant. Screen analyses of
the particulates, discus sed in the study, found in the vicinity of one coke plant are
listed in Table C-7. (33) Screen analyses of particulates generated at one coke-plant
quench tower are given in Table C -8. (35) The average weight of particulate
emission during a 2-minute quench cycle was calculated to be about 6 pounds.
.Circled numbers in this text refer to points of emission bearing the same circled numbers in Figure C- 9.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-26
TABLE C.7. SCREEN ANALYSES OF COKE-PLANT PARTICULATES, WEIGHT PERCENT
Screen Size, Coal Coke Pyrol ytic High-Temperature Mineral Fly
microns Dust Balls Char Carbon Coke Matter Ash Total
-74 17.4 1.2 1.8 2.0 2.8 13.8 5.1 44.1
+74 -250 11.8 1.3 4.9 3.4 6.4 5.9 0.5 34.2
+250 1.1 0.5 7.6 2.2 6.4 3.7 0.2 21. 7
TOTAL 30.3 3.0 14.3 7.6 15.6 23.4 5.8 100.0
TABLE C-8. SCREEN ANALYSIS OF QUENCH-TOWER PARTICULATES
Screen Size
Mesh Microns
Weight Percent
Retained Cumulative
6
16
30
50
100
200
-200
3327
1167
589
298
147
74
-74
o
1
9
35
39
13
3
o
1
10
45
84
97
100
Coke-Plant Emission Control. Control of air-polluting emissions generated in a
coke plant is difficult because of the nature of the process and the great amount of
material handling that is required in the making of coke.
Coke-plant operators seek to control dusting at transfer points because the coal
losses are costly and because the dust is a serious fire hazard. To this end, the
handling operations are usually well shrouded with steel plate and are conducted (mainly)
inside steel buildings that are swept or flushed regularly. The 'Ibug dust" (as it is .
called) can ignite spontaneously; spread of fire to a dusty area can produce a violent
explosion. These factors promote rigid house-keeping standards, at least within the
buildings. Dust escaping to the surrounding grounds is seldom reclaimed.
Coal-handling operations are not all shrouded, and most of the particulate
emission of a coke plant arises from the unguarded unloading, stocking, and reclaiming
of fine coal during dry, breezy weather. If the coal is unloaded or stocked by conveyor,
the discharge is onto an open pile with high exposure. If'the coal is recovered with a
gantry crane, the clam-shell bucket must dump into a rail car or receiving hopper from
an appreciable height because of clearances for the equipment. The only known control
over systems of this kind is careful operation and suspension of work in windy weather.
Newer plants avoid these kinds of transfer; either by underground reclaiming (as
in Japan) or by use of carriers similar to earth.:.moving equipment. Special tractor
carriers can carry coal out onto a low-profile pile and dump from a bottom slot with'
minimum free-fall and dust. During reclaiming, the slot opens as a scoop and no
free-fall is involved in the open-stock area. The coal storage space required is much
larger, however, and thus many steel companies resist'this practice.
The sealing of end doors on most postwar coke ovens is mechanical, and is
achieved by applying closure pressure between a thin stainless steel strip on the door
BATTELLE MEMORIAL INSTIT!JTE'- COLUMBUS LABORATORIES
-------
C-27
and a matching flat plate on the oven end wall. Older ovens may have doors sealed
("luted") with fire clay, but these require more labor. Either type of seal can be only
so good as the amount of maintenance attention given to it. Accretions on, or damage
to, the sealing surfaces must be corrected before leakage will stop.
The cast-iron lids on the charging ports seal under their own weight if the mating
surfaces on the ports are swept clean before the lids are replaced. As with doors, the
effectiveness of the seal is directly related to the attentiveness of the operator. Most
operators keep the top of the battery well-swept at all times.
The rush of gases generated during the charging of coal into a by-product oven
blows fines out of the coal descending from the larry car. At the completion of charging,
steam, flame, and smoke puff upward through the ports of the larry-car hoppers. A
tight fit between larry-car discharge and oven ports has been considered as a solution
to this problem, but because of the rush of gas, coal then would be blown out of the
larry-car hoppers themselves, and thus create an even more serious particulate-
emission problem. Close tolerances between equipment and oven are almost an
impos sibility because the larry car rides on widely separated rails, and even flexes
upward as coal is discharged. If aspiration on the standpipe were increased enough to
prevent the flow of gases out of the charging port, air would then be sucked into the
nearest open port and the coke oven could explode during charging.
A current promising approach to the control of charging emissions is with suction
applied on a shroud pipe arranged around the charging port and fitted to the rim of the
charging port. The exhausted air is passed through a disintegrator located on the
larry car, and flared to the atmosphere, hopefully yielding only carbon dioxide and
water vapor as final emission. Equipment of this type has been installed in France and
Canada.
There are severe problems even with the shrouding and scrubbing of gases
exhausted from coke ovens by this method. If the mixed coal and shroud air ignite
ahead of the scrubber, there may be a violent explosion because of the restricted space.
Such explosions have occurred with the installations in Canada. Fresh water for the
scrubber system must be taken on each time the larry car receives a charge of coal, and
the dirty water must be discharged. The volume of water required for each charging
cycle is about 500 to 1000 gallons per charge of coal, and the dirty water requires
elaborate waste treatment.
It has been found possible to reduce somewhat the amount of particulates generated
during the quenching operation by the installation of baffles in the quench tower. (35)
Particulate emissions to the atmosphere before the installation of the baffles in one
case was about 6 pounds per load of coke quenched. After installation of the baffles in
the quench tower, the emission of particulates was reduced to about 3/4 pound per
quench, or a reduction of over 1000 pounds per day of particulates to the atmosphere
from one quench tower.
Another problem may be created in the selection of the water used to quench coke.
Many coke-plant operators make use of phenol-laden and ammonia-laden waste liquors
mixed with the quench water as a means of disposing of the waste liquor. Some
materials (such as ammonia) are dispersed adequately to the atmosphere by evaporation,
but other materials either precipitate in the coke or are carried into the air by the
steam rising from the quench tower. Phenols and other organics entering the blast
furnace with the coke (as a result of the waste liquor - water quench) are probably
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-28
safely destroyed because of the high temperatures in the blast furnace and the closed-
circuit emission-collecting systems used with blast furnaces. The same compounds
rising from the quench tower would be dispersed to the atmosphere. The use of baffles
in a quench tower might also serve to lower significantly the fallout of liquid droplets
in the vicinity of the quench tower. (35)
Ironmaking
Historically, the making of pig iron, or hot metal asit is called in the industry,
has been done mostly in,blast furnaces. Processes which bypass the blast furnace
have been the subject pf much research and development for many years. They are
termed "direct reduction" processes and produce either (1) molten pig iron generally
similar to blast-furnace hot metal, or (2) solid "metallized" products such as sponge
iron for use directly in steelmaking furnaces. A plant of the first type (rated to produce
about 200,000 tons of pig iron per year) is under construction at Mobile, Alabama, by
the McWane Cast Iron Pipe Company. (44,45) A plant of the second type (rated to produce
about 500,000 tons of steel per year from sponge iron) is under construction in
connection with the operations of the Oregon Steel Division of Gilmore Steel Company,
in Portland, Oregon. (46,47) However, such processes account for only a very small
fraction of the production~of metallic iron in the United States (roughly 0.3 percent).
The production of blast-furnace pig iron in the United States has increased steadily
over the years to keep pace with market demands. Production from 1958 through 1967
is shown in Figure C -15. The increase in production has been achieved with a
95
C 90
o
...
H
C1 85
a:
- 80
Oil)
CC
00 75
,- ....
....
u""
:J<1>
uC 70
0-
0:0
II) 65
-c
g,Q
c=
C.- 60
-------
~ 95
~
+-
OJ
C 90
'+-
o
(/)
,9 85
E~ 80
o
+-
OJ
~
+-
o
I
'0 70
c
o
'+=
g 65
-0
o
~
0... 60
o
:J
C
c
-------
C-30
The blast furnace is one of the largest chemical reactors used by man. New
furnaces commonly have a hearth diameter of 30 feet or more. The blast furnace acts
as a countercurrent reactor in which solid materials descend by gravity from the top,
react with gases generated near the bottom, and then are forced upward.
Iron ore, fluxes, and coke are charged into the top of the furnace through a
succession of two or three seals that serve to limit leakage of gas at this point. Pre-
heated air (sometimes augmented with oil, gas, oxygen, or steam) is forced through
ports (tuyeres) arranged radially near the bottom of the furnace and just above the
hearth. The incoming air and admixed additives react between themselves and with the
hot coke to generate a reducing gas rich in hydrogen and carbon monoxide, at a flame
temperature of up to 3500 F. The hot reducing gases liberate some of their heat to melt
the iron and slag, then continue upward to carry energy and chemical potential to the
unreduced ore in the upper part of the furnace. Molten iron and slag drip down into the
hearth and are tapped intermittently through special ports.
The ascending gas (with a typical calorific content of about 80 Btu per standard
cubic foot) is removed from the top of the furnace, stripped of dust, then used to fire
regenerative stoves for heating more air to be blown into the tuyeres. Surplus blast-
furnace gas (not needed to heat air) frequently is burned in the powerhouse to generate
electricity for the blowing engines (and other purposes), and less commonly is used to
heat the flues of the coke -oven plant.
The molten iron tapped from the blast furnace is put into special, large railroad
cars for transfer to steelmaking operations. The slag, also molten, may be granulated
with water, or may be put into steel pots for conveyance to a dump area. Dust trapped
in gas -cleaning operations frequently is recycled via the sinter plant.
Identification of Emis sions
The points at which emissions are generated in the production of pig iron in a blast
furnace are described with respect to the several processing steps involved. Flow
diagrams of blast-furnace operations having different modes of operation are given in
Figures C-18 through C-24. The range of burdens covered in these flow diagrams is
from unscreened ore (by today's technology, an unusual situation generally regarded as
"old hat") to "modern" burdens consisting mainly of sinter, pellets, or mixtures of
sinter and pellets. Several of the flow diagrams show the use of preheated air without
the injectionof any auxiliary fuel or steam, while two of the diagrams allow for the
injection of natural gas with the air blast. Many other arrangements of component
parts of these flow diagrams are possible. Combinations of burden, fuel and air used
to describe variations here, however, cover a wide range, from the conditions shown
in Figure C -18 where manufacture of 1 net ton of pig iron requires about 3000 pounds of
coke and 9750 pounds of heated air, to the conditions shown in Figure C-24 where
manufacture of the same weight of pig iron requires about 1025 pounds of coke and
4300 pounds of heated air.
In Figures C-18 through C-24, circled numbers are inserted at major points of
expected potential air-polluting emissions, and these circled numbers will be used later
to identify types of emissions with the sources as given on the flow diagrams.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
CD
@
C-31
@
BLAST FURNACE
@
Slag,
1,050 pounds
G)
Blast air,
9,7.50 pounds
(a) Some values disclosed by
various sources are 15, 59,
85, 100, 182, and 296
pounds.
@
@
Pig iron,
2,000 pounds
Other plant use,
10,950 pounds
Heated air,
9,750 pounds
Com bustion
products,
5,650 pounds
@
FIGURE C-18. TYPICAL BLAST-FURNACE OPERATION ON A BURDEN OF 100 PERCENT UNSCREENED ORE
- .
BATTELLE MEMORIAL INSTITUTE - COLUMB'US LABORATORIES
-------
C-32
-------
C-33
Q)
Q)@@
@
@
Iron are,
1,250 pounds
Sinter,
1,690 pounds
Coke,
1,370 pounds
Limestone,
350 pounds
Heated air,
4,880 pounds
BLAST FURNACE
@)
@
@
Slag,
784 pounds
Top gas,
7,300 pounds
Pig iron,
2,000 pounds
Other plant use,
5, 500 pounds'
G)
Com bustion air,
1,225 pounds
Blast air,
4,880 pounds
Heated air,
4,880 pounds
Com bustion
products,
2,770 pounds
@
FIGURE C-20. TYPICAL BLAST-FURNACE OPERATION ON A BURDEN OF LUMP IRON ORE AND SINTER
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
.
-------
@@@
@
Sinter,
2,600 pounds
Screened ore,
800 pounds
@)
Slag,
830 pounds
CD
C-34
@
Coke,
1,445 pounds
BLAST FURNACE
@
Limestone,
200 pounds
Heated air,
5, 850 pounds
@
@
Pig iron,
2,000 pounds
Other plant use,
6,600 pounds
Combustion air,
1,500 pounds
Atmosphere
Blast air,
5,850 pounds
Combustion
prod ucts.
3,380 pounds
Heated air,
5,850 pounds
FIGURE C-21. TYPICAL BLAST-FURNACE OPERATION ON A BURDEN CONSISTING MAINLY OF SINTER
BATTELLE MEMORIAL INSTITUTE ",:"CQLUMBUS I..ABORATORIES
-------
C-35
(i)
@
@
Heated air,
4,660 pounds
BLAST FURNACE
@
@
@
Slag,
440 pounds
Dust,
40 pounds
Top gas,
6,990 pounds
Pig iron,
2,000 pounds
Other plant use,
5,240 pounds
G)
Com bust ion air,
I, 190 pounds
Top gas,
1,750 pounds
Blast air,
4,660 pounds
Heated air,
4,660 pounds
Combustion
products,
2,695 pounds
@
FIGURE C-22. TYPICAL BLAST-FURNACE OPERATION ON A BURDEN OF 100 PERCENT PELLETS
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-36
-------
C-37
G)@@
-------
C-38
Charging of Raw Materials. The immediate solid raw-material requirements of
the blast furnace are held in surge hoppers under the railway high-line used for the
transfer of raw materials from storage areas to the blast furnace. In the case of coke
(and sometimes sinter), these hoppers usually are filled by conveyor belts; in the case
of ore and flux, bottom-dump-transfer cars frequently are used. From the surge
hoppers (called "pockets"), the materials are drawn in a predetermined sequence into a
system that weighs the prescribed charges and deposits them into a skip hoist leading
up to the furnace top. In some newer installations, conveyor belts are used to charge
the raw materials to the blast furnace. (48) The use of a conveyor-belt system.for
charging requires more area than a s~ip-charging system, and this is a disadvantage in
established steel plants where ground space is usually at a premium. Future new blast-
furnace plants probably will make more extensive use of conveyor belts for charging.
. .. !
The transfer and weighing system can take.many forms. In conventional United
States stockhouses, coke i"s fed almost directly into the skip horst, whereas, ore and flux
are discharged from the pockets to ,a second transfer car that is scale-mounted.
Operations in the stockhouse usually "dusty", depending on the materials, the weather,
and the degree of shrouding of transfer points. Newer stockhouses rely on conveyor
belts replacing the transfer car, and can be considerably cleaner (but are not nece's sarily
cleaner) .
The skip hoist containing a component of the charge is ,hoisted to the topmost part
of the furnace and dumped into a receiving hopper. This transfer is highly exposed, but
partial shrouding is possible. From the first receiving hopper, the charge is dropped
stepwise through one or two more hoppers and closures into the furnace. The multiple-
closure system is used to contain the furnace gases as charge is added.
The degree of fit between hoppers and their closures (called "bells") is variable
with design, age of the equipment, quality of maintenance, and other factors. Generally
the top equipment tends to become leaky with use because of abrasion, wear, creep,
and other di storting factor s. The higher the operating pres sure of a furnace, the faster
leakage develops. In new high-pressure blast furnaces, three closures are usually used
and an artificial steam system is used to maintain back-pressure between the two closed
seals during each transfer.
The present state of affairs is that the t9P of an operating blast furnace is
considered to be an area of continuous gas hazard while the furnace is in operation. Men
who must do maintenance work on top are sent aloft in safety groups, with compressed-
air breathing packs or hose-connected masks. It is probably practically impossible for
blast furnaces, even new ones, to be operated without continuously discharging at least
some noxious gases to the atmosphere.
Two emis sions come from the top of a blastfurnace; top gas and the dust which it
entrains. The top gas is a mixture mai:nly of steam, nitrogen, carbon monoxide, and
carbon dioxide. On a dry basis, this gas may average 25 to 30 volume percent carbon
monoxide; thus it is toxic. Emissions to the atmosphere occur from leakage around
hoppers and seals. Top gas may leak from instrumentation such as ports for rods used
to determine the height of the charge materials inside the furnace. Dust entrained in the
top gas is a result of the abrasion sustained by the burden materials during charging and
during the initial stages of passage down the blast furnace. It is possible to minimize
particulate emissions by using good raw materials and sound operating practices and
thereby reduce the load on the dust-cleaning system. The effect of improving the burden,
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-39
by using increased amounts of sinter and pellets,
net ton of hot metal going into the dust-collecting
Figure C-25. (29)
on the amount of dust generated per
system is illustrated in
(/)
"0
C
:J 300
o
a.
~
E
(1)
-
(/)
>-
Cf)
CJI 200
c
+= -
U a
(1)-
-(1)
<5 E
u-
-0
~.c:
0- 100
o
.£c
(1)0
- -
0-
c:: (1)
_c
(/) ...
:J (1)
00.
FIGURE C-25.
a 1000 2000
Amount of Sinter Plus Pellets Charged,
net ton of hot metal.
(Balance of burden is i ran are.)
EFFECT OF BURDEN IMPROVEMENT ON DUST RATE
FROM A BLAST FURNACE TO ITS DUST-COLLECTING
SYSTEM
High above the charging system on a blast furnace, at the uppermost part of the
gas-collection mains (called uptakes), two or more safety valves ("bleeders") are
located to relieve "unusual" gas pressures within the furnace. During abnormalities
in furnace operation, sudden movements ("slips") of the burden materials may occur,
During slips, the bleeders will open automatically to relieve the high pres sures, and
will discharge dust and gas to the atmosphere. The operating factors producing ab-
normal gas pressures also tend to increase the dust loading of the gas. Fortunately,
the use of improved raw materials to reduce the dust loading in a blast furnace also
tends to minimize abnormal operating conditions and lower the frequency of slips. With
modern burdens in use, the discharge of dust and gas during slips occurs only occa-
sionally. American plants have advanced considerably in this respect since 1950, and
for some furnaces the bleeders rarely open. A blast-furnace plant containing eight
furnaces .has reported that only one or two short-interval openings of the bleeders occur
during any given week for all eight furnaces. (49)
Smelting of Iron, The conventional blast furnace is a steel shell lined with re-
fractory bricks (see Figure A-7), and there may be numerous ports in the steel shell
to admit piping for water-cooled plates, Recently, blast-furnace designs have been
changing toward complete water jacketing with a closed shell. To the extent that the
furnace shell is pierced for instruments and coolers, gases generated inside the blast
furnace may leak back through the brickwork and to the exterior of the furnace shell
where they usually burn as small flames.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-40
Internal reactions, besides generating considerable carbon monoxide, can be quite
complex. The contact of .nitrogen in the blast air with hot coke, for example, conceiv-
ably can generate a variety of carbon-nitrogen compounds including cyanide. Tramp
elements in the burden (such as selenium) might be trapped, accumulated, and re-
fluxed in the furnace, or slowly emitted in the gas as some compound. Blast-furnace
burdens used in the United States usually are relatively free from known noxious
materials. However, as blast-furnace campaigns (the period between blast-furnace
relinings) lengthen so as to produce perhaps 4 million tons of iron, which requires
the smelting of about 6 million tons of ore, 2 parts per billion of arsenic in the ore
conceivably could accumulate as a ~4-pound slug of reflux arsenic in the furnace.
No data are available yet on these types of accumulations or on the possibilities for
their emittance.
Casting and Flushing. Throughout the preceding discussion, the sulfur charged
as part of the coke and as part of other materials has not been mentioned. Curiously
enough, almost none of this sulfur enters the top gas. Most of it is trapped and fixed in
the furnace by the fluxes, and is almost completely partitioned between the iron and the
slag in the hearth.
A modern American blast furnace produces between 400 and 700 pounds of slag per
ton of hot metal. On a volume basis, slag weighs about one-third as much as iron.
Therefore, the volume of slag is equal to or somewhat less than the volume of iron.
Except in very fast operations ,the' slag is flushed out of the furnace twice for each tap
of iron; once with the iron and once through a special slag notch about an hour before the
iron is cast. Some fast operations with frequent iron taps and low slag volume omit the
preliminary slagging operation.
The sulfur that is not dissolved in the iron is combined in the slag. The sulfur in
the iron is kept under very close control, rarely exceeding 0.6 pound per ton of iron,
and usually maintained at 0.5 pound per ton. The extent of sulfur control by the slag can
be illustrated by the following discussion.
If we consider a burden with 1100 pounds of coke (at 0.65, percent sulfur) per ton
of iron, ,and containing 1 pound of sulfur per ton of iron (from other sources such as ores
and sinter), the total sulfur load in the charge is just over 8 pounds per ton of iron. A
normal American range is 7 to 10 pounds. Deducting 0.,5 pound for sulfur in the iron,
one may deduce that the sulfur load on the slag is 6.5 to 9.5 pounds per ton of iron. The
amount of sulfur in slag normally ranges from 1 to 1. 8 percent.
When the hot slag is flushed from the blast furnace, the sulfur reacts with oxygen
in the air to form sulfur dioxide near the slag runners. In damp weather hydrogen
sulfide may also be formed by these reactions:
2 S-- + 302 -+ 2 S02 + 2 0--
S-- + H20 -+ H2S + 0-- .
The air can be fouled by these reactions during flushing and casting alike. At older
furnaces, the slag-runner system may extend for 100 feet or more, so that considerable
surface is exposed. In newer practices, slag is often run a short distance and then
quick-cooled and granulated with high-pressure water. This increases the effective
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
G-4l
surface area of the slag, and.the formation of hydrogen sulfide may continue in the
granulation pit at low temperature.
The molten iron as tapped from the blast furnace is saturated with carbon dissolved
from coke in the hearth of the blast furnace, and rejection of flakes of graphite begins
almost immediately as the iron emerges from the furnace. This graphite (called "kish")
is very much lighter than the iron, so it rises quickly to the surface where currents of
heated air sweep it into the atmosphere. Other emissions from the surface of the hot
iron include manganese vapor, which oxidizes upon escape to form a fine dust.
These emissions from the surface of iron generally are considered harmless by
the steel industry, and the amount of manganese vapor is usually small. But the
kish is a substantial dirt nuisance because it is eas.ily borne for long distances by light
breezes and has an oily tenacity that makes it hard to wash away from the articles it
settles on. At present, kish control is a matter of minimizing the amount precipitated.
This is done by running the iron short distances into closed" submarine" or torpedo>:'
ladles that are well preheated. The combination of minimal temperature drop and small open
surface is helpful in minimizing the formation of kish. Kish rising to the opening of the
ladle can usually be raked away and disposed of before it becomes airborne. Those
plants still using open-top ladles and oil-fashioned long runners have a substantial kish
problem which is worsened when the iron is transferred out of the ladles to a mixer or
pig machine. In a few instances, hooding has been installed with modest (but not good)
effecti venes s.
Kish also is released when hot metal is transferred from the submarine ladles to
the pouring ladle used to charge the steelmaking furnace. Special emission control
equipment is being installed in some steel plants to remove the kish at the hot-metal
reladling stations.
Burners and Stoves. A high proportion of the gases generated during the blast-
furnace smelting reactions is burned in stoves to heat refractory-brick checker work
(see Figure A-4), which in turn releases the heat to the air that is blown through the
blast stoves on its way to the blast furnace. The amount of blast-furnace gas generated
is primarily a function of the coke rate for any particular smelting practice, as illus-
trated in Figure G-26. (50) The distribution of gas between blast-stove use and other
in-plant uses is also shown. It should be noted that as blast-furnace practice improves
so as to lower the coke rate, availability of blast-furnace gas for other in-plant uses
decreases.
Blast-furnace top gas is thermally lean, and as blast-furnace operations have
improved, the gases tend to get still leaner. Whereas gas at 95 Btu per cubic foot was
common in 1950, the more usual heating value is now 80 to 85 Btu per cubic foot. The
calorific value of blast-furnace top gas as it is affected by the coke rate (which is an
indication of the degree of technology used in smelting) is shown in Figure C-27. (50)
Some gases are so lean that they must be mixed with coke-oven gas or natural gas to
assure adequate combustion and a sufficiently high flame temperature.
The burners for the blast stoves are fired alternately according to a predetermined
cycle and may be highly automated. One of the three or four stoves is always "on blast"
(i. e., heating blast air), while the others are "on gas" (being fired and heated). The
stoves are changed from blast to gas when they cool to a temperature at which they can
"The word "submarine" is derived from the shape of the ladle.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-42
no longer attain the required temperature of blast air. A conventional stove system ducts
all combustion gas to a common chimney, although separate -chimney installations still
exist. The drafting of a common chimney is better, and it can be built higher on the
same budget as several separate chimneys.
140
o
....
(1) 120
E
....
~o
~.c 100
<.9-
(1)0
(.) 5 80
g....
~ ~
~ ~ 60
1i)CU
o~ 40
(1) .~
..c
B 20
o
8
Gas available for
other use
o
1000 1200 1400 1600 1800 2000
Coke Rate, pounds per net ton of hot metal
FIGURE C-26.
2200
EFFECT OF COKE RATE ON VOLUME OF BLAST-FURNACE
GAS PRODUCED
125
+-
o
o
-
120
()
B
:J
()
...
(1)
a.
:J
-
(1)
i 100
o
>
()
-
...
o
o
U
+-
Q)
Z
FIGURE C-27.
"",,- --;
" .
" . ..
,fI1I'. .. .:. --
""...... .... ~.--
"" .:. .".--
~'I."'. :or"
A.I ...,
~ .,. ~
.-" .' "".
~. . "
,.-... '\.""
/ .:~, Y
/ . . ..)
/":t .,t
, ..C. ... /...
. ..~.
, ... .'"'
/ .'/
, . lit .. /
85 / /
./
. .. "
80 '
1000 1200
Coke Rate,
115
110
105
95
90
1400 1600 1800 2000
pounds per net ton of hot metal
EFFECT OF COKE RATE ON THE CALORIFIC VALUE OF
BLAST-FURNACE TOP GAS '
Heat release in the stoves is by the burning of carbon monoxide and hydrogen, with
traces of methane if natural gas is being injected at the tuyeres. These fuels are burned
with excess air, regulated by combustion controllers, often aided by excess -oxygen
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-43
sensors. The chimney effluent, then, generally is relatively free of objectionable
components (unless sulfur-bearing coke-oven gas is used to improve heat release).
Fuel Injection. The coke consumed in a blast furnace is a relatively expensive
fuel. A case can be made for the fact that in terms of effective heat released it costs as
much as electricity. To minimize coke consumption, heavy oil, tar, coal, coal-oil
slurry, coke-oven gas, and natural gas are in use as tuyere injectants. They are
introduced through blowpipes arranged to' discharge into the stream of hot blast air just
as the blast enters the tuyeres of the furnace.
Under normal conditions, these auxiliary fuels do not alter the types of emissions
from ironmaking. But under start-up conditions, and when the amounts injected are
large, auxiliary fuels may be incompletely burned at the lower level. As with any flame,
incomplete reaction can produce soot, and it is usual'that such' soot will blow on through
the furnace, foul the dust system, and overload the wet collectors until the situation is
recognized and corrective Ci.ction taken. Hopefully, as blast-furnace operators gain
experience with the use of injected auxiliary fuels, such temporary problems will become
less frequent.
Characteristics of Emissions. Emissions generated in the making of iron in the
blast furnace and in its immediate auxiliaries have major chara<:;te'ristics thatnow will
be described. The circled numbers accompanying these descriptions refer to locations
of emissions as identified by corresponding circled numbers on the flow diagrams that
make up Figure s C - 18 through C - 24. .
CD Iron-ore dust - Particles are rounded to' elongated in shape and can be as
small as 2 microns. Larger particles are opaque and red orange in top
light. Individual sma~l grains are transparent and blood-red. (11)
Hardness: 5 (Mohs)
Specific gravity: 5.2
Chemistry:
Usually mostly Fe203 with some Fe304.
in HCl. Contains some silica, alumina,
oxide s.
Mostly soluble
and phosphorus
o Coke dust - Particles are opaque, irregularly shaped, quite porous and rough
with some straight, sharp edges. They are gray-black in reflected light. (11)
Chemistry:
(Typical) 85-90 % fixed carbon, 2 % maximum volatile
matter, U.6 - 1.5% sulfur, 0.018 - 0.040 % phosphorus,
balance ash. (51) ,
CD Limestone dust - Mineral name calcite. It'is colorless, with light-transmitting
characteristics varying from transparent to translucent. Particles generally
occur as rhombohedra because of their perfect rhombohedral cleavage. Frag-
ments may also occur as prisms. (11)
BATTEL'LE MEMORIAL INSTITUTE - COLU'MBUS LABORATORIES
-------
C-44
Hardness
: 3 (Mohs)
Specific gravity: 2. 7
Chemistry: Mostly CaC03
o Flue dust - The blast-furnace flue dust typically contains 15 percent metallic
iron, 40 percent red iron oxide, 40 percent magnetic iron oxide, and 5 percent
limestone, (11) but many variations reust exist.
Iron - Fragments are opaque, black, and sharp, ma,gnetic, with finely
granular surfaces,( 11)
. Red iron Oxide (hematite) - Particles are transparent, rounded grains,
usually less than 2 microns in maximum dim~~sion. (11)
Magnetic iron oxide (magnetite) - Particles are opaque, black, rough
fragments, partially or completely coveredwithred iron oxide.(l1)
Limestone dust - Transparent, colorless rhombohedra, and rounded
particles. Many particles may also be covered with red iron oxide. (11)
Results of chem"ical anC!-lyses on flue dust generated in blast-furnace practice
are tabulated in Table C-9. (52, 53) The size analysis of the same sampling of flue
dust is given in Table C-10.
. -
Flue dust out of United States blast furnaces is reported to vary from 40 to 90
pounds per net ton of hot metal produced in a multi-furnace operation, with one new
furnace generating 40 pounds of dust per ton of iron. (1) A blast furnace in the United
Kingdom operating with a high-pellet burden is reported to generate only 20 pounds of
dust per ton of iron. (12) The dust loadings for the four blast furnaces of Great Lakes
Steel Corp., operating with a sinter and pellet burden were reported in 1967 to be as
shown in Table C-11. (54) These are dust loading to the dust-c~llecting system as
determined by actual collection. Specific data on particulate emissions to the atmo-
sphere from leakage at the top of the blast furnace or from slips are not available to
the project.
@ Top gas - The chemistry of blast-furnace top gas is determined by the
nature of the burden used in any particular furnace and the operating
variables such as blast temperature, injection ofa~xiliary fuels, and
the additionof moisture. The relationship betw,een moisture in the blast
air and the hydrogen content in the top gas is illustrated in Figure C-28. (55)
Top-gas analyses for several differ-ent types of burden situations in the United
States are listed in Table C-12. The compositions of the top gases are fairly consis-
tent, except for the burdens using unscreened ore, in which instances the carbon
monoxide contents are high and the carbon dioxide contents low. The unscreened ore
itself probably is not the:: caus~ of this difference in gas composition. The difference
probably is the result of other variations in practice that accompany the use of
unscreened ore.
BA.TTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-45
TABLE C- 9. CHEMICAL ANALYSES OF DRY, BMST-FURNACE FLUE DUST(52, 53)
Weight Percent
Component Range for Several Plants Midwest Plant
Fe 36.5 - 50.3 47.10
FeO n.a. 11.87
Si02 8.9 - 13.4 8.17
Al203 2.2- 5.3 1.88
MgO 0.9- 1.6 0.22
CaO 3.8- 4.5 4.10
Na20 n.a. 0.24
K20 n.a. 1.01
ZnO n.a. 0.60
P 0.1 - 0.2 0.03
S 0.2- 0.4 n.a.
Mn 0.5- 0.9 0.70
C 3.7 - 13.9 n.a.
n. a. - not available.
TABLE C-IO. SIZE ANALYSIS OF BLAST FURNACE FLUE DUST FROM
U. S. BLAST FURNACES(52) .
Size
U. S. Series Microns Range, percent
20 833 2.5 - 20.2
30 . 589 3.9-10.6
40 414 7.0 - 11. 7
50 295 10.7 - 12.4
70 208 10.0 - 15.0
100 147 10.2-16.8
140 104 7.7 - 12. 5
200 74 5.3- 8.8
-200 -74 15.4 - 22.6
TABLE C-l1. DUST LOADINGS FOR GREAT
LAKES BLAST FURNACES(54)
Blast Furnace
Designation
Dust Loading,
pounds per net ton of hot metal
-------
----
A
B
C
D
60
39
28
36
-----------------.------
-----.-------------------'--
BATTELLE MEMORIAL INSTITUTE - COLUMBUS 'LABORATORIES
-------
C-46
4.0
o
o 0
~~
B
8 0
8 0 8
o
8 000
c}o
o
o
o
0
-
c
Q)
o
L..
Q)
a.
Q)
E
:J
(5
>
o
3.6
o
3.2
o 000
o
o
II)
o
~
a.
~
2.8
. .
. .. .
. ...
,.. .
. .. ..
.. ...
..",.. .... .
00 .
Legend
Natural moisture
Added moisture
o
2.4
-
o
+-
C
Q)
+-
C
8
.
.
....,
I
o
c
Q)
01
o
L..
'0
:>.
I
.
:s..
1.2
12
grains per cubic
2 4 6 8
Moisture Content of Blast,
foot of air
FIGURE C-28. RELATIONSHIP OF MOISTURE IN BLAST AIR TO HYDROGEN IN BLAST-FURNACE TOP GAS
TABLE C-12. TOP-GAS ANALYSES FOR DIFFERENT BLAST-FURNACE BURDENS
-------------------
--------------- _._---- ----------------
-----.-.--,-.-----.-------- --------.---------.---
Burden
Auxiliary
Fuel Injection
Gas Content, volume percent
CO C02 H2
Unscreened Ore None 41. 5 7.1 3.3
Unscreened Ore None 32.9 7.3 3.6
Screened Ore None 27.2 12.8 3.1
Screened Ore Natural Gas 26.7 13.7 2.1
100 perc~nt Sinter None 23.7 14.8 3.5
100 percent Sinter Natural Gas 25.9 13.2 1.7
100 percent Sinter Natural Gas 23.5 16.3 5.7
50 percent Sinter +
50 percent Unscreened Ore None 24.1 14.1 4.1
50 percent Sinter +
50 percent Screened Ore None 25.0 14.2 2.6
50 percent Sinter +
50 percent Pellets None 23.8 17.3 2.4
100 percent Pellets None 24.9 15.4 2 9
100 percent Pellets Natural Gas 23.8 19.0 3.6
100 percent Pellets Natural Gas 24.0 16.7 2.9
100 percent Pellets Coal 26.1 14.4 3.4
----------- --------'-------------'- ------
--------_._- ---.--- --'- -------_._--------------- - --.--
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-47
@ Kish - Carbon in the form of flaky graphite that is rejected by the molten
iron as it cools during flow from the blast furnace to ladles. Other types
of particles may be entrained with this kish. The graphite particles are
opaque, black, sharply angular flakes with smooth surfaces. Some are
in layered agglomerates, occasionally showing rounded lZ0-degree angles,
and even forming rounded hexagonal tablets. Other particles accompanying
the kish may consist of opaque, black, rather coarse, fragments of magnetic
iron oxide, and transparent, deep-red, rounded particles of hematite.
Traces of quartz and calcite may also be found with the kish. (11) Graphite
typically makes up about 90 percent of the emis sian, with the magnetic
oxide at 5 percent and hematite at 5 percent.
Chemistry: Graphite (C)
Magnetic iron oxide (Fe304)
Hematite (FeZ03)
Quartz (SiOZ)
Calcite (CaC03)
(2) Hydrogen sulfide - When blast-furnace slag comes in contact with water, a
reaction occurs that forms. small amounts of hydrogen sulphide. The reaction
takes place during granulation, when the slag is hot, as well as at ambient
temperatures when the slag is cold. No data are available as to the amounts
of hydrogen sulfide that escape into the atmosphere. Because hydrogen
sulfide is detectable in very low concentrations, and has an unpleasant odor,
it is considered to be an air pollutant. Complaints from the surrounding
neighborhood tend to emphasize the problem. This is particularly true when
the slag is hauled by truck through the neighborhood on a rainy day. Research
work is presently underway, sponsored by the American Iron and Steel
Institute, and is directed toward determining methods of suppres sing the
formation of hydrogen sulfide from this source. (56)
@ Combustion emissions - Carbon monoxide emission produced from the
burning of fuels in the firing of blast stoves is small because of the careful
controls maintained for combustion, and the fact that burning is carried out
under conditions of exces s air. Sulfur dioxide can be present in the products
of combustion, if coke-oven gas is used to heat blast stoves.
Blast-Furnace Emission-Control Equipment
The primary reason for cleaning blast-furnace top gas has been to make it suitable
as a fuel for heating blast stoves, and secondarily to provide clean fuel to other opera-
tions in the steel plant. Generation of a recycled fuel in the smelting of iron is one of the
features that contributes to the economy of the blast-furnace process. If blast-furnace
gas were not cleaned, the particulate matter would clog the holes in the regenerative
brickwork of the stoves, slagging reactions would be accelerated and might lead to
catastrophic failure of the large amount of brickwork in the stoves. Changes in the
technology of blast-furnace practice have led to the use of. higher blast temperatures
from an average of 1300 Fin 1960 to temperatures varying from 1550 F to 1850 F in
1969. This continuing trend toward higher blast temperatures has resulted in the design
and construction of blast- stove refractory tile with smaller holes and thinner walls,
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-48
which are necessary to improve the heat-transfer characteristics in the blast stoves.
It is interesting to note the contrast between statements made on this subject in 1956
and 1967:
1956 - "As a result of the processes described, the cleaned gas has a
dust content of less than 0.01 grains per cubic foot. Gas this clean permits
the use of smaller checkerbrick in the stoves, which provides greater heating
surface area per stove and makes possible the use of higher blast temperatures
with a consequent improvement in furnace efficiency. (57) .
1967 - "Future high-temperature operation at 2000 F or more will
necessitate the installation of high-energy burners and will create the need
for gas cleanliness in the 0.001 grain per scf range. (54)
The type of emission-control equipment used on blast furnaces is affected by
( 1) the operating blast temperature, which governs the openings required in the tile in
the blast stoves, and (2) the availability of space around the blast furnace. The various
systems used have certain common pieces of equipment such as dust catchers and
primary washers, while subsequent items of equipment may be dictated by other in-plant
use of the blast-furnace gas. Flow diagrams of some blast-furnace gas-cleaning sys-
tems are illustrated in Figures C-29, C-30, and C-31. (58, 59, 60,61,62,63,64,65)
The gas cleaning system on the newest blast furnace constructed in the United
States (Youngstown Sheet and Tube Co. 's No.4 blast furnace at Indiana Harbor) consists
of a dust catcher, two automatically adjusted venturi scrubbers, and a gas-cooling
tower. It is claimed that this system will clean the gases to a dust loading of 0.005
grain per cubic foot. (66)
BLAST STOVES
and
BOILERS
PLANT USE
COKE OVENS
Weirton Steel, Weirton, W. Va. - 1958
U. S. Steel, Geneva, Utah - 1959
FIGURE C-29. FLOW DIAGRAMS OF TWO EARLY BLAST-FURNACE GAS-CLEANING SYSTEMS
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
PLANT USE
Kaiser Steel,
Fontana, Cal. - 1961
C-49
PLANT USE
PLANT USE
BLAST STOVES
OTHER
PLANT USE
CF & I.
Pueblo, Colo. - 1962
Jones & Laughlin,
Cleveland, Ohio - 1963
Armco Steel,
Ashland, Ky. - 1963
FIGURE C-30. FLOW DIAGRAMS OF BLAST-FURNACE GAS-CLEANING SYSTEMS
PLANT USE
U. S. Steel,
Fairless, Pa. - 1965
PLANT USE
PLANT USE
Great Lakes Steel,
Ecourse, Mich. - 1967
Bethlehem Steel,
Sparrows Pt., Md. - 1968
FIGURE C-31. FLOW DIAGRAMS OF THREE RECENT BLAST-FURNACE GAS-CLEANING SYSTEMS
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
G-50
Steelmaking
Prior to 1960, by far most American steel was made in open-hearth furnaces,
but since then the introduction of basic oxygen steelmaking (BOF) furnaces and the
installation of larger and more efficient direct-arc electric furnaces has resulted in a
decline in the percentage of steel made in open-hearth furnaces. In 1967, about
55 percent of the steel produced in the United States was still made in open hearths.
However, one set of projections made in that same year indicated that by 1969 basic
oxygen furnaces may be producing 59.5 percent of the steel, open hearths 20.5 percent,
and electric furnaces 20.0 percent. (67) Battelle's present projections are given in Sec-
tion IV of the :nain body of this report.,
Open-Hearth Furnaces
Although the open-hearth furnace is well on its way to lowered importance as a
means for making steel, and some forecasters see virtual extinction by 1990(68), there
are still many open hearths in operation or on a stand- by basis to be placed into opera-
tion to meet peak demands for steel. About 467 open hearths are in this category as of
mid-1968; by 1973 this number will be reduced to 257. The number is decreasing as
new BOF steelmaking capacity is installed. The distribution of all United State s open-
hearth furnaces by nominal capacity for 1968 and 1973 is shown in Figure C-32, while
similar distributions for several geographical steelmaking districts are similarly shown
in Figures C-33 and C-34. It can be seen that in 1968 the highest concentration of open-
hearth furnaces is in the Chicago District with 102 furnace s (9 furnaces located in
Duluth, Minn., for a district total of 111), followed by the Pittsburgh District with 101
furnaces, Youngstown with 65, and the Northeast Coast District with 43. No open
hearth furnaces are in operation in the St. Louis district.
The open-hearth proces s (known outside of the United States as the Siemens-
Martin Process) was developed on the basis of the regeneration principle, where checker
chambers are used for regeneration of heat. Regeneration was a necessity for the classic
open-hearth furnace because a gaseous or liquid fuel burned with ambient-temperature
air produces a flame temperature only slightly above 3000 F, which is not high enough
to melt the charge materials. Transfer of heat from the flame to the charge is primarily
by radiation. By using regenerators to preheat the air to 1000 to 1200 F, the flame
temperature is raised high enough to melt the charge materials and carry out the making
of steel. With the availability of tonnage oxygen, checker chambers and the regenera-
tive principle became unnecessary, and the basic design of the furnace no longer was
completely satisfactory for the rapid production of steel from charges high in proportion
of hot metal.
Although many open-hearth furnaces operate today with oxygen for refining the
steel, information on the number of open-hearth furnaces operating today with and with-
out oxygen was not available during this study. (69) Conditions that would favor open-
hearth steelmaking without oxygen-refining include (1) use of cold-melt charges of scrap
steel, (2) manufacture of steel with high ~arbon contents, and (3) local economic or air-
pollution considerations.
Emission Identification. The conventional open-hearth process begins usually with
the charging of up to 50 percent home scrap and purchased scrap to the hearth of the
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
100
~
~ 50
E
:J
Z
90
C-51
Total of 467 open hearth furnaces in 1968.
1973 = 257 furnaces (forecast)
Shaded portion represents furnaces forecast to
be removed from operation by 1973.
125 175 225 275 325 375 425 475 525 575
100 150 200 250 300 350. 400 450 500 550 600
Nominal Open Hearth Furnace Capacity, net tons
FIGURE C-32. SIZE DISTRIBUTION OF OPEN-HEARTH FURNACES
IN THE UNITED STATES IN 1968 AND 1973
80
70
60
40
30
20
10
o
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
L
~
Q)
..CI
E 10
::J
Z
~ 20
Q)
..CI
E
::J 10
z
~
Q)
E 10
::J
Z
~
Q)
..CI I 0
E
::J
Z
~
Q)
..CI
~ 10
z
20
30
20
20
30
C-52
North East Coast district 43 furnaces in 1968
16 furnaces in /970
o
~ To be removed from operation
Pittsburgh District 101 furnaces in 1968
37 furnaces in 1973
o
,Buffalo District 28 furnaces' in 1968
6 furnace in 197 I
o
Cleveland District 18 furnaces in 1968
6 furnaces in 1971
o
20
Youngstown District 65 furnaces in 1968
42furnacesinl970
o
125 175 225 275 325 375 425 475 525 575
100 150 200 250 300 350 400 450 500 550 600
Nomina 1 Open Hearth Furnace Capaci ty I net tons
FIGURE C-33.
SIZE DISTRIBUTION OF OPEN-HEARTH FURNACES
IN SEVERAL GEOGRAPHICAL DISTRICTS FOR 1968
AND SUBSEQUENT YEARS
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-53
20
~
OJ
..0 10
E
::J
Z
0
40
Duluth
30
~
OJ
..0 20
E
::J
Z
10
o
20
~
OJ
E 10
::J
Z
0
20
~
OJ
..0 10
E
::J
Z
0
Detroit District 12 furnaces in 1968
No furnaces after 1969
(QJ To be removed from operation
Chicago District
102 furnaces
9 furnaces - Duluth, Minnesota
III furnaces total in 1968
78 furnaces in 1973
District 27 furnaces in 1968
21 furnaces in 1970
Southern District 26furnaces in 1968
20
Western District 31 furnaces
27 furnaces
~
OJ
..0 10
E
::J
Z
o
125 175 225 275 325 375 425 475 525 575
100 150 200 250 300 350 400 450 500 550 600
Nomina I Open-Hearth Furnace Capacity, net tons
FIGURE C-34. SIZE DISTRIBUTION OF OPEN-HEARTH FURNACES
IN SEVERAL GEOGRAPHICAL DISTRICTS FOR 1968
AND SU BSEQUENT YEARS
(Size range as follows: ISO-ton furnaces includes open
hearths having nominal capacities from 150 to 174 net
ton s , etc.)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-54
empty furnace. ':' After a period of intense preheating of the scrap, hot metal from the
blast furnace is added, and firing is continued until the scrap is fully dissolved in the
molten charge. During this period of meltdown, limestone is added to form a basic
slag and to release carbon dioxide. The CaO component of the limestone and the C02
react with phosphorus, silicon, and manganese to form their respective oxides, which
in turn are absorbed by the slag. Another source of oxygen for these reactions is the
preheated scrap which becomes highly oxidized as it is melting. Iron ore may also be
added as a source of oxygen to help continue oxidation reactions and to remove carbon
from the hot metal.
Availability of high-purity oxygen has enabled steelmakers to shorten the refining
stages and combine them into one step by the use of oxygen lances. The lances extend
into the furnace through the roof and are activated immediately after the addition of the
hot metal. The rate of oxidation of metalloids and carbon is two to four times faster
than when limestone and ore are the main sources of oxygen.
Particulate and gaseous emissions from the open-hearth process originate from
(1) the physical action of the flame on charged materials and the resulting pickup of
fines, (2) the chemical reactions in the bath, (3) the agitation of the bath, and (4) the
combustion of fuel. (70, 71) The emissions include sulfur dioxide, carbon monoxide,
carbon dioxide, and fly ash from the fuels, plus iron oxide and other metallurgical fumes
from the steelmaking process. The metallurgical fumes include fine silica from the
burning of silicon monoxide released from the molten bath, manganese oxide from
manganese vaporized from the bath and subsequently oxidized, and iron oxide from the
rust on scrap, or in later parts of the heat, and from iron droplets oxidized in the
open area above the steel during refining.
Emis sions from open-hearth steelmaking come from the materials and fuels
used in the process. Variations in the types and amounts of emissions vary according
to the stage of the process. Some minor particulate emissions may be generated dur-
ing the charging of materials into the furnace and during tapping of the heat, but these
are of a minor nature. Flow sheets for various operating practices are given in Fig-
ures C-35 through C-41. While some of these practices may rarely be used, the flow
sheets are given as a matter of information. Figure C-4l (oxygen practice with
60 percent hot metal) is most typical of today' s operations in large steel plants.
Open-hearth furnaces generate four major types of particulate emission.
are indexed below with circled numbers keyed to Figures C-35 to C-4l.
CD Open-hearth dust - Charging period. Two components appear in the dust
generated during charging of the furnace. One is a magnetic iron oxide of
black, opaque spheres, and elongated, rough particles with sharp jagged edges,
all generally coated with red iron oxide. The second component comprises
transparent, rounded particles of red iron oxide, usually less than 2 microns
in dimension. They occur free or in simple agglomerates. (11)
These
*In tonnage steelmaking operations, the use of 40 percent scrap and 60 percent hot metal is common. Less common is the
"cold-melt" practice that uses no hot metal.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
G-55
Heated fuel oil,
and steam,
400 pounds
CD
@
@
Scrap yard,
60 pounds
@
FIGURE C-35. OPEN-HEARTH FURNACE OPERATING WITH A COLD-METAL CHARGE CONSISTING OF
30 PERCENT PIG IRON AND 70 PERCENT STEEL SCRAP (ORE PRACTICE)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-56
Heated fuel oil,
and steam,
326 pounds
-------
Heated fuel oil
and steam,
275 pounds
C-57
Rolling
operation,
2.000 pounds
\.
CD
@
Scrap yard
@
@
FIGURE C-37. OPEN-HEARTH FURNACE OPERATING WITH HOT-METAL PRACTICE CONSISTING OF
50 PERCENT HOT METAL AND 50 PERCENT STEEL SCRAP (ORE pRACTICE)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-58
Heated fuel oil
and steam.
258 ounds
Q)
@
@
@
FIGURE C-38. OPEN-HEARTH FURNACE OPERATING WITH HOT-METAL PRACTICE CONSISTING OF
60 PERCENT HOT METAL AND 40 PERCENT STEEL SCRAP (ORE PRACTICE)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-59
Heated fuel oil
and steam,
267 pounds
Steel scrap,
647 pounds
CD
-------
C-60
Heated fuel oil
and steam,
183 ounds
Oxygen,
55 pounds
CD
@
@
@)
FIGURE C-40. OPEN-HEARTH FURNACE OPERATING WITH HOT-METAL PRACTICE CONSISTING OF
50 PERCENT HOT-METAL AND 50 PERCENT STEEL SCRAP (OXYGEN PRACTICE)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-61
Heated fuel oil
and steam,
167 pounds
Steel scrap,
907 pounds
CD
@
@
@)
FIGURE C-41. OPEN-HEARTH FURNACE OPERATING WITH HOT-METAL PRACTICE CONSISTING OF
60 PERCENT HOT METAL AND 40 PERCENT STEEL SCRAP (OXYGEN PRACTICE)
BATTELLE MEMORIAL INSTITUTE';" COLUMBUS LABORATORIES
-------
C-62
o Open-hearth dust - Hot metal to lime-up. Three components make up the
dust from this period of open-hearth operation. (1) Loose agglomerates of
tiny transparent grains usually less than 1 micron in dimension. It is a
hydrated iron oxide such as HFe02. Individual grains and agglomerates are
yellow under top light. (2) Tiny, rounded, transparent, red grains of iron
oxide usually less than 1 micron in dimension. (3) Opaque, black spheres
and rounded particles of magnetic iron oxide. Some particles are covered
the hydrated iron oxide and/ or the red iron oxide. (11)
o Open hearth dust - tap to charge. The same as Item CD with the addition of
black, opaque, frothy, rounded particles of coke. (11)
@ Particulates in combustion product. About 85 percent of the material is
transparent, deep red, rounded grains, of iron oxide, usually less than
1 micron in dimension. The remaining 15 percent is black, opaque spheres
3 to 5 microns in dimension of magnetic iron oxide. The smaller grains
are orange in top light and tend to form simple agglomerates or loose
lumps.
All lime dust does not occur as a visually apparent particulate, it is present
in open hearth dust in very small quantities as shown by chemical analyses.
Sulfur in the form of sulphates also occurs in open hearth dust, but informa-
tion is not available in the published literature on visual characteristics.
No data have been located in the United States literature on the composition of the
products of combustion, but these do contain sulfur compounds originating from the
sulfur contained in the open-hearth fuels (such as oil, tar, and coke-oven gas). It has
been reported that the type of fuel (differentiating between tar and oil) has an effect on
the rate of dust emission, with average dust loadings being higher for tar and lower for
oil. (71) Results from an investigation conducted in Germany for similar fuel combi-
nations are shown in Figure C-42. (72) These data confirm results reported in the
United States. (71) Data from the same German investigators on the amounts of carbon
dioxide and sulfur dioxide in the waste gas during various stages of an oxygen-lanced
heat are shown in Figure C-43.
The amount of dust generated during the open-hearth steelmaking process varies
according to the different stages of the process (see Figure C-42) and according to the
practice. Oxygen lancing produces more particulate emission than open-hearth
practice without lancing. Pertinent information on the amounts of dust generated during
various stages is given in Table C-13. (70, 73, 74, 75)
The dust loadings for the various stages of the process vary over quite a range of
values. For example, the difference between the first two examples in Table C-13 is
quite large. However, the differences can be attributed to the time at which oxygen was
introduced into the process and the length of time it was used. In the first example,
oxygen was not used until after the hot metal had been added. In the second example,
the practice was to use oxygen from the time scrap was charged into the furnace until
the required levels of carbon in the steel were reached in the refining period. Similar
information on an 80-ton open-hearth operation in Germany, using a tar-oil fuel mixture
in combination with oxygen lancing, is shown in Figure C-44.
Typical dust loadings per net ton of raw steel are estimated as about 20 to 22
pounds for open hearths operating with oxygen lancing. (49,75) German literature
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
0.30
-
g 0.25
..-
<.>
:g 0.20
<.>
'0
~
o 0.15
'0
c
o
-
III
~
~ 0.10
III 0.09
c
'0 0.08
~
01 0.07
.~ 0.06
'0
g 0.05
...J
-
~ 0.04
o
FIGURE C-42.
C-63
40% Crude tar +
60 % Coke oven gas
Fuel oil
40% Fuel oil +
60% Coke oven
gas
!--Charging+-Melt - Down~-+ Lime-Ore Boil+ Refining-1
Time --
EFFECT OF DIFFERENT FUELS ON DUST LOADINGS FROM AN
80-TON OPEN-HEARTH FURNACE OPERATING WITHOUT
OXYGEN INJECTION (GERMAN PRACTICE)
-
111°
c.E 0 016
(!) u .
~:Q
~ ~ 0.012
3:~
cQ)
.- a. 0.008
NIII
OC
U)'o
~ 0.004
0.000
o
FIGURE C-43.
,,'
- - ..," \
,
\
\
CO2 ,,---
........... /'
"t, r, J
I \ 1\ I
, \ "
I \ , , I
\ I
, \ I 'I
, ,I \1
, 'J \I
(.1. "
,,,
5 III
o
(!)
4~
III
o
3:
3 c
2 3
Time, hours
N
20
u
4
5
SULFUR DIOXIDE AND CARBON DIOXIDE CONTENTS OF AN
OXYGEN-LANCED.OPEN HEARTH FIRED WITH A TAR-OIL
FUEL (GERMAN PRACTICE)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-64
TABLE C-13. DUST GENERATION IN OPEN-HEARTH STEELMAKING
---------------.----------.
-------"- - --------_._-------------
------------------------------
---------------
Plant
Practice
Year
of
Data
Dust Loading, grains per standard cu. fr.
Charge to Hot Metal Lime-Up ';ap to
Hot Metal to Lime-Up to Tap Charge
--_._-
"--------------.--------.---.----.-----
U. S. Steel Corp.
Edgar Thomson Works
National Steel Corp.
Weinon Steel Div.
Oxygen 1959 0.35 0.45 0.82 0.87
Oxygen 1965 0.78 1.90 2.70 0.21
Oxygen 1967 0.25 0.65 1.61 n.a
No oxygen 1963 0.56 0.61 0.18 0.11
Youngstown Sheet & Tube
Co.. Indiana Harbor Works
Steel Company of Canada
Hilton Works
----.-------.
----- ----------.------.- ---'----'------'---.-
'---'-'--------------'---
----------------------
Note: n. a. - data not available.
'0 5.0
~o
0' '+-
C
:u.~ 4.0
c.c
0:;)
-.Ju
- ~ 3.0
en 0)
~a.
o
~ 2.0
c
~
0' 1.0
lancing --+Tar-oil firing~
6
Sulfur as
sulphate
~
0)
5;
.s=.
a.-
4 ~ c
(j)0)
u
en ~
C 0)
a.
3--
en.s=.
~O'
0.-
2 .~ ~
3
Time, hours
FIGURE C-44. DUST LOADING DURING OXYGEN-LANCED OPEN-HEARTH
PRACTICE WITH A TAR-OIL FUEL 1,fiXTURE (GERMAN
PRACTICE)
contains a reported value 'of 11 pounds per net ton of raw steel(72), but here again, the
time at which oxygen is introduced into the process (and its duration) has an effect on
the amount of dust generated. An open hearth operating without oxygen injection had a
reported value of 7.95 pounds of dust per net ton of raw steel. (73)
00
~
I~
~
(j)
o
Data on the size distribution of particulate emissions from open-hearth furnaces is
almost nonexistent; information having been located from only one published source in
the literature. (70) Figure C-45 shows the size distribution of particulate emissions
from the No.4 Furnace of the Edgar Thomson Works of the United States Steel Corpora-
tion during the lime boil, as well as a composite sample. It should be noted that the dust
particles for the lime-boil sample average smaller in size than the composite sample.
About 77 percent of the particles in the lime -boil sample were smaller than 5 microns,
while the composite sample had only 46 percent of its particles smaller than 5 microns.
Chemical compositions of open-hearth particulate emissions are listed in
Table C-l4. (70,71,72,73,76,77) .
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
--
c: 25
Q)
u
~ 20
a..
-- /5
L:.
01
'(ii 10
3:
FIGURE C-45.
C-65
35
30
Composite sample
5
o
1-11-2 1-5 1-101-15 1-201-251-301-401
+2 +5 +10 +15 +20 +25 +30
Particle Size, microns
SIZE DISTRIBUTION OF OPEN-HEARTH PARTICULATE EMISSIONS
. FOR OPERATION WITH OXYGEN LANCING (D. S. PRACTICE)
TABLE C-14. CHEMICAL COMPOSITIONS OF OPEN-HEARTH PARTICULATE EMISSIONS,
OXYGEN LANCING, WEIGHT PERCENT
El em em
or U. S. Steel Corp. Steel Co. of Canada United Kingdom Germany
Compound Edgar Thomson Homestead Hilton Works United Steel Co. Dillingen U. S. Plant
Fe203 89.07 88.70 n.a. 88.5 79.65 n.a.
FeO 1. 87 3.17 n.a. 2.2 0.31 n.a.
Total Fe 63.70 n:a. 63. 5 - 68. 0 Q.a. 55.90 59.40
Si02 0.89 0.92 1.16 - 1. 56 0.4 0.47 2.00
A1203 0.52 0.67 0.15 - 0.44 0.4 0.52 0.48
CaO 0.85 1. 06 0.68 - 1. 06 0.9 0.88 1. 85
MgO n.a. 0.39 0.32 - 0.44 1.5 1. 86 1.12
MnO 0.63 0.61 n.a. n.a. 0.61 n.a.
Mn n.a. n.a. 0.43 - 0.55 n. a. n.a. 0.28
CuO n.a. 0.14 n.a. n.a. n.a. n.a.
Cu n.a. n.a. 0.11 - 0.16 n.a. n.a. 0.08
ZnO n.a. 0.72 0.26-2.04 n.a. n.a. n.a.
Zn 1. 70 n.a. n.a. n.a. n.a. 0 - 3.0
PbO n.a. n.a. n.a. n.a. n.a. n.a.
Pb 0.50 n.a. 0.05 - 0.95 n.a. n.a. n.a.
Sn02 n.a. n.a. n.a. n.a. n.a. n;a.
Cr n.a. n.a. 0.06 - 0.11 n.a. n. a. n.a.
Ni n.a. n.a. 0.03 - 0.05 n.a. n. a. 0.07
P205 0.47 1. 18 n.a. n.a. 1. 52 n.a.
P n.a. n.a. 0.06-'0.12 0.3 n.a. 0.15
S 0.40 0.92 0.34-0.70 1.4 2.69 2.78
Alkalies 1. 41 n.a. 0.56-1.71 n.a. 2.72 2.88
Note: n.a. - data not available.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-66
Open-Hearth Emission-Control Equipment. Open-hearth furnace installations
operated without the use of oxygen lances were in themselves fairly efficient dust coHec-
tors because of the auxiliary units needed to achieve efficient operation. These
auxiliary units are the slag pockets, checker chambers, and flues to the waste -heat
boilers. Dust-laden gases pass out of the open-hearth chamber through the downtakes
into the slag pockets, where there is a considerable decrease in the velocity of the gases
and a change in direction. This results ina deposition. of large particles in the slag
pockets. The waste gases then pass on to th~ checker chambers (the heat regenerators) .
where, because of additional changes in direction, additional deposition of dust particles
takes place. This is followed by further changes in direction as the gases pass into the
flues that conduct them to the waste heat boilers, where still more dust settles out of
the gases. However, the advent of oxygen la~cing in open-hearth furnaces resulted in a
need to install dust collectors to eliminate the large volumes of fume generated during
the lancing operation. .
The first installation of an electrostatic precipitator on an open hearth was made by
the Kaiser Steel Corporation in.the late 1940's. (78) This was a commercially available
unit which had not been designed specifically for use with open hearths. Many design
and operating problems were encountered and resolved as a result of this installation.
The first electrostatic precipitators designed specifically for open-hearth use were
installed in 1953 at the Fairless Works, United States Steel Corporation and by Kaiser
Steel at Fontana, California. (79,80)
Electrostatic precipitators have been the principal choic'e for emission control on
open hearths. However, the use of venturi scrubbers and bag houses has also been
investigated for the collection of open-hearth particulate emissions. In 1955, pilot-
plant work was started on the application of venturi scrubbers to open hearths, and
resulted in 1959 in the installation of a full-sized scrubber on the No.4 open hearth of
of the Edgar Thomson Works, United States Steel Corporation. (70) An experimental
bag-house program was initiated in 1960 at the Lackawanna Plant, Bethlehem Steel Cor-
poration, and resulted in the installation of full-sized equipment at the Sparrows Point
plant in 1963. (81, 82) Flow diagrams of dust-cleaning systems for open hearths ar~
illustrated in Figures C-46 and C-47. (70, 71, 75, 82-85)
It should be emphasized that the slag pockets, checker chambers, and waste-heat
boilers also act as part of the dust-collection system and are common to all systems in
open hearths. The waste -heat boiler and heat exchanger are shown in the flow sheet
(Figure C-47) with the bag house at the Sparrows Point plant because the heat exchanger
is required to assure that the gases entering the bag house are below the maximum
service temperature of the bag fabric.
Basic Oxygen (BOF) Furnaces
Basic oxygen furnaces are becoming the principal means of making steel in the
United States, and are expected by 1969 to be producing well over half of American
steeL The trends in installed BOF capacity and actual production are shown in Fig-
ure C-48. (86, 87) The lag between rated installed capacity and actual production from
1961 through 1965 can be attributed to the learning that must take place when new tech-
nology involving large, massive equipme!lt is brought on stream, and also to a lag in
additional sales necessary to absorb the new extra'steelmaking capacity. The lag be-
tween rated installed capacity and actual production decreased from 1965 through 1967,
as can be expected when operators become more knowledgable in a new technology. If
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-67
Six
electrostatic
precipitators
Four
electrostatic
U. S. Steel Corp.
Fairless, Pa. - 1953
Bethlehem Steel Corp. United States Steel Corp.
Sparrows Pt.. Md. - 1961 Homestead Works - 1963
FIGURE C-46. FLOW DIAGRAMS OF OPEN-HEARTH DUST-COLLECTING SYSTEMS
USING ELECTROSTATIC PRECIPITATORS
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
U. S. Steel Gorp.
Edgar Thomson Works- 1959
C-68
Youngstown Sheet & Tube Go.
Indiana Harbor Works - 1963
Republic Steel Gorp.
Buffalo Disuict- 1964
Bethlehem Steel Gorp.
Sparrows Pt.. Md. - 1963
FIGURE C-47. FLOW DIAGRAMS OF OPEN-HEARTH DUST-COLLECTING SYSTEMS USING SCRUBBERS AND BAG HOUSES
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-69
the trends of instal.led capacity and actual production continue to 1970, it may be antici-
pated that by that year BOF furnaces will be producing over 70 million net tons of raw
steel per year. .
80
10
-- .."
--.
,
,
,
Future .'
planned '
installations i/
I
~~
Rated 7 '/
installed.
- -- capacity IJ
I ''/- -Production
"..~ I
~
"..' "..,
~ .."".
I I I I I I
VI 70
c
o
+-
Q) 60
c
-
~ 50
c
~
"E 40
Q)
Q) 30
en
~
fi. 20
o 1958 1960 1962 1964 1966 1968 1970 1972
Year
FIGURE C-48.
TRENDS IN INSTALLED CAPACITY AND RAW -STEEL
PRODUCTION IN BOF FURNACES
Emission Identification. In bas"ic oxygen steelmaking, the need for a large-surface
bath, (such as is required in open-hearth steelmaking) is overcome by forcing a jet of
high-purity oxygen below the surface of the metal under pressure. The jet also provides
violent agitation, and therefore increases the area of the slag-metal interface. The
BOF process is exothermic to the extent that up to 30 percent (or more) of steel scrap
can be melted using as fuel only the carbon and other metalloids dissolved in the metal.
No conventional fuel is added. In terms of emis sions, the sulfur dioxide and unburned
hydrocarbons associated with open hearths are nonexistent with BOF furnaces.
In the initial stage of basic oxygen steelmaking, the charging of carbon- saturated
hot metal upon cold scrap results in a release of kish as the molten iron is rapidly
cooled. Only a part of this kish is contained by the furnace vessel. The initiation of
oxygen blowing is marked briefly by a heavy dark-brown smoke' (caused by the direct
burning of iron) which persists until the metalloids begin to oxidize and refining begins.
J.ne first elements burned are silicon, manganese, and phosphorus; their. oxides enter
the furnace slag, but absorption is imperfect, and some white silica and lime fume with
minor amounts of manganese enter the fume-exhaust system.
As most of the metalloids .become oxidized, the oxidation of carbon increases in
rate to consume the rest of the oxygen blown, and the volume of gas leaving the furnace
mouth increases noticeably. An. excess of air often is permitted to mix with the exhaust
gases as they pass into the fume-exhaust system. This is taken as a safety precaution
to prevent the existence of a high carbon monoxide content in the flue system and elimi-
nate a possible explosion hazard. An analysis of operating data relating the amount of
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-70
oxygen blown to the volume of exhaust gases (C\.t temperature of combustion) generated
has shown that the volume of exhaust gases is about 25 times greater than the volume of
oxygen blown, This relationship is shown in Figure C-49. (88)
-
Q)'
Q)
-
u
:0 16
~
u
g 14
o
~ 12
c::
~ 10
CD
c:: 8
Q)
C\
~ 6
o
- 4
o
Q)
E 2
~
0
> 0
o
FIGURE C-49.
18
Vessel tonnage,
nominal rating
. '"
165T
25 SCFM gas = I CFM02
50
100 150 200 250 300 350 400 450
Exhaust Gas Volume,IOOO cubic feet
RELATIONSHIP BETWEEN THE VOLUME OF OXYGEN
BLOWN AND VOLUME OF EXHAUST GASES
During the oxidation of carbon, the fuming appears to be limited to iron dust either
from iron vaporized from the bath or as iron droplets ejected by the carbon monoxide
rising from the molten bath. Factors that determine the amount of fumes generated
during the blowing process include the type of oxygen lance used, the velocity of the
oxygen, the carbon content of the iron, and the temperature of the iron. An effect of
the number of holes in an oxygen lance is shown in Figure C-50. (89) The s'ingle-hole
lance shows a higher pE'!ak of gas emission ,than a multiple-hole lance, but the total
amount of gas evolved is about the same for both designs. The s,ingle-hole lance is no
>.
-
+-
c:
C
:::J
a
If)
c
(!)
Q)
+-
If)
C
~
FIGURE C-50.
"',
i \
i \
j \ Single-hole lance
-- I, ,',
I , '
8' I -,
: \. .L' -----..l....Multiple-holelance
: "', ...-- \ \
'I ' ,
., \ ,
! i ~
,/ "',
10 15 20 25
Blowing Time. minutes
30
35
EFFECT OF NUMBER OF HOLES IN OXYGEN LANCE
ON EMISSIONS DURING OXYGEN BLOWING
BATTELLE MEMORIAL .INSTITUTE - COLUMBUS LABORATORIES
-------
C-71
longer used, because the multiple-hole lances provide other better operating charac-
teristic s in addition to having lower peaking volumes than the single-hole lances.
The effect of velocity on the amount of emissions obtained from laboratory experi-'
ments is illustrated in Figure C-51. (90) The effects of increasing the velocity of the
oxygen stream are quite pronounced, but at the higher velocities, noise becomes a prob-
lem in the vicinity of the ve sse 1.
c
- 1000
~
o
5 10 50 100 5001000 5000
Calculated Oxygen Velocity, feet per second
8000
0'
E 7000
-t:i 6000
QJ
>
05
>
w 400
QJ
-a 3000
E
C/) 2000
Blown with oxygen
at 0.14 cubic foot
per minute
"'C
c
~
~u.
00
;>.10
:-=~
~-
Q) C
>Ic
Q)IQ)
- CI
~I~
.)( 10
°lc
~.-
0.1.
'$....
I
FIGURE C-51.
EFFECT OF VELOCITY ON EMISSION DURING OXYGEN
BLOWING OF BOF FURNACE (LABORA TORY RESULTS)
The effect of carbon content in the iron on the amount of emissions generated is
illustrated in Figure C-52 from two independent laboratory investigations. (90,91)
10
()
Q)
en
...
Q)
a. 8
CI
E
Q) 6
-
C
a::
c
,2 4
:;
(5
>
w
Q) 2
.JC
0
E
(/)
00 1 2 3 4
Carbon, percent
:t 1.5
~
.c
Q)
:;
c:
'E
,
~ 1.0
...
Q)
a.
,
,
,
t
,
.
I
I
~
en
c:
o
...
~
c:
.2 0.5
-
~
o
>
W
Q)
.JC
o
E
(/)
o
A
o
o
2
Car bon I percent
6
FIGURE C-52.
EFFECT OF CARBON CONTENT ON EMISSIONS FROM
BOF FURNACE (LABORATORY RESULTS)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-72
The effect of metal temperature is shown in Figure C'- 53, again from labo rator)"
investigations. (90)
12
u
Q)
en 10 .
.....
C'
E
Q)
- 8
c
a:::
c:: .,
0
- 6
~
0
>
w
Q) 4
...: '
o
E
(f) 2
.
o
2300 2400 2500 2600 2700 2800 .2900 . 3000
Temper'ature I F
FIGURE C-53.
EFFECT OF METAL TEMPERATURE ON GENERATION OF
EMISSIONS FROM BOF FURNACE (LABORATORY RESULTS)
From the effects of the various factors illustrated in Figures C-49 to C-53, it is
evident that the amount of evolution of emissions. from a BOF furnace is dependent upon
the interaction of numerous factors. These factors are controlled mainly in such a way
as to make steel most effectively and economically and are controlled secondarily to
inhibit generation of emi s sions. .
The predominant particulate emission is brown iron oXide, and the only gas of
concern is carbon monoxide. A predominance of submicron sizes in the oxide dust
makes it especially difficult to trap and collect: The calculated mean diameter for dust
obtained from specific surface measurements was reported to range from l..s to 2.9
microns. (92) More recent work using electron-microscope counting techniques has
indicated a particle-size distribution with a count median diameter of 0.012 micron. (93)
If galvanized scrap is used as part of the scrap charge to the basic oxygen furnace, the
fume will contain 7.inc ferrites from vaporization and oxidation of the zinc. In some
plants, zinc oxide may comprise from 5 to 8 percent of'the BOF dust and render it
useless for recycling to the blast furnace via the sinte~ plant. (Zinc is a recognized
destructive agent of blast-furnace refractories.) The problem of handling galvanized
scrap is a considerable one for companies manufacturing large tonnages of galvanized
product. If open hearths are available, it is considered preferable to divert galvanized
scrap to the open hearth, rather than use it in basi~ oxyg~n fu~naces.
Many systems have been designed to handle the carbon monoxide generated. The
newer ones seem to emphasize minimizing the amount of aspirated air because of prob-
1ems of overheating of tubes and other parts of the hood and duct work. Many basic
oxygen furnaces are equipped with waste-heat boilers; a few try to recover chemical heat
from the carbon monoxide by burning it; some seek only to flare the gas safely to the
atmosphere.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-73
The Kaldo rotary steelmaking process (used by only one plant in the United States)
has the same characteristic emissions as the BOF steelmaking process, except for the
amounts and size of emissions. The carbon monoxide content of the exhaust gases is
much lower from the Kaldo process than from the BOF steelmaking process, and the
size of particulate emissions is reported to be larger. (94) . .
Typical £lowsheets for the BOF steelmaking process and for the Kaldo process are
given in Figures C-54, C-55, and C-56.
The major emissions from BOF furnaces are described below. The circled num-
bers below refer to emission locations indexed with the same numbers in Figures C-54
through C-56.
CD "Kish" - Carbon in the form of graphite is rejected by the molten iron as it
cools during charging into a BOF steelmaking vessel on top of cold steel
scrap. The graphite particles are opaque ,black, sharply' angular flakes
with smooth surfaces. Some are in layered agglomerates, occasionally
showing rounded 120-degree angles, and even forming rounded hexagonal
tablets. Other particles may consist of opaque, black, rather coarse,
fragments of magnetic iron qxide and transparent, deep red, rounded
particles of hematite. Trances of quartz and calcite may also be found
with kish. (11)
Chemistry:
Graphite (C)
Magnetic iron oxide
Hematite (Fe203)
Quartz (Si02)
Calcite (CaC03)
(Fe304)
o Silica fume - Approximately 50 to almost 100 percent silica, often containing
small quantities of iron, manganese, magnesium, and carbon. Color of the
collected material is gray to off-white. Its bulk density is about 10 to
12 pounds per cubic foot. (95)
Chemistry: Si02
Size distribution of silica fume as determined by counting techniques using
1500 X photomicrographs is given in Table C-15.
TABLE C-15.
SIZE DISTRIBUTION OF SILICA FUME
Particle Size,
microns
Particle Larger,
number percent
Particle Larger,
cumulative percent
0.28
0.20
O. 16
0.11
0.085
0.059
0.040
0.022
-0.022
0.40
0.25
2.35
7.00
17.00
21.00
27.00
20.50
4.50
0.40
0.65
3.00
10.00
27.00
48.00
75.00
95.50
100.00
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-74
BASIC OXYGEN FURNACE
@
@
Raw steel,
2, 000 pounds
Slag, ,
298 pounds
Dust, ,
47 pounds
Off gas,
192 pounds
Rolling
operation,
2, 000 pounds
Scrap yard,
78 pounds
(a) Charge scrap plus cooling scrap = 456 pounds
FIGURE C- 54. BASIC OXYGEN FURNACE OPERATING WITH 80 PERCENT
HOT METAL AND 20 PERCENT STEEL SCRAP
BATTELLE- MEMORIAL INSTITUTE":' COLUMBUS LABORATORIES
-------
C-75
Oxygen,
138 ounds
BASIC OXYGEN FURNACE
@
@
Ra w steel,
2,000 pounds
Off gas,
168 pounds
Rolling
operation,
2,000 pounds
Scrap yard,
77 pounds
Dust collector
(a) Charge scrap plus cooling scrap = 678 pounds.
FIGURE C-55. BASIC OXYGEN FURNACE OPERATING WITH 70 PERCENT
HOT METAL AND 30 PERCENT STEEL SCRAP
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-76
Steel Scrap,
988 pounds
Hot metal,
I, 208 pounds
Ferro alloys..
14 pounds
Oxygen,
159 pounds
CD
ROT ARY OXYGEN FURNACE
@,
GY.
Scrap,
70 pounds
.Dust,
22 pounds
Off gas,
181 pounds
Scrap yard,
70 pounds
Dust collector
FIGURE C- 56. ROT ARY OXYGEN FURNACE OPERATING WITH 55 PERCENT
HOT METAL AND 45 PERCENT STEEL SCRAP
BATTELLE MEMORIAL INSTITUTE - COLUMBUS l:ABORATORIES
-------
C-77
Q) Basic oxygen process dust - Tiny (less than 1 micron in dimension) rounded,
transparent particles of red iron oxide, which tend to agglomerate. Shiny
black spheres of magnetite covered with red iron oxide. (11) The particle size
of this dust is discus sed in more detail below.
There is a great discrepancy in the reported size analysis of BOF dust. Reports
have been made of 95 percent less than 1 micron in size(96), as well as 99 percent less
than 0.2 micron in size. (97) These reports show median sizes of 0,45 and 0.065 mic-
rons respectively. A more recent report states a count median diameter of 0.012
micron. (93) The Kaldo proce ss is reported to have a particle distribution in which
6 percent of the material is less than 1 micron in size, which makes the Kaldo dust
larger than the BOF dust. (94) The larger particulate size for the Kaldo process is ex-
plained as a result of agglomeration of finer particles.
Data on grain loadings per cubic foot of off gas are limited to a single source which
reports that the dust concentration ahead of a precipitator varied from 2.02 to 4.96
grains per cubic foot and averaged 3.59 grains per cubic foot. (98) Data on the amount
of dust generated per net ton of raw steel produced by the BOF process varies over a
large range. In 1959, the dust generated was reported to be between 14.5 and 27.4
pounds per net ton. (99) In 1965, results for an operation in Europe were reported to
vary from 19.6 to 46.6 pounds of dust per net ton of steel. (100) In 1968, an average
figure of 40 pounds of dust per net ton of steel was reported for an operation in the
United States. (101) Dust generation for steel made by the Kaldo process was reported to
be 10 pounds per net ton of raw steel. (102)
Chemical compositions of basic oxygen furnace dusts are given in Table
C-16. (12,88,103) The high zinc content in the dust from two plants is due to the use
of galvanized- steel scrap in the BOF charge. A variatiOn in the composition of BOF
dust will exist from plant to plant, depending on the particular compositions of hot
metal and scrap charged. Variations will also exist within any given plant, depending
on the particular type of steel produced in each heat.
Compositions of some off gases from ,BOF and Kaldo steelmaking ve ssels are given
in Table C-17(88, 102, 104), and a log of off-gas composition before combustion with
aspirated air is given in Figure C-57. (104)
BOF Emission-Control Equipment. The first BOF steelmaking furnaces were in-
stalled in 1954 by the McLouth Steel Company at their T renton, Michigan, plant. Air-
pollution-control equipment consisted of a wet washer and disintegrator. (105) A BOF
steelmaking plant placed into operation the same year in Canada by Dominion Foundries
and Steel Ltd. made use of a wet-washing system with venturi scrubbers. (106) The sec-
ond and third BOF plants installed in the United States were at the Jones and Laughlin
Aliquippa Works and at the Fontana plant of Kaiser Steel Corporation, where electro-
static precipitators were used for dust collecting. (98)' The choice of electrostatic pre-
cipitators at Fontana was influenced to some degree by the fact that Southern California
is a water- short area, and the water system at the plant was a recirculating system
that would have required extensive expansion to handle the additional load from a wet-
washing system. Flow diagrams for typical gas-cleaning systems are shown in Fig-
ure C-58 for wet-cleaning systems(107-109) and in Figure C-59 for electrostatic
systems. (110, Ill)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-78
TABLE C-16. CHEMICAL COMPOSITIONS OF BASIC OXYGEN
FURNACE STEELMAKING DUST, WEIGH:T
PERCENT
Element 0 r BOF Dust from
Compound Typical U. S. Plant s
FeO 1.5 n. a. n. a. n. a.
Fez03 90.0 80.00 n. a. n. a.
Fe n. a. n. a. 56.0 57.68
Mn304 4.4 n. a. n. a. n. a.
Mn n. a. 0.35 I.Z 1. 54
SiOZ 1. Z5 z.oo 1.9 1. Z9
AlZ03 O. Z O. 15 0.4 O. 13
CaO 0.4 5. 10 3. 1 3.59
MgO 0.05 1. 10 n. a. 0.63
S n. a. O. lZ 0.09 O. lZ
P n. a. O. 10 O. Z 0.09
PZ05 0.3 n. a. n. a. n. a.
Cu n. a. 0.04 0.03 n. a.
Zn n. a. Trace 1. 93 4.80
Note: n.a. - data not available.
TAB LE C - 17.
COMPOSITION OF OFF GASES FROM OXYGEN
STEELMAKING PROCESS, VOLUME PERCENT
Gas
BOF Process
Before Combustion After Combustion
With Aspirated With Aspirated
Air Air
Kaldo Process
Before Combustion
With Aspirated
Air
COZ
CO
NZ
5.0-16.0
74.0-90.5
0.7-13.5
0.0-0.3
7Z.9
ZZ.7
3.0-8.0
74.5-78.9
4.4
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-79
95
90
85
+- 80
c:
<1>
u 75
...
<1>
Q.
70
<1>
E 20
:J
.g 15 ',Carbon dioxide
10 """"""""'-
........
5 Nitrogen """'- -- --
Oxygen
0 ......
0 2 4 6 8 10 12 14 16
Time, minutes
FIGURE C-57.
OFF-GAS ANALYSIS FROM A 60-TON BASIC OXYGEN
CONVERTER (BEFORE COMBUSTION)
Electric Steelmaking Furnaces
The price of electric energy has declined during the last decade. This and greater
availability of steel scrap have made electric furnaces attractive economically for the
production of plain carbon steels, as well as for the manufacture of alloy and stainless
steels. During the last 10 years, the annual tonnage of steel produced in electric fur-
naces in the United States has doubled. In 1957, electric furnaces accounted for 7 per-
cent of all the steel produced in the United States; this increased to 11 percent in 1967.
Electric-furnace steel production from 1957 through 1967 is shown in Figure C-60. (8)
The trend in recent years has been toward larger and more powerful electric furnaces
in the integrated iron and steel industry. The distribution of furnaces by size, as of
1968, is shown in Figure C-61. (112, 113, 114) Of the total of 195furnaces, 40 percent
have capacities less than 50 tons, 36.4 percent have capacities between 50 and 90 tons,
and the remaining 23.6 percent have capacities greater than 100 tons. Electric furnaces
are combined with continuous casting machines in some of the newer installations.
These range in size from 25 -ton furnaces used in mini steel plants that cast billets to
150-ton furnaces that are used for producing the steel for multi strand continuous cast-
ing machines such as are being installed at Jones and Laughlin's Aliquippa Plant and
National Steel's Great Lakes Division Plant at Ecourse, Michigan.
Emission Identification. Emissions generated during electric -furnace steelmaking
originate from the physical nature of scrap used, the cleanliness of the scrap, the na-
ture of the melting operation, and oxygen lancing. Thin steel scrap will oxidize easily
and result in heavy fuming and a high metal loss during melting in electric -arc furnaces.
For this reason, thin steel scrap is considered generally undesirable by electric -furnace
operators. Dirty scrap is a major source of emissions. Dust emissions as high as
30 pounds per net ton of steel can result from the use of particularly dirty scrap. (115)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
Inland Steel Company
Indiana Harbor, Ind. - 1965
C-80
Bethlehem Steel Corp.
Sparrows Point, Md. - 1965
Wheeling Steel Corp.
Steubenville, O. - 1965
FIGURE C-58. EXAMPLES OF WET-CLEANING SYSTEMS FOR BOF STEELMAKING FURNACES
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
generating
hood
Great Lakes Steel
Ecourse Mich. - 1962
C-81
Wisconsin Steel
South Chicago 111. - 1964
FIGURE C-59. EXAMPLES OF ELECTROSTATIC-PRECIPITATOR GAS-CLEANING SYSTEMS
FOR BOF STEELMAKING FURNACES
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
FIGURE C-60.
FIGURE C-61.
C-82
~ 16
o
Q.I
:>. 15
~
Q.I
a. 14
I/)
C
o 13
+-
+-
Q.I
C 12
-
o
I/) II
C
o
E 10
C 9
o
+-
U
:J
"tJ
0
~
a..
6
1957 1959 .1961 1963 1965 1967
1958 1960 1962 1964 1966 1968
Year
ANNUAL PRODUCTION OF RAW STEEL IN ELECTRIC
FURNACES IN THE UNITED STATES
28
26
24
III 22
~ 20
g 18
~ 16
IL.. 14
15 12
~ 10
.a 8
E
:J 6
Z 4
2
o
10 20 30 40 50 60 70 80 90 150 200 250
15 25 35 45 55 65 7585 100 175 225
Furnace Capacity. net tons
SIZE DISTRIBUTION OF ELECTRIC STEELMAKING FURNACES
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-83
Figure C-62 shows the reported dust emissions (in pounds per net ton of steel) for 22
operations using various combinations of normal scrap (scrap with little to moderate
rust), dirty scrap (scrap with heavy rust), and oxygen landng.(llS-122) It can be seen
that electric-furnace melting with dirty scrap can generate as much dust as an electric-
furnace heat using oxygen lancing. It has been estimated that 20 percent of the emis-
sions are produced during oxygen lancing, with the remaining 80 percent attributed to
the meltdown period. (119) Dust loadings during the oxygen-blowing phase of a 40-ton
electric -furnace heat are shown in Figure C -63. (119)
- 25
Q)
Q)
+-
(I)
.....
o
c: 20
o
+-
+-
Q)
!=
~
~ 15
(I)
"'0
c:
:::J
o
a. 10
(I)
c:
o
(I)
(I)
E
w
+-
(I)
:::J
o
FIGURE C-62.
FIGURE C-63.
3
100
. Normal scrap- with oxygen lanCing
o Normal scrap- no oxygen lancing
o Dirty scrap - no oxygen lancing
.
o
5
....6
. "
....'
....""
Q.""
....""
....
....
....
....
....
. . ....~"
. .....
. . "
"
0'
...,0"
-...0'"
.
.
o
o
o
o
o
00
30
10 15 20 25
Melting Rate, net tons per hour
DUST EMISSIONS DURING ELECTRIC-FURNACE
MELTING OF STEEL
4
~ 3
~ -
-
-
-
-
-
--
01 U
c: .-
.- .Q
"0 ::J
C u 2
o ...
...I Q)
Co
-
-
-
-------
C-84
Flow diagrams for electric-furnace melting are shown in Figures C-64, C-65,
and C -66. All practices shown are for cold-melt practice. Hot metal (as a part of the
electric-furnace charge) has been used on occasion in the making of electric-furnace
steel(l23, 124), but this is not a routine type of operation and is used mainly when short-
duration scrap shortages may occur in an integrated iron and steel plant that has access
to hot metal from a blast furnace. Depending upon the practice and raw materials used,
principal emissions from electric-arc steelmaking furnaces are dust from the furnace
itself, dust from scrap preheaters, and furnace off gas. These are discussed separately
below using circled numbers to key the descriptions to location of emission in
Figures C-64 through C-66.
CD Electric-furnace dust - Opaque, rounded grains that are peach to reddish in
color in top light. Small agglomerates are present, but are not common. (11)
The chemical composition of electric -furnace dusts will be influenced by the com-
position of the steel being melted. Because of this, optical characteristics of the dust
may also vary because of the different alloying-element oxides that may be present.
Because electric furnaces are used for melting a wide range of alloy and stainless steels,
the chemical composition of any particular dust will reflect the composition of the alloy
melted. Some reported compositions of dust emis sions are given in
Table C-18. (125-128)
TAB LE C - 18 .
CHEMICAL COMPOSITIONS OF ELECTRIC-FURNACE DUSTS
,
WEIGHT PERCENT
Element or Sample Designation
Compound A B C D E F
FeO 4.2 n.a. n.a. n.a. n. a. 4 - 10
Fe203 35.04 50.55 52.62 52.05 50.05 19 - 44
Cr203 0.00 0.56 0.00 0.15 13.87 0 - 12
MnO 12. 10 12.22 5.34 1.29-2.58 n. a. 3 - 12
NiO 0.30 n.a. tr tr 3. 18 0 - 3
PbO n.a. n. a. 3.47 0.81-1.08 n. a. 0 - 4
ZnO n.a. n.a. 8.87 1.24-2.48 n. a. 0 - 44
Si02 8.80 5.76 6.78 3.85 5.50 2 - 9
A1203 12.90 5.85 2.55 14.61 n. a. 1 - 13
CaO 14.90 2.60 6.72 1.40-4.20 9.80 5 - 22
MgO 7.90 7.78 3.49 1.66-4.98 6.64 2 - 15
S 0.26 tr 0.59 n.a. n. a. 0 - 1
P 0.10 0.28 n. a. n.a. n.a. 0 - 1
C 2.30 n.a. n. a. n. a. n. a. 2 - 4
Alkalies 1.20 4.76 n.a. n.a. 2.50 1 - 11
Note: n. a. - not available, tr - trace
Sample A - Single 20-ton furnace. Plant specializing in tool and die steels.
Sample B - Representative sample from plant with four 75-ton and two 200-ton furnaces producing
low-alloy and stainless steels.
Sample C - Single IOO-ton furnace producing low-alloy steels for plate.
Sample D - Single IOO-ton furnace producing low-alloy steels for plate.
Sample E - Single 70-ton furnace producing stainless steel.
Sample F - Representative samples from multiple-furnace shop. Furnaces vary in size from 4 to 200-ton, producing
low-alloy and stainless steels.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-85
ELECTRIC FURNACE
@
CD
Rolling
operation
FIGURE C-64. EXAMPLE OF ELECTRIC-FURNACE STEELMAKING USING A CHARGE OF
COLD STEEL SCRAP (ORE PRACTICE)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS' LABORATORIES
-------
C-86
Atmosphere
ELECTRIC FURNACE
@
-------
C-87
Atmosphere
FIGURE C-66. EXAMPLE OF ELECTRIC-FURNACE STEELMAKING USING A CHARGE OF COLD STEEL SCRAP
(OXY-FUEL BURNERS FOR MELTDOWN; OXYGEN PRACTICE)
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-88
The high contents of zinc and lead oxide in some of the above analyses can probably be
attributed to the use of galvanized scrap and scrap containing lead (terne plate). The
change s in composition of electric -furnace dust during various periods of making a
particular heat of low-alloy steel are given in Table C-19. (129)
TABLE C-19.
CHANGES IN COMPOSITION OF ELECTRIC-FURNACE
DUST DURING A SINGLE HEAT
Composition, weight pe rcent
Period Fe203 Cr203 MnO Si02 CaO MgO Al203 P205 S02
Melting 56.75 1. 32 10.15 9.77 3.39 0.46 0.31 0.60 2.08
Ore oxidation 66.00 1. 32 5.81 o. 76 6.30 0.67 0.17 0.59 6.00
Oxygen lancing 65.37 0.86 9. 17 2.42 3. 10 1.83 O. 14 o. 76 1.84
Refining 26.60 0.53 6.70 Tr 35.22 2.72 0.45 0.55 7.55
Tr - trace.
The size of particulate emissions from electric -furnace melting in the United
Kingdom has been reported to be 30 percent by weight below 10 microns for furnaces
operating without oxygen lancing, and 40 percent by weight below 3 microns for furnaces
operating with oxygen lancing. (130) Size distributions for some electric steelmaking
furnaces operating in the United States are given in Table C-20. (122)
TABLE C-20. SIZE DISTRIBUTION OF PARTICULATE
EMISSIONS FROM ELECTRIC STEEL-
MAKING FURNACES, PERCENT
Particle Size, Electric Furnace Size, net tons
microns 3.5 4 14 17 19 50
o to 5 57.2 63.3 59.0 72.0 43.3 71.9
5 to 1 0 37.8 17.7 33. 1 10.5 17.7 8.3
10 to 20 3.4 8.0 4.9 2. 7 6.4 6.0
20 to 40 1.6 8. 1 3.0 4.7 14.6 7.5
greater than 40 0.0 2.9 o. 0 10. 1 18.0 6.3
o Dust from scrap preheaters - Information on the physical and optical
characteristics of dust generated during the preheating of scrap has
not been located. However, it can be as sumed that the composition of
the dust will be influenced mostly by the cleanliness of the scrap, its
content of volatile matter, and presence of surface coatings on some
of the steel.
Information on dust generated during scrap preheating is not available for el~ctric-
steelmaking furnaces in the United States. Investigations carried out in Norway have
resulted in data relative to dust emissions during scrap preheating. (131) Work done
with heat sizes varying from 6.6 to 26.4 net tons has shown that particulate emissions
varied from 0.040 to O. 111 grains per cubic foot, which is equivalent to 0.07 to 0.27
pound per net ton of scrap charged. Although dust generation varied considerably, no
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-89
correlation was found between dust loading and any of the operating variables. The
variations in grain loadings are shown in Figure C-67, and the size distribution of the
particulate emis sions is shown in Figure C - 68.
0.12
(; 0.10
- 0
0'-
.!:: u 0.08
"'C .-
0..0
o ~
....J u 0.06
~
+-Q)
enQ.
c5 en 0.04
c
o
~0.02
--,
I
~-----..,
L_-
---
S5
-H3 Test I
Numbe~
H2
S4
-----
0.000
40
FIGURE C-67.
DUST LOADINGS DURING PREHEATING OF SCRAP IN NORWA Y
24
22
"'C 20
Q)
.: 18
o
Q) 16
a:::
+- 14
c
~ 12
~
~ 10
+-
.J::. 8
0'
Q) 6
~
4
2
o
850 200 100
400 150 75
Size, microns
I 40 I 100 I 200 I 325 I
20 70 140 270 -325
Equivalent U.S. Series
55 -45
45
FIGURE C-68.
SIZE DISTRIBUTION OF DUST PARTICULATES
DURING SCRAP PREHEATING IN NORWAY
According to this information, while the dust generated from melting in an electric
steelmaking furnace varies from 82 to 100 percent less than 40 microns in size, 98 per-
cent of the dust generated during preheating of scrap is larger than 45 microns in size.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
+-
C 80
Q)
u
~
Q) 70
c.
E 60
::J
g 50
+-
~ 40
+-
C
<3 30
(/)
o
(!)
C-90
100
Ore addition
Start of boil ~
~, Slag Off. ~ ~ ~
'0- -q ~
, , ,
A 1\
N2 \ ,\ if
, \ /,
\ ,. 'r/'
\' ,
\' ,
\ ,
9--...,
I ........
I ........ _0
I '0.. -(Y'" - -
I N2
I
I
J
Oxygen lancing
90
20
10
o
o
2
3
Time, hours
4
5
FIGURE C-69.
"U
Q) Q) 800
>+-
-::J
°c
>.-
wE
Q)
"U~
.- Q)
xc.
o
g (U 400
~Q)
-
Cu
.8:0 200
~ ::J
Ou
U
FIGURE C-70.
GASES GENERATED DURING THE PRODUCTION
OF A BALL- BEARING STEEL IN A 22- TON
ELECTRIC FURNACE
1000
600
16
CARBON MONOXIDE EVOLUTION DURING OXYGEN LANCING
IN A 16. 5-TON ELECTRIC STEELMAKING FURNACE
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-9l
o Electric -furnace off gas - The off gas is made up mainly of carbon dioxide,
carbon monoxide, oxygen, and nitrogen. The composition of the off gas
varies somewhat with the slag practice used, the stage of the heat, and the
use of oxygen for lancing. The change in composition of the gas during the
course of one heat is illustrated in Figure C-69. (129) Carbon dioxide stayed
below 15 percent during the entire course of the heat, while the carbon
monoxide content varied inversely with the nitrogen content. The carbon
monoxide content of the off gas during oxygen lancing went as high as
85 percent.
Carbon monoxide evolution during the oxygen-lancing period is shown in Fig-
ure C-70 for a 16. 5-ton electric furnace producing high-chromium steels. (132) For
the same furnace, it also was possible to relate the evolution of particulate emissions
with that of carbon monoxide, as shown in Figure C-71. (132)
~ 5.0
::J
c:
E
~
~ 4.0
tJ)
"'0
c:
::J
~ 3.0
E
w
100 200 300 400 500 600
Carbon Monoxide, cubic feet per minute
700
FIGURE C-71.
RELATIONSHIP BETWEEN THE EVOLUTION OF DUST
PARTICULATES AND CARBON MONOXIDE IN A 16.5-
TON ELECTRIC FURNACE
Electric-Furnace Emission-Control Equipment. Control of emissions from
electric steelmaking furnaces is affected by the design of the furnace. Two designs of
electric-arc furnaces used in the integrated iron and steel industry are (1) the movable-
roof top-charging furnace, and (2) the fixed-roof door-charging furnace. The top-
charging furnace has an advantage of permitting more rapid charging, while the door-
charging furnace permits somewhat better control of melting practice and improved
refractory life because of an essentially closed system and lower impact on the furnace
hearth during charging.
Emissions from electric -arc steelmaking furnaces are controlled and collected
by three main types of systems: (1) collection of emissions by the use of hoods over and
around the furnace at points of emission, (2) direct extraction from the furnace interior,
and (3) shop-roof extraction and collection. Emissions leave an electric furnace around
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
L---
C-92
the electrode ports in the roof of the furnace, the tapping spout, slagging door, and in
the case of top-charged furnaces, through the open furnace top during charging.
Hooded Collection. A hooded collection system consists of close-fitting hoods at
the points of emission to collect the particulate and gaseous emissions and carry them
via a duct system to the dust collector. Several hoods are required, and must be
movable in the case of top-charging furnaces. A hooded system tends to obscure
visibility from the crane -operator's cab, and results in added operational hazards.
Direct Extraction. This system consists of an exhaust opening located in the roof
of the furnace so as to draw off the emissions and direct them to the collector. The
system is operated in such a manner that air flows into the furnace and out the exhaust,
thus minimizing the discharge of emissions through the various doors and electrode
ports. This method has shown a tendency to cause the loss of some alloying elements,
as shown in Figure C-72 for manganese. (133) Difficulties have also been reported in
1:: 90
Q)
~ 80
Q)
c: 70
>-
Qj 60
>
8 50
Q)
a::
Q) 40
en
~ 30
a
:? 20
a
~ 10
o
FIGURE C-72.
0.05 0.10 0.15 0.20
Carbon at Tap, percent
0.25
MANGANESE RECOVERY IN STEEL AS AFFECTED
BY THE METHOD OF EXTRACTION OF ELECTRIC-
FURNACE EMISSIONS
operating with special carbide slags in refining of special steels. (120) Otherwise, there
are no apparent metallurgical difficulties associated with direct-extraction methods. (120)
Advantages of direct extraction with respect to furnace operation have been reported to
include decreased electrode consumption and increased roof life. Electrode consumption
has been reported to decrease 8 percent and roof life to increase 16 percent. (118) An
increase in roof life from 170 heats to 320 heats per roof reline has been reported for
Roanoke Steel on a 20-ton furnace. (134) Direct-extraction systems usually require
lengthy duct work or heat exchangers to cool gases to a safe temperature before they
enter a bag house. It has been determined that a minimum of 600 feet of ducting is
required to cool direct-extracted off gases by radiation before their entry into a bag
house. (135) -
Shop-Roof Extraction Systems. Emission collection by this method consists of
making the shop building itself a large collecting hood for all emissions generated in a
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-93
multiple-furnace shop. Special exhaust ducts and hoods must then be utilized to ensure
proper removal of the emissions from the shop.
Flow diagrams of examples of direct-extraction and furnace- shell extraction sys-
tems with bag-house collectors are given in Figure C-73(l20, 127, 136, 137), dust-
collecting systems using wet scrubbers in Figure C-74(l28, 138,139), and shop-roof
extraction systems in Figure C-75(l34, 140, 141).
A new operating technique that is making an appearance in electric-furnace melt-
ing is the continuous charging of metallized pellets into the electric furnace through the
roof of the furnace. The technique is also possible with fragmentized scrap, which can
conceivably develop into a system for preheating scrap and for exhausting emissions
from an electric furnace.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
m
»
~
~
111
r
r
111
~
111
~
o
::0
»
r
Z
III
~
~
c
~
111
,
o
o
r
c
~
m
c
III.
r
»
m
o
::0
»
~
o
::0
111
III
Lukens Steel Co.
Coaresville, Pa. - 1964
Lukens Steel Co.
Coatesvi1le, Pa. - 1966
U. S. Steel Corp.
Chicago, 111. - 1961
()
I
-..0
~
Bethlehem Steel Corp.
Seattle, Wash. - 1959
FIGURE C-73. EXAMPLES OF DIRECT-EXTRACTION EMISSION-CONTROL SYSTEMS WITH BAG HOUSES
-------
C-95
Armco Steel Corp.
Butler, Pa. - 1959
Armco Steel Corp.
Houston, Texas - 1966
FIGURE C-74. EXAMPLES OF ELECTRIC-FURNACE DUST-COLLECTING SYSTEMS USING WET SCRUBBERS
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-96
Ingot
pouring
area
Jones & Laughlin Steel Corp.
Warren, Mich. - 1966
Top-charge
.furnaces,
one 70 ton
two 100 ton
Bethlehem Steel Corp.
Los Angeles, Cal. - 1966
FIGURE C-75. EXAMPLES OF ELECTRIC-FURNACE SHOP-ROOF EMISSION-CONTROL SYSTEMS
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-97
Vacuum Degassing
Vacuum degas sing of molten steel (as discussed in Appendix A) was first developed
as a means of removing hydrogen from steel, and later was developed further to use the
carbon- oxygen reaction to deoxidize the molten steel. These technological developments
resulted in cleaner steels with improved properties.
Lmission Identification
The use of vacuum-degassing processes results in off gases that contain hydro-
gen, oxygen, nitrogen, carbon dioxide, carbon monoxide, and methane. Dust is also
generated because of the violent agitation of the metal in the vacuum chamber, and the
combined effects of high vapor pressures of the metallics and low pressure in the
vacuum chamber. The off gases are combustible, as indicated by a report that when
removing 3 parts per million of hydrogen and O. 04 percent carbon at a treating rate of
5 tons of steel per minute, the gases leaving the steam- ejector system when ignited pro-
duced a flame 10 feet high. (142) When the same amounts of hydrogen and carbon are
removed at a treating rate of 30 tons of steel per minute, the exhaust gases have a net
combustion value of about 250,000 Btu per minute. The dust particulates will vary in
composition depending on the alloying elements in the steel and their respective vapor
pressures at the pressures in the vacuum vessels. Flow diagrams for the two general
types of vacuum-degassing processes are shown iri Figures C-76 and C-77.
The major emissions are described below, using circled numbers to key the
descriptions to the locations in Figures C-76 and C-77.
(.0 Gases liberated from the molten steel during degassing are principally
(a) the hydrogen that has been dissolved in the steel, and (b) a carbon
monoxide - carbon dioxide mixture resulting from the reaction between
carbon and oxygen in the steel. Although nitrogen occurs in the off gas,
its source is air in the system or aspirated air. Small amounts of
water, oxygen, methane, and argon may also be present. The compo-
sition of gases released will vary during the course of the degassing
treatment as shown in Figure C-78. (143)
The variation in the composition of off gases during vacuum- stream degassing for
different grades of rotor steel, at variou9 times and levels of vacuum, are given in
Table C-21. (144) ..
BATTELLE MEMORIAL INSTITUTE --COLUMBUS LABORATORIES
-------
C-98
Molten steel
STEAM EJECTOR
@
@
Atmosphere
Dust.
Vacuum-de assed molten steel
LAD LE
INGOT MOLD
FIGURE C-76. VACUUM-STREAM DEGASSING OF MOLTEN STEEL
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-99
Molten steel
VACUUM -DEGASSING CHAMBER
-------
I
I
C-IOO
-
c::
Q)
o
~ 40
a.
70
60' ~\
50 \ I ',\Yd'Ogen
\! '\,
~
30 /\
,
-
c::
~
c::
8
en
o
C>
15
20
25
30
Time I minutes
FIGURE C-78.
VARIATION IN GAS CONTENTS DURING
VACUUM-STREAM DEGASSING
TABLE C-21. COMPOSITION OF OFF GASES DURING V ACUUM-STREAM DEGASSING OF STEEL
Typical Steel Analysis, weight percent Time, Pressure, Gas Analysis, mole percent
C Mn Si Ni Cr Mo V W minutes microns C02 CO H2 02 H20 N2 A
0.24 0.60 0.28 2.75 0.30 0.08 0 80 1.6 12.9 34.9 50.1 0.5
5 320 1.9 21.7 40.3 1.5 3.5 30.9 0.2
10 360 0.6 26.0 59.0 0.5 2.3 11.5 O. 1
15 340 2.5 38.4 31. 4 0.8 1.7 25.0 0.3
22 1000 1.8 49.7 24.5 0.6 0.6 22.6 0.3
0.38 0.70 0.75 2.80 1. 20 0.50 0.18 0 175 3.2 18.5 1.6 9.0 6.7 68.2 0.6
5 400 0.7 26.1 62.1 0.1 0.1 10.8 0.1
10 410 0.4 19.6 67.2 0.2 0.6 11. 9 --
16 300 2.1 27.0 50.0 4.4 1.8 14.7 --
24 4.3 38.2 24.0 0.1 0.3 32.6 0.4
26 590 4.1 33.0 19.6 0.5 1.7 40.5 0.5
0.25 0.88 0.25 0.88 12.00 0.88 0.20 0.88 0 80 0.4 21.1 1.9 75.6 1.0
5 540 1.7 19.5 65.6 0.0 0.2 12.9 0.1
14 550 4.6 44.3 37.0 0.0 0.5 13.3 0.2
16 550 6.6 47.0 26. 1 0.2 1.0 18:8 0.2
0.28 0.88 0.28 0.25 1. 05 1. 25 0.25 0 160 0.3 1.2 -- 20.7 10.4 66.5 0.8
5 620 2.4 29.8 41.4 0.1 3.6 22.5 0.2
11 680 1.1 37. 5 41. 5 0.1 1.6 18.0 0.1
19 600 1.1 48.3 33.7 0.2 1.8 14.9 0.1
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-IOl
Examples of composition during vacuum- stream degassing of ball- bearing steels
are given in Table C-22. (145)
TABLE C-22.
OFF-GAS COMPOSITION DURING VACUUM-STREAM
DEGASSING OF BALL- BEARING ST EELS
Steel Analyses, weight percent Gas Analyses, volume percent
C Mn Si Ni Cr Mo V C02 CO H2 02 N2 CH4
0.23 0.75 0.14 0.26 O. 33 2. 7 15.8 29.5 50.7 1.3
O. 32 1. 31 0.14 O. 36 0.37 0.27 1.2 36.4 42. 0 1. 1 15. 6 3. 7
O. 38 1.35 0.22 0.04 0.26 0.26 1. 3 47.9 27.0 1.8 20. 1 1.5
0.40 O. 60 0.86 0.22 4.91 1.40 O. 91 7.4 24. 7 10. 1 O. 7 51. 2 5.9
O. 33 1.34 O. 16 O. 18 O. 37 0.29 0.8 40.4 28. 7 1.3 27.3 1.5
0.32 0.46 0.20 O. 17 .3.58 0.53 1.3 44. 0 19.9 24.9 9. 9
@ The compositions of the off gases emitted to the atmosphere are essen-
tially the same as those of the gases evolved from the steels in the
vacuum-treating units. Some solution of the gases in water may occur
as the gases are exhausted with steam and pas s through intercondensers
and a hot-well before being exhausted to the atmosphere. Analyses made
of the gases as they exhaust to the atmosphere from D-H vacuum-
degassing installations have shown them to have an average composition
of up to 80 percent carbon monoxide, up to 15 percent hydrogen, and up
to 20 percent carbon dioxide. (146) .
G) Metallic dusts - are deposited on the walls of the steel chambers used
@ in stream degassing. Descriptions of the size consist of the dusts
have not been located in the literature. However, the dusts are very
finely divided and are pyrophoric in nature to the extent that safety
precautions must be exercised in opening the tanks after the steel has
been processed. (144, 147)
Chemical composition of one deposited dust has been reported to include the follow-
ing: 2. 1 percent carbon, 1. 1 percent silicon, 0.4 percent aluminum, 78.0 percent
manganese, and 12. 3 percent iron. (144) Analyses of other dusts generated during
vacuum- stream degassing and analyses of the metal in the ladle after treatment are
given in Table C - 23. (147)
TABLE C-23.
DUST AND METAL ANALYSES FOR VACUUM-
TREA TED STEELS
Elements, weight percent
Material C Mn Si Ni Cr V Mo Cu Fe
Steel in ladle 0.33 0.73 0.25 .2.86 o. 99 0.22 0.53 O. 17
Dust 1.66 46.30 1. 63 O. 38 o. 36 0.01 O. 05 1. 60 17.60
Steel in ladle 0.33 0.83 0.26 O. 17 1. 01 0.23 1.21 O. 14
Dust 1.69 47.70 1.40 O. 13 0.38 O. 04 O. 09 1. 20 15.50
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-102
@ Dusts generated during vacuum degassing are not emitted to the atmos-
phere because of the dust-washing characteristics of the steam ejectors
used to create. the vacuums required. All of the vacuum-degassing
installations used in the integrated iron and steel industry in the United
States use steam ejectors. Detailed reports on the amount of dust gen-
erated during vacuum treatment have not yet been located in the pub-
lished literature. A single report does state that about 10 pounds of
dust are generated during a single cast and collected in the vacuum
system, and because the installation is designed to handle lOO-ton heats,
the generation of dust is about one-tenth of a pound per ton of steel. (148)
Vacuum Degassing Emission-Control Equipment
As has been stated in the preceding discussion, practically no particulate emissions
enter the atmosphere from vacuum-degassing installations. This can probably be attrib-
uted to the fact that the steam ejectors used for creating the necessary vacuum are to a
certain extent venturi scrubbers. A vacuum-degassing installation may have from 4 to
6 ejectors and related intercondensors, depending on the amount of steel to be treated
and the vacuum levels required. A typical steam- ejector system is illustrated in
Figure C-79. (144) An ejector is shown schematically in Figure C-80. (149) Its simi-
larity to a simple venturi scrubber shown in Figure B-15 is apparent.
XH"'un
r VACUATOIt
#. IT[AW-EJECTOR PUMP
II 2 INTIACONDENSIUt
HOT WILL
If J INTEACONOfNIER
E"ICTOIt
11 I 'TE'M ..nCTOR
TO' 0' CHARGINO 'J.9!!.
-.-
AtUUM
TANK
1iiiT'TOiiO,- PIT
FIGURE C -79.
TYPICAL STEAM-EJECTOR SYSTEM FOR VACUUM DEGASSING
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C -1 03
STEAM CHEST
DISCHARGE
FIGURE C~80.
CROSS SECTION OF A STEAM EJECTOR
Gases passing through the steam- ejector system are usually emitted to the atmos-
phere, except where they may be passed through a water hot-well and be a hazard to
working personnel, in which case they are ignited and flared to the atmosphere. (142)
.l ne amount of gases which may be discharged during vacuum degassing is influenced by
the carbon content in the steel and the amount of hydrogen to be removed. The weight
and volume of gases, as equivalent air, for a typical situation are listed in Table
C~24. (142)
TABLE C-24.
EFFECT OF TREATING RATE AND AMOUNT OF CARBON
REMOVED AS CARBON MONOXIDE ON THE GAS LOAD
DURING VACUUM TREATMENT WITH STEAM
EJECTORS(a)
Carbon Removed,
percent
Gas Load, equivalent air per hour
5-tons-per-minute Treating Rate 30-tons-per-minute Treating Rate
Pounds Cubic Feet Pounds Cubic Feet
0,00
0, 01
0, 05
13
156
760
161
1,933
9,416
80
940
4, 360
991
11, 647
54,020
(a) Treating rate is in terms of tons of steel per minute. Gas loading is based on removal of 4 ppm of hydrogen from the steel.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C -1 04
Manufacture of Semifinished Products
Semifinished products, as considered in this study, include ingots, billets,
blooms, and slabs. These are the intermediate shapes that are produced in the manu-
facture of the various end products of steel. However, even the semifinished products
as considered here are a marketable item accounting for 3.3 percent of American steel
shipments in 1967(8). Production of billets, blooms, and slabs from 1958 through 1967
is shown in Figure C-81. The data shown are based on the assumption that tube ~ounds,
light merchant shapes, reinforcing bar, joint bars, tie-plate bars, wire rods, and
forging billets were produced from billets; piling, rails, and heavy structural shapes
from blooms; and hot flat-rolled products and skelp from slabs. The production data
include manufacture by conventional ingot casting and rolling, by continuous casting,
and by pressure casting.
If) 70
c:
o
If)
+-
g 50
"0
o
~
a..
"0 40
Q)
L:.
(/)
:~ 30
......
-
'E 60
E
Q)
(f) 20
......
o
c:
o
---~----
Billets --...... --
. ---------
........---------- .-
10
+-
U
::J
"0
o
I-
a..
Blooms
----- --
-
-
o
1958
1959
1960
1961
1962
1963
Year
1964
1965
1966
1967
1968
FIGURE C-81.
ANNUAL PRODUCTION OF SEMIFINISHED STEEL
PRODUCTS
Ingot Casting and Rolling
The casting of ingots long has been the conventional method for converting molten
steel into a solidified shape suitable for further proces sing into semifinished products.
Many sizes, types, and designs of ingot molds are used in the production of ingots. The
particular type of mold is influenced by the end product to be made and by the chemistry
of the grade of steel cast. Steel is classified into three types: (1) killed steel, (2) semi-
killed steel, and (3) rimmed steel. Schematic illustrations of their respective ingot
structures are given in Figure C-82.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-I05
Hot top
r------l
I I
QQotOlVllvPIIVII",
\~ e' 0 6 Q " . ' 0 '" D'
o 'IIO'O~ ,
0'0:0,.0,1
Do, ~
0\.11'1 ,'"
00,.0',
, 6"00'
/JODO,\)\) ,.
P(J " ""00
"01100 '
di P 01) 0 ,,, ..
, '0 9000
" 000 0 '<:I 0
o 0
0011 00
~ 0
b '"
Po
Gas
holes
Killed Steel Ingot
Semikilled Steel Ingot
FIGURE C-82.
r--------'
I I
I ,
I I
J \
;' .....
:"~,, ;~OD
, 0 II 0
d D q'\> "0 00
, "0,0 I
II Q". 0 '11'00
:0::'09/10
, 0°0 ~
, ,,, 6
o 0 ,
II
o , 1:1
o ,,0 0
II ~
II 0
o 0
o 0
o ,
o ..
o :
o 0
o
o 0
o II
o '
o ,
o 0 0 Q
CI> D 0 't:.
c:::I" fJ ~
;;; Do"" " . ~
<==' C>
c::> ~
<=> "'"
~ 0
-;00 D6 00 00 OOOOOOgOo6lloai
Rimmed Steel Ingot
INGOT STRUCTURES FOR DIFFERENT TYPES OF STEEL
Kill"ed steels are generally used where a homogeneous structure is. required in the
finished product. Alloy steels, forging steels, and carburizing steels are typical ex-
amples of killed steels. Steels with more than 0.30 percent carbon are generally made
as killed steels. The term l'killed'l means that the molten steel has been thoroughly
deoxidized with various elements (frequently aluminum) so that it will be quiet when
poured into a mold. Ceramic hot tops are placed on the top of ingot molds poured with
killed steel. The function of the hot top is to delay solidification of the steel at the top
of the mold so as to supply molten metal to the solidifying ingot. If the hot top is not
used, the shrink cavity will penetrate deep into the body of the mold and will result in
less usable steel from the ingot. Steel sheet (as for example, automobile body stock)
conventionally is cast as rimmed steel.
Once the ingots have been poured, they are retained in the ingot mold for a spec-
ified period of time depending on the size of the ingot and the chemistry of the steel.
This period is usually referred to as "track time" because the ingots are usually
handled on special railroad flat cars. At the proper time, the ingots are removed from
the molds or the molds are removed from the ingots, depending on whether the ingots
are of "big-end-up" or "big-end-down" configuration. Ingots are next transferred to
soaking pits where they are heated to equalize the temperature throughout the ingot and
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-106
increase the temperature of the ingot to that required for rolling. The combustion
characteristics of the soaking pits are often controlled so that a certain thickness of
scale is developed on the surface of the ingots. This is done so that when the scale
breaks off during the first pass in the rolling operation, it takes with it certain minor
surface defects and the result is a better surface on the ingot. This practice results
in a major source of "mill scale".
Ingots, depending on their size and the desired end product, are rolled in blooming
mills or slab mills. Mills that perform both functions are often called universal mills.
The billets, blooms, and slabs that are the semifinished products in this stage of pro-
duction are cooled to ambient temperature and transferred to a storage area where
they are inspected, and where the surfaces are prepared or "conditioned" for the follow-
ing rolling operations.
Emis sion Identification. When molten steel is exposed to air, some type of fume
IS generated. This is particularly true during pouring when the molten steel is subjected
to turbulent action during the time it is flowing into the mold. Gaseous emissions occur
during the heating of ingots in the soaking pits, and a minor amount of particulates are
released during the primary rolling operations. A flow diagram illustrating the casting
and rolling operations is shown in Figure C-83, in which numbers in circles refer to
the source s of emis sion discus sed below.
CD
Fume - Minor amounts of iron oxide fume are generated in the tapping
of molten steel into ladles and during pouring of the molten steel from
the ladle into the ingot molds. Data concerning amounts, size, and
chemical composition are not available; however it can be assumed
that the particulates are of a very fine size and are primarily iron
oxide.
@
Gases from Mold Coatings (frequently carbonaceous) - Mold coatings
are applied to the inner surface of ingot molds to minimize certain
types of defects that are detrimental to the .quality of the final prod-
uct. One type is a nonvolatile coating that relies on surface texture
to accomplish the desired surface improvement. The other general
type is a coating that relies on volatilization to shield metal splashes
from the mold wall and produce an improved surface on the ingot.
The volatile type of mold coating include s coal-tar products,
petroleum derivatives, or naturally occurring materials such as
Gil sonite( 150).
Q)
Gases from Hot-Top Materials - Depending on the materials used to
make hot tops for use with ingots of killed steel, gases of varying
amounts and composition can be generated. Data on the chemical
compositions of the gases and their amounts are not available. Hot
tops can be divided into three classifications: (1) permanent, (2) in-
sulating, and (3) exothermic(l51-156) Permanent hot tops are made
of castable or preformed refractories. Amounts of gases generated
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
I~
~r
Slab,
pounds
STORAGE
C-107
Molten steel,
2020 pounds
Fume
LADLE
Solid ingot,
2000 pounds
(Heating)
SOAKING PIT,
2000 pounds
G)
o
o
~
G)
Scrap,
390 pounds
Atmosphere in vicinity
of operation
o
Combustion products
Atmosphere
Heated ingot,
2000 pounds
Mill scale,
10 pounds
Billets,
pounds
Scrap,
390 pounds
Scrap,
390 pounds
Bloom,
pounds
FIGURE C-83. CASTING OF STEEL INTO INGOTS; AND ROLLING TO SLABS, BLOOMS, AND BILLETS
STORAGE
(Numbers in circles refer to sources 'of emissions described in the text. )
BATTELLE MEMORIAL. INSTITUTE - COLUMBUS LABORATORIES
-------
C-108
by these in use are very small, and their compositions are harmless.
Insulating hot tops contain materials that have a tendency to char and
become porous. Smoke may be generated when molten steel comes in
contact with the insulating materials. Exothermic hot tops are gen-
erally composed of mixtures of aluminum, iron oxide, oxidizers, and
inert materials. Contact with molten steel starts the exothermic reac-
tion in the hot top, and causes generation of smoke and fumes. Some
compositions of exothermic hot tops cause copious fumes. Commer-
cial exothermic hot-top materials are proprietary and their composi-
tions are not publicized. Data on th.e amounts and chemical composi-
tion of emissions generated during pouring of ingots are not available.
Usually the emissions generated are treated as a nuisance, but their
contribution to air pollution can be substantial.
@)
Lead Fumes (Lead oxide fumes generated during the addition of lead
to free-machining steels) - Certain types of free-machining steels
contain lead which acts as a chip breaker during machining. The
lead is added in the form of shot during the time the molten steel
is poured into the ingot mold. Because lead fumes are considered
to be a health hazard, facilities for their collection usually are
provided where this type of steel is produced.
0)
Combustion Products (Carbon dioxide, carbon monoxide, nitrogen,
and sulfur dioxide) - Ratios of carbon dioxide to carbon monoxide
are determined by specific combustion practices in each steel
plant. The amount of sulfur dioxide is dependent mostly on the
amount of coke-oven gas used and on the degree t~ which sulfur has
been removed from the coke-oven gas. The use of coke-oven gas
as a fuel in the firing of soaking pits and the amounts used is deter-
mined by the energy availability and energy requirements in a spec-
ific steel plant. Tolerance of hydrogen sulfide in coke-oven gas
is determined to some extent by the types of steels produced and
whether the existing level of hydrogen sulfide is considered a factor
in air pollution. Some steel companies use a minimum amount of
coke-oven gas or desulfurize it before use( 107). Others use the
coke-oven gas as it is received from the coke plant without any
special desulfurizing. (12,49,157) If the sulfur content of the coal
used for making coke is low enough, the sulfur content of the coke-
oven gas is correspondingly low and the steel company does not
consider desulfurization to be necessary. (158)
Primary rolling operations do not contribute emissions to air pollution, unles s
steam generated from the flow of water in the primary mill is considered as an emis-
sion. Heavy iron-oxide scale (mill scale) that is formed on the ingots during heating
in the soaking pits is broken from the ingot during the first pas ses in the mill and is
removed from the mill by water which carries it to scale pits from whence it is further
transported to scale-recovery systems that are associated with the water-treatment
facilities of the steel plant.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-109
Ingot Casting and Rolling Emission Control. The emissions generated during the
pouring of ingots can be minor or major, depending on the practice in the particular
plant. When detrimental emissions are generated (such as in making leaded steels,
using volatile coatings on the molds, or using some exothermic hot tops) the necessary
facilities for exhausting the emissions should be provided. Bag houses sometimes are
used for these applications. To some degree, fuming of the steel during teeming can
be minimized by shrouding the stream of molten metal with an inert gas such as
argon. (159, 160) This technique has found use in the production of high-quality vacuum-
degassed steels, but the economiCs at this time do not favor its use for tonnage-steel
applications. In some electric-furna.ce shops where the entire building is considered
as an emission-control system, the fumes generated during ingot casting are ex-
hausted by the cornmon emission-control system for the entire operation. (139)
Continuous Casting
Continuous casting eliminates the need for ingot molds, soaking pits, and rolling
mills to produce billets, blooms, and slabs; and in some situations is thought to pro-
vide an economic advantage over the conventional method of casting ingots. High-
tonnage production of continuous-cast products is just starting in the integrated iron
and steel industry. The estimated annual production from 1962 through 1968 is shown
in Figure C-84 in comparison with the total product.ion of billets, blooms, and slabs
by all methods. It should be pointed out that the estimated actual production for 1968 of
100
95
Total production of
billets, blooms, and
slabs
III
g 90
+-
+-
Q)
c:
..... 85
o
III
c:
o
.- 80
:E
+-
U
::::J
"0
o
...
a..
10
5
Estimated continuous-:-cast "-
production -\..--"
---
. --
--
o
1962
1963
1964
1965
Year
1966
1967
1968
FIGURE C-84.
CONTINUOUS-CAST PRODUCTION AS COMPARED WITH TOTAL
PRODUCTION OF BILLETS, BLOOMS, AND SLABS
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-110
about 4. 5 million net tons of continuous-cast steel is far short of the 1968 estimated
capacity of 7 million tons. This difference is caused mainly by delays encountered in
the start-up of the larger continuous casting machines.
Emission Identification. The generation of emissions during furnace tapping and
filling of the ladle for continuous casting is the same as for conventional casting of in-
gots. However, the mechanics of the continuous-casting proces s and equipment design
tend to minimize emissions in the actual continuous-casting operation. A flow diagram
showing the points of emission in the continuous casting of steel is shown in Fig-
ure C-85, in which circled numbers refer to the following discus sion.
CD
Fume - Minor amounts of iron oxide fume are generated during the
filling of the ladle and during further handling of the molten steel
on its way to the continuous-casting machine. Data are not available
concerning amounts or size consist of the fume. Fume generated
during the actual pouring of the molten steel into the continuous-
casting machine is minimal. This can be attributed to the fact that
the molten steel is falling from the ladle to the tun-dish and finally
'into the continuous-casting mold over very short distances. This
reduces its exposure to the air to a very short period of time, and
minimizes the opportunity for iron oxide fume to form. The tun-
dishes sometimes are covered or blanketed with inert or reducing
gases which tend to further minimize the formation of oxide fumes.
@
Carbonaceous Gases - Lubricants are used in the continuous-casting
mold to prevent seizure between the solidifying steel and the mold.
Rape seed oil is the usual lubricant. Data concerning the amounts
and compositions of the gases generated are not available. However,
the amounts generated are so small that this emission is not con-
sidered as a problem by steel companies.
Q)
Fume - Oxide particles that are generated during the cut-off opera-
tion. No data are available on amounts. Chemical analysis would
show that the particulates are mainly iron oxide.
Continuous -Casting Emis sion Control. The continuous -casting proces s by its de-
sign tends to minimize the generation of emissions, and emission-control equipment,
as such, usually is not incorporated into the process. Some exhausting systems may
be employed to evacuate the steam that is generated in the secondary cooling system
by the impingement of water sprays on the hot steel( 161). Cutoff torches used in cut-
ting the continuous-cast steel into manageable lengths generate a small amount of
particulate emissions. In the production of plain carbon steels, this is not considered
a problem by steel companies. In a few installations in which- stainles s steels or free-
machining leaded steels are made, emission-control systems using filters are used to
collect the fumes generated(l6l). Many installations have hydraulic shears to cut con-
tinuous billets or blooms into the desired length. This in effect eliminates the emission
problem. In comparison to the casting of billets, the casting of large slabs creates a
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
Molten steel,
2060 pounds
CONTINUOUS-CASTING
MACHINE - SLABS
G)
o
o
Solidifying slab.
2013 pounds
Scale,
S pounds
Continuous solid slab.
2005 pounds
C-lll
LADLE
G)
o
CONTINUOUS -CAS TING
MACHINE - BILLETS AND BLOOMS
MECHANICAL SHEARS
Billet and bloom.
2005 pounds
FIGURE C-S5. CONTINUOUS CASTING OF STEEL
(Circled numbers refer to emissions discussed in the text. )
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-1l2
somewhat greater amount of fume in the cutoff operation involving torches, simply be-
cause a thicker cross section of steel must be cut. Shears could be used to cut the
continuously cast slab, but the high cost for such a shear usually is considered to make
this method of cutting economically unattractive. It has been estimated that the cost
of such a shear would be in excess of $1 million. (162)
. Pres sure Casting
Pres sure casting, like continuous casting, eliminates much of the equipment re-
quirements associated with conventional ingot casting and rolling. Pressure casting
usually is not economically competitive with continuous casting in the production of
billets and blooms, but in the casting of slabs it doe s appear to be competitive for some
situations. More information on the competitive situation will undoubtedly be forth-
coming when a 500, OOO-net-ton-per-year plant for pressure casting of carbon steels is
put into operation in 1969 in Oregon. Production figures for pres sure-cast slabs are
given in Table C-25(l63).
TABLE C-25.
PRODUCTION OF PRESSURE
CAST SLABS
Year
Annual Production, net tons
1964
1965
1966
1967
1968
35,000
50,000
70,000
85,000
150,000
Emission Identification. Like other steelmaking processes, fumes are generated
during the tapping of the furnace and filling of a ladle. From that point on, emissions
from pressure casting probably are less than for continuous casting. Points of emission
are shown in the flow sheet in Figure C-86, in which circled numbers refer to the fol-
lowing discussion.
-------
C-113
Q)
Fume - Iron oxide fumes generated during torch cutting of the
riser and its removal from the slab. These are considered
insignificant in contributing to an emission problem from the
steel plant, and can be readily collected by a suitable exhaust
system.
, Molten steel.
2150 pounds'
G)
o
TORCH CUT OFF
. FIGURE C-86. PRESSURE CASTING OF STEEL SLABS
(Circled numbers refer to emissions discussed in the text. )
Manufacture of Finished Products
The manufacture of finished products, for the purposes of this study, is con-
sidered in two steps. The first is the rolling of billets, blooms, and slabs into hot-
rolled and cold-rolled products. The second is the coating operations associated with
the production of finished products. .
Rolling Operations
Rolling operations for the production of finished products are generally carried
out on continuous mills. Billets are usually processed into products such as reinforcing
bar, hot-rolled bar, cold-rolled bar, wire for cold-heading operations (this may be as
large as 1-1/4-inches in diameter), high-quality bar products depending on the chemis-
try of the steel, and small angles and channels. Blooms are used to make larger
BATTELLE MEMORIAL INSTITUTE - COLUMBUS' LABORATORIES
-------
C-114
angles, channels, and structural products. These are normally furnished in the condi-
tion they are produced and without any further finishing or cleaning operations. Flat-
rolled products, on the other hand, are supplied hot-rolled, cold-rolled, tin-plated,
galvanized, terne-coated, painted, or plastic-coated to commercial users of the
products.
Hot-Rolling Emission Identification. As the hot-rolling of steel proceeds toward
finished products, the generation of air-polluting emissions is minor. Flow sheets for
the rolling of finished products are shown in Figure C-87 for bar and merchant products
from billets and blooms, and in Figure C-88 for flat-rolled products from slabs. Con-
ditioning of semifinished products is generally done at the primary-mill facilities before
billets, blooms, or slabs are delivered to the finish-rolling mills. However, this will
vary from plant to plant.
Conditioning at the primary mill is primarily spot conditioning that is done with a
variety of equipment. Hand grinding and chipping hammers are used along with special
equipment, known as "peelers", that mechanically removes defective portions of the
billets or blooms. Scarfing with hand torches is also done. The type of conditioning
treatment is determined by the chemical composition of the steel and by the quality re-
quirements specified by the consumer for the finished product. Grinding, chipping, and
peeling techniques do not generate emissions as considered in this study. Grinding
operations do produce some particles, but these are collected at the grinding station.
Generally these particles are of such a size that they do not become extensively air-
borne, and settle in the vicinity of the grinding operation. The following circled numbers
refer to points of emission as marked on Figures C'-87 and C-88.
CD
Fume - Iron oxide fume generated from hand scarfing of the billets
and blooms. No information is available concerning the size or the
amount of fume generated per net ton of billets or blooms. How-
ever, it has been reported that the loss in yield for scarfing billets
varies from 3 to 6 percent depending on the type of steel (164). Most
of this metal loss is in the form of metal splatter rather than fume.
Hand scarfing of slabs poses the same proht'ems as hand scarfing of
billets and blooms. However, hand scarfing of slabs usually is
limited to the removal of deep defects that would not be removed by
machine scarfing prior to rolling.
@
Products of combustion - Mixtures of carbon monoxide, carbon
dioxide, and nitrogen resulting from the firing of reheat furnaces
prior to rolling of billets, blooms, or slabs. The semifinished
products in this stage of manufacture are at a point where quality
control is getting to be a very large factor in the process. Re-
heating must be done under closely controlled conditions so the
semifinishe~ products are at the proper temperature for rolling
and are heated uniformly throughout the thicknes s of the product.
Natural gas is the usual fuel for firing the reheat furnaces) and,
under the closely controlled combustion practices, pollution is not
considered to be a problem. As the semifinished products proceed
toward a finished product, any losses in material contribute more to
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
BILLET AND BLOOM
STORAGE
(0
o
(0
Pickled
Grit
blasted
FIGURE C-87. TYPICAL HOT-ROLLING OF BAR AND
MERCHANT MILL PRODUCTS FROM
BILLETS AND BLOOMS
(Circled numbers refer to discussion
in text. )
C-115
SLAB STORAGE
8
CD
G)
Cold
rolling
FIGURE C-88. TYPICAL HOT-ROLLING OF SHEET AND
S TRIP FROM SLABS
(Circled numbers refer to discussion
in text. )
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-116
economic loss than those of any preceeding steps. Induction heat-
ing of large, continuous-cast slabs is being implemented at the
McLouth Steel Corp., to provide uniform heating of slabs as well
as to reduce the amount of scale formed on the steel. Needles s
to say, this method of heating steel for rolling is a pollution-free
method.
Q)
Fume - Iron oxide fume generated during machine scarfing of slabs,
blooms, or billets immediately prior to hot rolling. Fume is re-
ported to be generated at a rate of 2 to 3 grains per standard cubic
foot of gas, and the dust-collecting systems typically are designed
to operate between 75,000 and 135,000 C£m during the short dura-
tion of the scarfing operation(1, 101, 157). Several factors that enter
into the amount of metal that is removed during machine scarfing in-
clude: (1) speed of the semifinished product through the machine,
(2) oXYyen pres sure at the scarfing head, and (3) temperature of the
steel. ( 65) Metal removal by machine scarfing results in yield
reduction which can normally vary between 0.85 to 1.54 percent,
but can go as high as 2.5 percent for slabs, and as high as 7 per-
cent for blooms. (166-168) Machine scarfing is increasingly used
as a means for surface conditioning prior to hot rolling. The trend
in the tonnage of steel that is machine scarfed is shown in Fig-
ure C - 89. (168-170) It has been estimated that this trend will level
off at 50 percent of all tonnage, and that based on information re-
ceived from Europe, a considerable amount of continuously cast
steel will have to be machine scarfed. (169) Steel technologists in
the United States are directing a great deal of effort toward improving
the surface quality of continuously cast steel so as to eliminate the
need for extensive surface conditioning.
50
45
40
1968~
~
/
/
/
/
r
I
~ ""
.-
- V
-
35
30
....
~ 25
&. 20
15
10
5
o
1935
1940
1945 1950 1955
Year
1960 1965
1970
FIGURE C-89.
TREND IN INGOT TONNAGE OF STEEL THAT IS
MACHINE SCARFED
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-1l7
@)
Fume - Iron oxide fume generated during rolling in last few finishing
stands of a hot-strip mill. The emissions are submicron in size,
but no further information is available as to size distribution and
amounts. (157)
Hot-Rolling Emission Control. The principal emission associated with hot rolling
is iron oxide fume generated during the scarfing operation. Generally emis sion-
control equipment is not used with hand scarfing operations, because this is usually
an intermittent type of operation that is combined with chipping and grinding. How-
ever, some steel mills, especially those concerned with meeting increasing customer
quality requirements, may scarf almost all of the billets and blooms by hand. In such
situations, exhaust hoods are used to remove the fumes from the scarfing area. Infor-
mation concerning the collection and/ or disposal of the fumes and dust collected is not
available.
Fume-control from machine scarfing of slabs varies from no control to complete
control. (1,12,49,101,157,158) Both electrostatic precipitators and high-energy
scrubbers are used to collect the fume generated during machine scarfing. (1,157,158)
The small amount of iron oxide fume generated in the finishing stands of a hot
mill is reported to be collected in one plant by a high-energy scrubber. (157)
Cold-Rolling Emission Identification. Particulate and gaseous emissions asso-
ciated with preceding steps in the manufacture of steel are not encountered in the cold-
rolling operation. The major emis sions associated with cold-rolling operations are
acid fumes generated in the pickling of hot-rolled steel strip or dust resulting from
mechanical cleaning operations that may be used to prepare the surface of hot-rolled
strip for cold rolling. A flow sheet illustrating the cold-rolling process is shown in
Figure C-90, in which circled numbers refer to the following discussion.
Pickling of hot-rolled strip is required to provide a clean metal surface for
cold rolling.
CD
Acid Fumes - Either sulfuric or hydrochloric ac.id fumes depending
on the individual plant situation with respect to product require-
ments and availability and cost of acid. Sulfuric acid has been used
traditionally for pickling operations to prepare hot-rolled steel for
cold rolling. The increasing cost of sulfuric acid, combined with
problems of disposal of spent sulfuric acid liquor has prompted a
trend to hydrochlo ric acid pickling. (171, 172) It has been estimated
that, by 1970, hydrochloric acid and sulfuric acid will each be used
for 50 percent of the steel-pickling operations. (171)
~
Water-Oil Mist - Mixtures of water with water-soluble oils generate
mists during the cold- rolling operations.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-1l8
Hot -rolled coils
ACID PICKLING TANK
G)
Re -coiled hot -rolled
strip
Pickled hot -rolled
coils
COLD -ROLLING MILL
o
FIGURE C-90. COLD ROLLING OF STRIP STEEL
(Circled numbers refer to discussion in text. )
Cold-Rolling Emission Control. Acid fumes from the pickling tanks generally
have been controlled by steel plants. The pickling lines are hooded and exhausted to
fume-control systems. Wet scrubbers and packed towers are used to remove the acid
mists from the pickling tanks. (1,13,101,157) Where tonnage requirements do not
economically warrant the installation of acid-fume collection system1 the fumes may
be exhausted to the atmosphere by building roof-exhaust systems. (15 ) Collection of
mists generated by cold-rolling operations is done with mechanical mist eliminators or
with wet s c ru b be r s ( 1 57, 1 73) .
Coating of Finished Products
Surface coating of steel fo r protection or appearance is tending toward continuous-
line proces sing so that operating economies can be realized. The continuous lines also
lend themselves to control of emissions that may be generated. A listing of the number
of continuous and batch units in the integrated iron and steel industry is given in
Table C-26. In addition to the facilities listed, there are 3 batch-galvanizing facilities
for pipe and 15 continuous-galvanizing facilities for wire. The various facilities are
located in 43 steel plants throughout the country. Annual production of several different
types of coated products from 1958 through 1967 are shown in Figure C-91. The produc-
tion of hot-dipped tin and terne plate which are made primarily in batch operations that,
from pollution aspects, are more difficult to control, has decreased from 450,000 net
tons in 1958 to 30,000 net tons in 1967.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
(/)
c:
0
-
-
Q)
c:
'0
(/)
c:
0
E
~
c:
0
+=
0
::J
"0
e
a..
O.
O.
o.
o.
o
C-1l9
12.0
/ -
/
Total /
J~
....... /
'" I
Electrolyte tin plate ,-
/ ...--..., ---- --- ~"
-'--I.... ,
/' ~"".. " ~ ""
'.... I
', I ~
~
/" Galvanized sheet
----
.,.:> >
5
~.... -,',
4 ....
~ \Hot-dipped tin and terne plate
\
3 \
,
"', k? ~ ~
'.... -
2 Long terne ~ --~ ~
/" " ---
--- -," "''''....
.1 -
-'~ "
~ Other metallic coated ...
I I I ---
11.0
10.0
9.0
8.0
7.0
6.0
5.0
4.0
3.0
2.0
0.0
1958
1962 1963
Year
1965 1966
1967 1968
1964
1959
1960 1961
FIGURE C-91. PRODUCTION OF SHEET STEEL COATED PRODUCTS
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C -120
TABLE C-26.
SURFACE-TREA TMENT FACILITIES FOR SHEET
PRODUCTS IN THE INTEGRA TED IRON AND
STEEL INDUSTRY
Type of Facility
Continuous
Batch
Acid pickling
Galvanizing
Electrolytic tin plating
Long terne sheets
Aluminum coating
Chromium coating
Nickel coating
Copper coating
Painting
Total (Excluding pickling)
124
62
40
6
2
4
2
2
6
42
12
15
6
2
o
o
o
o
35
124
Emission Identification. Emissions associated with the surface-treatment lines
are not the particulate and gaseous emissions associated with other steelmaking prac-
tices. Emissions originate mainly from the acid-pickling operations and from the coat-
ing operations. Many of the coating processes are mechanically similar - the differ-
ences being in the type of coating applied. Processing is similar for the batch opera-
tions and for the continuous operations, except that the continuous lines are designed to
perform the coating operation around the clock rather than on an intermittent, stop-
and-go basis as for batch treatments. The continuous operations usually lend them-
selves to better control of emissions than do the batch operations. A flow sheet for
typical galvanizing operations is shown in Figure C-92, in which circled numbers refer
to the following discussion. .
CD
Products of Combustion - Normal mixtures of carbon monoxide,
carbon dioxide, and nitrogen as sociated with the combustion of
natural gas in heat-treating furnaces.
@
Acid Fumes - Sulfuric or hydrochloric acid fumes. One steel plant
has reported that hydrochloric acid fumes are emitted to the atmo-
sphere from a tower scrubber at a rate of 0.5 gallons per hour when
treating steel at the rate of 100 tons per hour. (174) This rate of emis-
sion is equivalent to 108 grains per ton of steel treated.
Q)
Flux Emissions - Emissions generated from the cover flux which can
be ammonium chloride or zinc ammonium chloride. No data are avail-
able on the chemical composition of the emissions. However, data
from galvanizing job shops provided the compositions given in Ta-
ble C-27. (175) In the absence of data from iron and steel industry
installations, these data may be considered as typical of emissions
from batch operations in steel-plant facilities. Particulates from
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-121
batch galvanizing installations have an average particle size of
about 2 microns as they evolve from the pots, but have a tendency
to agglomerate. A size analysis of par'ticulates taken from a bag
house showed that 23 percent by weight of the particulates were
greater than 250 microns in size. (175)
(2)
(2)
HEA T -TREA TING
FURNACE
HEA T -TREA TING
FURNACE
G)
o
G)
o
FIGURE C-92. TYPICAL GALVANIZING PROCESSES
G)
HEAT-TREATING
FURNACE
G)
(Circled numbers refer to discussion in text. )
TABLE C-27.
CHEMICAL COMPOSITION OF GAL V ANIZING EMISSIONS
Component
Source One, weight percent
Source Two, weight percent
H20
ZnCl2
ZnO
Zn
NH4CL
NH3
Oil
Carbon
Not identified
2.5
3.6
15.8
4.9
68.0
1.0
1.4
2.8
0.0
100.0
1.2
1-5.2
6.5
0.0
23.5
3.0
41. 4
0.0
9.2
100.0
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
,(
-------
C -122
Coating installations for' the application of other metals', paints,. and plastics are
essentially the same as those illustrated in the flow sheets in Figure C-92. They con-
sist of a preparation section followed by plating, dipping, or spraying, which is fol-
lowed by cleaning and drying operations.
Auxiliary Operations
Two auxiliary operations considered in this study are (1) the foundry facilities
associated with steel-plant operations, and (2) the incineration facilities required for
the disposal of solid combustible wastes generated in the manufacture of iron and steel.
Foundry Facilities. Foundry facilities associated with steel plants are primarily
used to supply needed replacement parts as a maintenance function. Some companies
have facilities for making their own ingot molds (neces sary for the conversion of molten
steel to solidified steel) and rolling mill rolls (required to make semifinished and fin-
ished products). Foundry facilities for steel plants in the United States are given in
Table C-28, with data on ~elting facilities and types of metals melted. (176)
Three major companies that supply ingot molds to the steel industry are:
(I) Shenango Incorporated, with plants at Neville Island and Sharpsville, Pennsylvania,
and Buffalo, New York, (2) Valley Mold and Iron Corporation, with plants at Chicago,
Illinois, Cleveland, Ohio, and Hubbard, Ohio, and (3) Vulcan Mold and Iron Company,
with plants at Lansing, Illinois, Latrobe, Pennsylvania, and Trenton, Michigan.
Emission Identification. Emissions from the aluminum-melting operations are
generally limited to chlorides that are generated by the use of chlorine gas in the re-
moval of gases from the molten aluminum. Emissions generated from the melting and
casting of brass are generally limited to zinc oxide fume generated from the oxidation
of zinc in the alloy. No specific data are available on emissions from aluminum and
bras s melting, but it is reasonable to expect that they have the same characteristic s
as those generated in the respective foundry operations in ordinary commercial
operation.
Specific data on emissions from cupolas also are not available with respect to
specific steel-plant operations. Again,. it can be expected that the emissions generated
will have the general characteristics of emissions generated in commercial gray iron
foundry operations. It is understood that a project has been initiated by the National
Air Pollution Control Administration to investigate the problems of iron and steel
foundry emissions in detail. An indication of the size characteristics of particulates
emitted from gray iron cupolas is shown in Table C-29. These data are from litera-
ture sources on the gray iron foundry industry. (179,180) Chemical analyses of par-
ticulates from a cupola are given in Table C-30. (181) The combustibles noted in
Table C-30 are primarily particles of coke that are blown from the cupola. The many
operating and material variables in cupola melting influence the characteristics of
emissions from gray-iron melting cupolas. These variables include (1) the ratio of
coke to metallics charged,. (2) characteristics of the coke, (3) type of scrap, (4) clean-
liness of scrap, and (5) the rate at which air is blown into the cupola.
BATTELLE MEMORIAL INSTITUTE - COLl:JMBUS LABORATORIES
-------
C-123
TABLE C-28. FOUNDRY FACILITIES LOCATED AT STEEL PLANTS IN THE UNITED STATES
Company Plant Metals Melted Special Items Melting Facilities
Armco Steel Corp. Ashland, Ky. Brass, aluminum,
gray iron
Torrance, Cal. Steel
Bethlehem Steel Co. Bethlehem, Pa. Brass, gray iron, Three 25-tOn electric
ductile iron induction furnaces
Foundry and Machine Gray iron and ductile Rolls
Division, iron
Bethlehem, Pa.(177,178)
JohnstOwn, Pa. Brass, gray iron
CF & I Corp. Pueblo, Colo. Brass, aluminum, gray
iron, steel
Crucible Steel Corp. Midland, Pa. Brass, gray iron, steel
Ford MotOr Co. Dearborn, Mich. Gray iron Ingot molds
Granite City Steel Co. Granite City, ill. Gray iron Ingot molds
Inland Steel Co. East Chicago, Ind. Gray iron Ingot molds
Jones & Laughlin Pittsburgh, Pa. Brass, aluminum, gray Two cupolas, 9 tons per
Steel Corp. iron, steel hour each
Kaiser Steel Corp. Fontana, Cal. Brass, aluminum, gray Ingot molds
iron, ductile iron, steel
Lone Star Steel Co. Lone Star, Texas Gra y iron Ingot molds Two cupolas, 30 tons per
hour each
Repu bUc Steel Corp. Gadsden, Ala. Brass, aluminum, gray Two cupolas, 10 tOns per
iron, steel hour each
United States Steel Roll and Machine Works, Gray iron, ductile iron Rolls Two cupolas, 9 tons per
Corp. Canton, Ohio hour each
South Chicago, ill. Aluminum, gray iron, One cupola, 10 tons per
ductile iron hour
Gary, Ind. Gray iron, ductile iron,
steel
Fairfield, Ala. Brass, aluminum, gray
iron, steel
Johnstown, Pa. Gray iron, ductile iron, One cupola, 12 tons per
steel hour
Lorain, Ohio Gray iron, steel One cupola, 9 tons per
hour
Braddock, Pa. Gra y iron Ingot molds
Provo, Utah Brass, aluminum, gray
iron, steel
Youngstown Sheet and Indiana Harbor, Ind. Gra y iron
Tube Co.
BATTELLE MEMORIAL INSTITUTE '- COLUMBUS LABORATORIES
-------
C -124
TABLE C-29. SIZE CHARACTERISTICS OF PARTICULATES EMITTED FROM GRAY IRON CUPOLAS
Particle Size Screen Size, Test I, Test 2,
Range, microns Weight Percent microns weight percent weight percent
o - 4 4 - 10 +833 0.7 2.9
5 - 9 2 - 15 +252 12.0 19.6
10 - 24 4 - 15 +147 18.1 23.8
25 - 49 5 - 15 +97 16.6 17.5
50 - up 45 - 85 +74 9.3 8.8
+47 12.7 11.5
-47 30.6 15.9
TABLE C-30. CHEMICAL ANALYSIS OF PARTICULATES
FROM A GRAY IRON CUPOLA
Component Range, percent
Si02 20 - 40
CaO 3 - 6
Al203 2 - 4
MgO 1 - 3
FeO, Fe203' Fe 12 - 16
MnO 1 - 6
Combustibles 20 - 50
Emissions generated in the melting of steel for castings in a steel plant are the
same as those described in the sections on steelmaking emissions, because the steel
used for castings is made by the facilities already located in the steel plants. Infor-
mation on these emissions can be found on pages C-50 through C-93 of this Appendix.
Incineration. The large amounts of miscellaneous combustible solid materials
that must be disposed of in a steel plant create a disposal problem. In older steel
plants, several small incineration facilities may be located in various parts of the
steelworks complex to handle the disposal of wastes. However, with the emphasis now
placed on the control of emissions, many of the multiple-site incineration facilities
probably will be consolidated into one major incineration facility with the required
emis sion-control equipment as part of the facility. This trend is taking place with the
new plants under construction and during modernization of existing steel plants. No
specific data are available on emissions generated in steelworks incinerating operations.
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
C-125
REFERENCES FOR APPENDIX C
(1)
Ziercher, J. L., Report on Official Travel, Battelle Memorial Institute, Columbus
Laboratories, October 4, 1968.
(2)
Tsujihata, K., et al., "Developments in Ironmaking at Yawata Iron and Steel Co. ,
Ltd. ", Blast Furnace and Steel Plant, 53 (3), 242-248 (March 1965).
(3) Send, A., and Wimzer, G., "Trends in Burden Preparation and Their Effect on
Blast Furnace Operation in Germany", Journal of Metals, 19 (7), 58-64 (July
1967). -
(4) Knepper, W. A., and Sciulli, C. M., "Operation of an Experimental Blast Furnace
with Sized Coke", AIME Ironmaking Proceedings, 24,38-40 (1965).
(5 )
White, R. H., "The Effect of Coke Sizing on Blast Furnace Operation", Blast
Furnace and Steel Plant, 54 (3), 241-245 (March 1966).
(6) White, R. H., and Meyer, V., "Blast Furnace Operation With Washed Burden
Materials", Journal of Metals, ..!:1 (6), 52-54 (June 1967).
(7) Nitchie, C. M., "Effect of Screened and Sized Sinter on Blast Furnace Operation",
AIME Ironmaking Proceedings, 26, 15- 19 (1967).
(8) Annual Statistical Report, American Iron and Steel Institute, New York, N. Y.
(1962, 1967).
(9) Harris, E. R., and Beiser, F. R., "Cleaning Sinter Plant Gas With Venturi
Scrubber", Journal of the Air Pollution Control Association, 15 (2), 46-49
(February 1965). -
(10)
Baranyi, J. F., "Results of Design Changes in Sinter Plant Fans", Iron and Steel
Engineer, 42 (12), 85-90 (December 1965).
(11) McCrone, W'. C., et al., "The Particle Atlas", Ann Arbor Science Publishers,
Inc., Ann Arbor, Michigan (1967), pp 153-236.
(12) Young, P. A., et al., II The Generation and Treatment of Sinter Plant Dusts",
AIME Blast Furnace, Coke Oven and Raw Materials Proceedings, 20, 299-313
(1961). -
(13)
Ziercher, J. L., Report on Official Travel, Battelle Memorial Institute, Columbus
Laboratories, October 17, 1968.
(14) "Symposium on Sinter P\ants", Iron and Steel Engineer, 36 (6), 101-122 (June
1959).
(15) Chapman, H. M., "Experience With Selected Air Pollution Control Installations in
the Bethlehem Steel Company", Journal of the Air Pollution Control Association,
~ (12), 604-606 (December 1963).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
(16)
(17)
( 18)
(19)
(20)
(21 )
( 22)
(23)
(24)
(25)
(26)
(27)
(28)
(29)
(30 )
C-126
Frame, C. P., "The Effects of Mechanical Equipment on Controlling Air Pollu-
tion at No.3 Sintering Plant, Indiana Harbor Works, Inland Steel Company",
Journal of the Air Pollution Control Association, 13 (12), 600-603 (December
1963). -
Young, T. A., "Gary Steel Works Experience With Dust Control at Number 3
Sinter Plant", Preprint, Blast Furnace and Coke Plant Association Meeting,
Chicago, Illinois, October 4, 1968. 18 pp.
"Iron Ore News Highlights qf 1967", American Iron Ore Association, January 15,
1968. pp. 20- 23.
Behrens, et al., "The Effects of Lime Properties on Basic Oxygen Steelmaking",
AIME O~en Hearth Proceedings, 48, 74-82 (1965).
Tartaron, F. X., and Ruschak, J. D., "Effect of Lime Structure in Oxygen Steel-
making", Report of Investigations 6901, U. S. Bureau of Mines, 1967. 41 pp.
"Lime for Steelmaking: Tailored to Fit New Demands", 33/The Magazine of
Metals Producing, ~ (1), 58-93 (January 1967).
Thomas, F. H., "The Production of Blast Furnace and Sinter Plant Flux and
BOF Lime", Journal of Metals, .!.2 (6). June 1967. pp. 86-89.
Graf, D. L., and Lamar, J. E., "Properties of Calcium and Magnesium Car-
bonates and Their Bearing on Some Uses of Carbonate Rocks", Economic Geology,
Fiftieth Anniversary Volume, Part II, The Economic Geology Publishing Company,
(1955), pp. 639-696.
Boynton, R. S., and Gutschick, "Lime", Industrial Minerals and Rocks, AIME
Seeley W. Mudd Series, Third Edition (1960), pp. 497-517.
"New Lime Plant Begins Production", Iron and Steel Engineer, 42 (12), 162
(December 1965).
MacNamara, J., "Lime Plant at Algoma", Iron and Steel Engineer, 44 (10), 126-
130 (October 1967).
"Republic Steel Plans Limestone Plant on Lake Erie", Mining Engineering, 19
( 1 0 ), 2 9 - 3 2 (0 c to b e r 1 967 ) .
"CF & I Smoking With a Filter and the Air's a Pure Delight", Metalworking News,
p. 14 (October 21, 1968).
Uys, J. M., and Kirkpatrick, J. W., "The Beneficiation of Raw Materials in the
Steel Industry and Its Effect Upon Air Pollution Control", Journal of the Air Pollu-
tion Control Association, g (1), 20-32 (January 1963).
Ramm, A. N., "Minimum Theoretically Possible Coke Oven Consumption for
Pig Iron Production Under Modern Conditions", Stal, No. 10, pp. 860-871
(October 1964).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
(31 )
(32)
(33 )
(34)
(35)
(36)
(37)
(38)
(39)
(40)
(41)
(42)
(43)
(44)
(45)
C-127
Nakatani, F., et al., "Theoretical Consider.a tion on Blast Furnace Coke Rate",
Transactions of the Iron and Steel Institute of Japan, ~, 263-280 (1966).
"Practical Suggestions for the Reduction of Emissions, Dust, and Grit at Coke
Ovens", Special Publication No.5, British Coke Res earch As sociation,
Chesterfield, Derbyshire. (1962) 12 pp.
Herrick, R. A., and Benedict, L. G., "A Microscopic Classification of Settled
Particulates Found in the Vicinity of a Coke-Making Operation", Paper No. 68-
137, Annual Meeting of the Air Pollution Control Association, St. Paul, Minnesota.
June, 1968. 23 pp.
Sellars, J. H., and Hornsby-Smith, M.P., . "Smoke Emissions During the Charging
of Coke Ovens", Coke and Gas, 23, 411-420 (1961).
Fullerton, R. W., "Impingment Baffles to Reduce Emissions from Coke Quenching",
Journal of the Air Pollution Control Association, 17 (12), 807-809 (December
1967). -
Schapiro, N., 'and Gray, R. J., "Petrographic Classification Applicable to Coals
of All Ranks", Proceedings of the Illinois Mining Institute, 68, 83- 97 (1960).
Schapiro, N., et al., "Recent Developments in Coal Petro'graphy", AIME Blast
Furnace, Coke Oven, and Raw Materials Proceedings, 20, 89-109 (1961).
Benedict, L. G., and Berry, W. F., "Further Applications of Coal Petrography",
American Conference on Coal Science, Advances in Chemistry Series, Vol. 55,
American Chemical Society (1966), pp. 577-601.
Thompson, R. R., et al., "The Use of Coal Petrography at Bethlehem Steel
Corporation", Blast Furnace and Steel Plant, 54 (9), 817-824 (September 1966).
Bayer, J. L., and Denton, G. H., "Applications of Coke Microscopy.to Plant
Problems I', Blast Furnace and Steel Plant, 54 (12), 1133-1142 (December 1966).
Denton, G. H., et al., "Progress and Problems in Routine Petrographic Evalua-
tion of Coals for Coke Plant Use", Journal of Metals; .!.1 (5), 88-92 (May 1967).
Benedict, L. G., et al., "Relationship Between Coal Petrographic Composition
and Coke Stability", Blast Furnace and Steel Plant, 56 (3), 217-224 (March, 1968).
Thompson, R. R., and Benedict, L. G., "Goals, Accomplishments, and Limita-
tions of Petrographic Methods of Coal Evaluation'l, Journal of Metals, 20 (3), 79-
84 (March 1968). -
"McWane Company to Market Merchant Iron", American Metal Market, October 13,
1966.
Ban, T. E., and Violetta, D. C., "D-LM New Commercial Ironmaking Process",
Iron and Steel Engineer, 45 (9), 101-111 (September 1968).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
(46)
(47)
(48)
(49)
(50)
(51 )
( 52)
(53)
( 54)
(55)
(56 )
(57)
(58)
(59)
(60)
(61 )
C-128
"Gilmo.re Steel Corporation to Construct Integrated Steel Producing Plant in the
Pacific Northwest", Blast Furnace and Steel Plant, 55 (12), 1123 (December 1967).
"Some Economic Tests Coming at Oregon Steel", Steel, 163 (14), 40 (September 30,
1968).
"Phase II of Burns Harbor Project Started", American Metal Market, (March 15,
1968), p. 5.
Hoffman, A. 0., Report on Official Travel, Battelle Memorial Institute, Columbus
Lab.oratories, October 23, 1968.
Heynert, Von G., et al., "Charge Preparation and Its Effect 'on Operating Results
of the Blast Furnace", Stahl und Eisen, ~ (1), 1-12 (January. 5'., 1961).
"The Making, Shaping, and Treating of Steel", Eighth Edition, .1964, United States
Steel Corporation, Pittsburgh, Pennsylvania.
"Dust Recovery Practice at Blast Furnaces", Steel Industry Action Committee,
Ohio River Valley Water Sanitation Commission, 36 pp. (January, 1958).
Gaffney, L. J., and Holowaty, M. 0., "Inve stigation of the Effects of Burden
Constituents on Blast Furnace Refractory Linings'., AIME Ironmaking 'Proceedings,
~, 11 - 14 (1966).
Hipp, N. E., and Westerholm, J. R., "Developments in Gas Cleaning - Great
Lakes Steel Corp. ", Iron and Steel Engineer, 44 (8), 101-106 (August 1967).
Carney, D. J., et al., "Continuous Analysis of Iron Blast Furnace Top Gas",
AIME Blast Furnace, Coke Oven, and Raw Materials Proceedings, ~, 142-157
(1954).
Woehlbier, F. H., and Rengstorff, G. W. P., "Preliminary Study of Gas Formation
During Blast-Furnace Slag Granulation With Water", Paper No. 68-136, Annual
Meeting of the Air Pollution Control Association, St. Paul, Minnesota, June 1968.
Weise, W. H., "Blast Furnace Flue Dust Treatment Facilities", Sewage and
Industrial Wastes, 28, 1398-1402 (November 1956).
Ess, T. J., "Weirton Steel Company", Iron and Steel Engineer, 35 (11), W-42 -
W-61 (November 1958).
Ess, T. J., "United States Steel's Geneva Works'., Iron and Steel Engineer, 36
(6), G-2 - G-27 (June 1959).
Ess, T. J., "Kaiser Steel - Fontana Plant", Iron and Steel Engineer, 38 (2), K-1 -
K-23 (February 1961).
Crawford, C. C., "CF & I Steel at Pueblo", Iron and Steel Engineer, ~ (4), P-2 -
P-27 (May 1962).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
(62)
(63)
(64)
( 65)
(66)
(67)
(68)
(69)
(70)
(71 )
(72)
(73)
(74)
(75)
(76 )
(77)
C-129
Longenecker, C., and Lassen, E. G., IIJones and Laughlin Rebuilds and Expands
Purchased Plant, and Converts It Into a Modern Producer", Blast Furnace and
Steel Plant, ~ (8), 652-668 (August 1963).
Brady, J. L., IIAmanda Blast Furnace", AIME Ironmaking Proceedings, ~ (1964).
Sieger, E. W., and Dean, A. F., IINo. 1 Blast Furnace Reline'l, Iron and Steel
Engineer, 42 (2), 105-113 (February 1965).
Jewell, C. J., IIOver the 4000 Ton Barrier - Sparrow s Point' J'II, Blast Furnace
and Steel Plant, ~ (1), 59-63 (January 1968).
IINew Blast Furnace Operating at Indiana Harbor", Iron and Steel Engineer, 45
(1),161 (January 1968).
Stone, J. K., IIL-D Steelmaking at Mid-1967'1, Journal of Metals, .!1. (7), 10
( J ul Y , 19 6 7) .
Sims, C. E., IIWhat is Ahead in the Next 25 Years for Electric Furnace Steel-
making", Journal of Metals, 20 (2), 44-50 (February 1968).
Varga, J., Jr., Communication with the American Iron and Steel Institute,
November 7, 1968.
Bishop, C. A., et aI., IISuccessful Cleaning of Open-Hearth Exhaust Gas With a
High-Energy Scrubber", Journal of the Air Pollution Control Association, 11 (2),
83-87 (February, 1961). -
Schneider, R. L., IIEngineering, Operation and Maintenance of Electrostatic
Precipitators on Open Hearth Furnaces", Journal of the Air Pollution Control
Association, ~ (8), 348-353 (August 1963).
Zimmer, K. 0., IIDust-Laden Waste Gases Emitted in the Basic Open Hearth
Process With the Usual Melting Practice and With Oxygen Injection; Also Re-
moval of the Dust From Them", Stahl und Eisen, 84 (17), 1070-1075 (1964).
Elliot, A. C., and Lafreniere, A. J., IICollection of Metallurgical Fumes From
Oxygen Lanced Open Hearth Furnaces'l, Journal of Metals, 18 (6), 743-747
June, 1966. -
Smith, W. M., and Coy, D. W., IIFume Collection in a Steel Plant", Chemical
Engineering Progress, 62 (7), 119-123 (July 1966).
Broman, C. U., and Iseli, R.
With Venturi Type Scrubber",
(February 1968).
R., liThe Control of Open Hearth Stack Emis sions
Blast Furnace and Steel Plant, ~ (2), 143-148
Thomas, F. A., IIOxygen Lancing Means Open Hearth Waste Gas Cleaning Sys-
tems, II. Venturi Gas Scrubbers", Journal of Metals, ..!...2. (3), 264-266 (March
1965).
Jackson, A., IIFume Cleaning in Ajax Furnaces", Fume Arrestment, Special
Report 83, The Iron and Steel Institute, pp. 61-64 (1964).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
(78)
(79)
(80)
(81)
(82)
(83)
(84)
(85)
(86 )
(87)
(88)
(89)
(90)
(91 )
(92)
(93)
C-130
Akerlow, A. K., "Modification to the Fontana Open Hearth Precipitators",
Journal of the Air Pollution Control Association, 7.. (1), 39-43 (May 1957).
"Smoke Eaters for Nine Fiery Furnaces", Air Repair, ~ (3), 113 (February 1953).
Akerlow, A. K., "Design and Construction of Fontana Open Hearth Precipitators",
Iron and Steel Engineer, 34 (6), 131-138 (June, 1957).
Herrick, R. A., "A Baghouse Test Program for Oxygen Lanced Open Hearth
Fume Control", Journal of the Air Pollution Control As sociation, 13 (1), 28-32
(January 1963). -
Herrick, R. A., et al., "Oxygen-Lanced Open Hearth Furnace Fume Cleaning
With a Glass Fabric Baghouse", Journal of the Air Pollution Control As sodation,
~ (1), 7-11 (January 1966).
Speer, E. B., "Operation of Electrostatic Precipitators on O. H. Furnaces at
Fairless Works", Air and Water Pollution in the Iron and Steel Industry, Special
Report No. 61, The Iron and Steel Institute. (1958), pp. 67-74.
Dickinson, W. A., and Worth, J. L., "Waste Gas Cleaning System at Sparrows
Point Plant's No.4 Open Hearth", AIME Open Hearth Proceedings, 47, 214-225
(1964). -
Johnson, J. E., "Wet Washing of Open Hearth Gases", Iron and Steel Engineer,
44 (2), 96-98 (February, 1967).
"L-D Process Newsletter", Kaiser Engineers Division of Kaiser Industries,
Newsletter No. 46, October 21, 1968.
"Annual Statistical Report, American Iron and Steel Institute, 1967", American
Iron and Steel Institute, New York City, 1968.
Gaw, R. G., "Gas Cleaning", Iron and Steel Engineer, 22. (10), 81-86 (October
1960) .
Koenitzer, F., and Zimmermann, K. A., "Gas Cooling in the Oxygen Steelworks
of August Thyssen-Hutte A. G. ", Engineering Experience in Oxygen Steelworks,
Publication 98, The Iron and Steel Institute. (1966), pp. 8-9.
Rengstorff, G. W. P., "Factors Controlling Emissions From Steelmaking Pro-
cesses", AIME Open Hearth Proceedings, 45, 204-219 (1962).
Kosaka, M., and Minowa, S., "Effect of Rate of Carbon Elimination Upon the
Formation of Oxide Fumes in Oxygen Blowing", Tetsu-to-Hagane, 50 (11), 1735-
1738 (1964). -
Watkins, E. R., and Darby, K., "The Application of Electrostatic Precipitation
to the Control of Fume in the Steel Industry", Fume Arrestment, Special Report
83, The Iron and Steel Institute (1964), pp. 24-35.
McShane, W. P., and Bubba, E., "Automatic BOF Stack Monitoring", 33/The
Magazine of Metals Processing, ~ (5), 97-104 (May 1968).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
(94)
(95)
(96)
(97)
(98)
(99)
( 100)
( 101)
(102)
(103)
(104)
(105)
( 106)
( 107)
( 108)
( 109)
C-131
Kalling, B., et al., "Metallurgical Characteristics of the Kaldo Oxygen Steel-
making Proces s", Preprint, AIME Open Hearth Meeting, Chicago, Illinois
(April 1960), 19 pp.
Yocom, J. E., and Chapman, S., "The Collection of Silica Fume With a Venturi
Scrubber", Air Repair, ! (3), 155-158 (November 1954).
Trenkler, H., and Hauttmann, H. F., "LD-Process of Steelmaking With Oxygen
Jet", Metals Progress, 69 (1), 49-56 (January 1956).
Behrendt, A., "Gas Cleaning in Relation to Oxygen Pre-Refining and the Rotor
Proces s at Oberhausen", Air and Water Pollution in the Iron and Steel Industry,
Special Report 61, The Iron and Steel Institute (1958), pp. 90-96.
Smith, J. H., "Air Pollution Control in Oxygen Steelmaking", Journal of Metals,
~ (9), 632-634 (September 1961).
Loughrey, D. R., "The Basic Oxygen Process at Jones and Laughlin", AIME
Open Hearth Proceedings, 42, 274-285 (1959).
Massobrio, G., and Santini, F., "Some Starting and Operating Experiences With
the 300-Ton Oxygen Furnaces at the Taranto Works", AIME Open Hearth Pro-
ceedings, 48, 115-119 (1965).
Ziercher, J. L., Report on Official Travel, Battelle Memorial Institute,
Columbus Laboratories, October 10, 1968.
Dormsjo, T. 0., and Berg, D.
Its Metallurgy and Economics",
1959) .
R., "The Kaldo Oxygen Steelmaking Proces s -
Iron and Steel Engineer, 36 (4), 67-77 (April
Krijgsman, M., "Recovery and Utilization of Dust From the Basic Oxygen Steel-
making Processl1, Blast Furnace and Steel Plant, ~ (4), 44-62 (April 1964).
Rudnitskii, Ya. N., et al., "Determining the Amount of Gases Evolved in Con-
verters and the Time for Which Oxygen Should be Injected in an Oxygen-Blown
Converter", Stal, No.1, pp. 15-20 (January 1968).
Vajda, S., "Symposium on Basic Oxygen Furnaces - Equipment Layout", Iron and
Steel Engineer, 37 (10), 73-78 (October 1960).
McMulkin, F. J., "Oxygen Steelmaking in Canada", AIME Open Hearth Pro-
ceedings, 38, 241-254 (1955).
Nelson, F. D., "Progres s of New Basic Oxygen Shops - 1. At Inland Steel Co. ",
Journal of Metals, ~ (7), 785-787 (July 1965).
"Sparrows Point, Another Basic Oxygen Shop for Bethlehem Steel'., Journal of
Metals, ~ (5), 556-557 (May 1966).
Holloway, W. P., "Progress of New Basic Oxygen Shops - II. At Wheeling Steel
Corp.", Journal of Metals, ~ (7), 788-792 (July 1965).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
( 110)
(111)
(112)
( 113)
( 114)
( 115)
(116)
(117)
( 118)
(119)
(120)
( 121)
( 122)
(123 )
( 124)
( 125).
( 126)
C-132
Jones, M. A., "Engineering Aspects of the Basic Oxygen Furnace Plant at Great
Lakes Steel", Iron and Steel Engineer, !!.. (3), 123-128 (March 1964).
Nickel, M. E., "Progress of New Basic Oxygen Shops - IV. At Wisconsin
Steel Works, International Harvestor Co. ", AIME Open Hearth Proceedings, 48,
131-135 (1965). -
"Electric Furnace Round-Up", 33/The Magazine of Metal Producing, i (6),
pp. 71-80 (June 1966).
"Everybody is Getting Into the 'High-Power' Act", The Iron Age, 202 (2), pp. 22-
23 (July 11,1968).
Bennett, K. W., "Steelmakers Turn to Electrics", The Iron Age, 201 (26), 56-57
(June 27, 1968).
Coulter, R. S., "Smoke, Dust, Fumes, Closely Controlled in Electric Furnaces",
The Iron Age, 173 (2), 107-110 (January 14, 1954).
Kane, M., and Sloan, R. V., "Fume Control-Electric Furnace Melting Furnaces",
American Foundryman, ~ (11), 33-35 (November 1950).
Brief, R. S., et al., "Properties and Control of Electric-Arc Steel Furnace
Fumes", Journal of the Air Pollution Control Association, .~ (4), 220-224
(February 1957).
Hohenberger, A., "Dust Removal From Electric Arc Furnaces", Stahl und
Eisen, ~ (15), 1001-1005 (1961).
Davies, E., et al., "The Control of Fume From Electric Arc Furnaces", Journ::ll
of the Iron and Steel Institute, 201 (2), 100-110 (February 1963).
Bintzer, W. W., "Design and Operation of a Fume and Dust Collection System
for Two 100-Ton Electric Furnaces", Iron and Steel Engineer, 41 (2), 115-123
(February 1964). -
Baum, A., "Removal of Dust From Electric Furnace Waste Gases", Stahl und
Eisen, 84 (23), 1497-1500 (1964).
Danielson, J. A., "Metallurgical Equipment", Air Pollution Engineering Manual,
Public Health Service Publication No. 999-AP-40. (1967), pp. 235-257.
Hoff, W. A., "Use of Hot Metal in Electric Furnaces", AIME Electric Furnace
Proc eeding s, .!.i, 293 - 295 (1956).
Schmudde, A. W., "Use of Hot Metal in Electric Furnaces", Journal of Metals,
~ (4), 501-503 (April 1966).
Finkl, C. W., "Dust and Fume Control in a Modern Melt Shop", AIME Electric
Furnace Proceedings, .!.i, 272-278 (1956).
Peterson, H. W., "Gas Cleaning for the Electric Furnace and Oxygen Process
Converter", AIME Electric Furnace Proceedings, .!.i, 262-271 (1956).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
(127)
( 128)
(129)
( 130)
(131 )
(132)
(133 )
( 134)
(135)
( 136)
(137)
(138)
( 139)
( 140)
C-133
Campbell, W. W., and Fullerton, R. W., "Development of an Electric-Furnace
Dust-Control System", Journal of the Air Pollution Control Association, 12 (12),
574- 590 (1962). -
Pettit, G. A., "Electric Furnace Dust Control System", Journal of the Air
Pollution Control Association, ~(12), 607-609, 621 (December 1963).
Harms, F., and Riemann, W., "Measurement of Fumes and Dust Volumes From
70-Ton Electric Arc Furnaces Operated Partially on Oxygen", Stahl und Eisen,
82 (20), 1345-1348 (1962).
Hipkin, A. S., "Cleaning of Fume From Arc Furnaces", Air and Water Pollution
in the Iron and Steel Industry, Special Report 61, The Iron and Steel Institute
(1958), pp. 108-114.
Nestaas, 1., and Romslo, R., "Measurement of Particulate Emissions From
Scrap Pre-Heaters", Report for Schjelderups Industriovner A/S, Oslo, Norway,
by The Engineering Research Foundation, Technical University of Norway (1968),
10 pp.
Kahnwald, H., and Etterich, 0., "Determination of the Volume, Composition,
and Temperature of the Waste Gas and the Dust During Meltdown and Oxidati~n
by Oxygen Lancing in a IS-Ton Electric Arc Furnace", Stahl und Eisen, 83 (17),
1067-1070 (1963).
Walker, W. S., and Harris, T. H., "Some Operational Details of a Large Elec-
tric Melting Shop", Journal of the Iron and Steel Institute, 198, Part 1, 5-12
(May 1961). -
"How Electric Arc Furnaces Pay Off at Roanoke Steel", Carbon and Graphite
News, .!.Q (1), 6-7, 1966 (Metal Progress, 90 (6), December 1966).
Varga, J. Jr., Report on Official Travel, Battelle Memorial Institute, Columbus
Laboratories, August 27, 1968.
Jenison, R. E., "Lukens Steel Co. 's Electric Melt Shop Complex", Journal of
Metals, .!.2. (6), 41-43 (June 1967).
"How Electric Arc Furnaces Pay Off at Bethlehem-Seattle", Carbon and Graphite
News, .!.Q (1), 2-3 (1966). (Metal Progress, 90(6), December 1966).
Ess, T. J., "Armco at Butler", Iron and Steel Engineer, 38 (8), A-2 - A-20
(August 1961).
Rankin, W. M., "Electric Furnace Steel Production, Houston Works, Armco
Steel Corp. ", Journal of Metals, 20 (4), 104-107 (May 1968).
"J & L Combating Air Pollution at Stainless and Strip Div. ", Iron and Steel
Engineer, 45 (3), 139-140 (May 1968).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
( 141)
(142)
(143 )
( 144)
( 145)
( 146)
( 147)
( 148)
(149)
( 150)
(151)
(152)
( 1 53 )
(154)
(155)
(156)
C-134
Venturini, J. L., "Historical Review of the Air Pollution Control Installation at
Bethlehem Steel Corporation's Los Angeles Plant", Preprint No. 68-134. Air
Pollution Control As sociation Annual Meeting, St. Paul, Minnesota (June 23 -27,
1968), 19 pp.
Hornak, J. N., "Vacuum Degassing - Why and How?", Iron and Steel Engineer,
42 (6), 73-79 (June 1965).
Hornak, J. N., and Orehoski, M. A., "Vacuum Casting of Steel'l, Preprint of
paper published in Journal of Metals, .!2. (7), 471-475 (July 1958).
Hornak, J.
Properties
(1958).
N., and Orehoski, M. A., "Effect of Vacuum Stream Degassing on
of Forging Steels", AIME Electric Furnace Proceedings, ~, 68-84
Hobson, J. D., "Results From a Pilot Plant for Vacuum Stream Degassing, and
Some Theoretical Consideration of the Process", Hydrogen in Steel, Special
Report 73, The Iron and Steel Institute (1962), pp. 30-61.
Schempp, E. G., Communication to J. Varga, Jr., Battelle Memorial Institute,
Columbus Laboratories, October 14, 1968.
Lehman, A. L., "Vacuum Stream Degassing", AIME Electric Furnace Pro-
ceedings, ~, 84-93 (1958).
Forster, G. B., "R-H Degassing'l, Journal of Metals, ~ (4), 628-633 (May
1966) .
Wilson, L. H., and Unick, T. F., "Ladle To Ladle Vacuum Stream Droplet
Degassing Facility and Operations At Sharon Steel Corporation", Blast Furnace
and Steel Plant, ~ (9), 823-832 (September 1965). .
Geogiadis, J. F., and Hendrick, R. B., "Method for Evaluation of Volatile Mold
Coatings", AIME Open Hearth Proceedings, 45, 326-337 (1962).
Bunting, R. L., "Lukens Cement Hot-Top Practice", AIME Open Hearth Pro-
ceedings, !!., 338-346 (1958).
Roloff, D. V., and Smith, K. V., "A Study of Hot-Topping Practice", AIME Open
Hearth Proceedings, 42, 18-29 (1959).
Bayers, W. E., and Boyle, C. D., "Exothermic Sideboard Hot Tops", AIME
Open Hearth Proceedings, 44, 430-443 (1961).
Wayne, T. J., "Automated Preparation of Low Volume Hot Tops", AIME Open
Hearth Proceedings, 47, 267-269 (1964).
Selky, J. L., and Bingham, R. C., "Disposable Hot Tops'l, AIME Open Hearth
Proceedings, 47,272-287 (1964).
Lamont, J. A., et al., "Investigation of Hot-Top Materials", Journal of Metals,
~ (6), 738-742 (1966).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
( 157)
( 158)
( 159)
(160)
(161)
( 162)
(163)
(164)
(165)
(166 )
( 167)
(168)
(169)
(170)
( 171)
C-135
Hoffman, A. 0., and Ziercher, J. L., Report on Official Travel, Battelle
Memorial Institute, Columbus Laboratories, November 14, 1968.
Hoffma~, A. 0., and Ziercher, J. L., Report on Official Travel, Batte).le
Memor-ial, Institute, Columbus Laboratories, December 19, 1968.
Hoffman, M. F., et al., "Argon Casting for Improving Steel Quality"', AIME
Electric Furnace P'roceedings, 43, 375-386 (1960).
\ -
Wilson, W., "Argon'Teernfng of Degassed Steel", Journal of Metals, Q (4), 350-
352 (1961).
Communication from K. L. Backhaus, Concast, Inc., to J. Varga, Jr., Battelle
Memorial Institute, C6lumbus Laboratories, January 31, 1969.
Communication from T. Sullivan, Mesta Machine Co., to J. Varga, Jr., Battelle
Memorial Institute, Columbus Laboratories, September 9, 1968.
Communication from V. Navis, Amsted Research Laboratories, to J. Varga, Jr.,
Battelle Memorial Institute, Columbus Laboratories, December 16, 1968.
Discussion to article: Glossbrenner, A. B., "Tirnken Steel and Tube Division's
Approach to Bloom and Billet Conditioning", AIME Metallurgical Society Confer-
ences, Vol. 13. Bar and Applied Products, ?:J.., Interscience Publishers, New
York, N. Y. (1961).
Trilli, L. J., "Hot Machine Scarfing of Semi-Finished Carbon Steels", AIME
Metallurgical Society Conferences, ~, Flat Rolled Products II: Semi-Finished
and Finished, 3-17 (1960).
Whittaker, R., and Long, R. L., "Factors Affecting the Yield of Free-Cuttil1;g
Steels at Park Gate", Optimization of Steel Product Yield, ISI Publication 107,
The Iron and Steel Institute. (1967), pp. 47-55.
Keefe, J. M., "Optimization of Yield in Wide Strip Rolling, Part 2: Ingot to
Pickled Coil", Optimization of Steel Product Yield, ISI Publication 107, The Iron
and Steel Institute. (1967), pp. 64-72.
McLean, C. J., "Control of Defects - Flat Rolled Products", Blast Furnace and
Steel Plant, 54 (3), 231-240 (March 1966).
Communication from A. L. Hodge, Linde Company, to J. Varga, Jr., Battelle
Memorial Institute, Columbus Laboratories, January 30, 1969.
Glossbrenner, A. B., "Timken Steel and Tube Division's Approach to Bloom
and Billet Conditioning", AIME Metallurgical Society Conferences, Vol. 13.
Bar and Applied Products, 21-25. Interscience Publishers, New York, N. Y.
(1961).
"Trends in Steel Pickling and Waste Acid Treatment", 33/The Magazine of Metals
Producing, i (3), 65-76 (March 1966).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
I
(172)
(173 )
( 174)
(175)
(176 )
(177)
(178)
(179 )
(180 )
(181 )
C-136
Thompson, H. J., "Conversion from Sulphuric to Hydrochloric Acid in a Hori-
zontal Pickle Line", Iron and Steel Engineer, 45 (2), 102-108 (February 1968).
Communication from P. R. Klauss, Swindell-Dressler Company to J. Varga, Jr.,
Battelle Memorial Institute, Columbus Laboratories, October 18, 1968.
Miltenberger, R. S., liThe Use of Hydrochloric Acid in Conventional Pickling
Facilities", Blast Furnace and Steel Plant, 53 (9), 833-836 (September 1965).
Lemke, E. E., et al., "Air Pollution Control Measures for Hot Dip Galvanizing
Kettles", Journal of the Air Pollution Association, .!.Q. (1), 70-77 (February 1960).
Penton's Foundry List, 1967-68, Penton Publishing Company (1967).
"Iron Roll Foundry Expansion Under Way at Bethlehem Plant", American Metal
Market, p. 5, September 8, 1964.
"Takes Delivery of 3 25-Ton Furnaces", Metal Working News, p. 13, October 11,
1965.
Sterling, M., "Foundry Air Pollution Problems ", paper presented at the National
Conference on Air Pollution, December 1966, ,Washington, D. C.
"Foundry Air Pollution Control Manual", American Foundrymen' s Society, Des
Plaines, Illinois (1956), p. 39.
Cowen, P. S., "Roundup on Air Pollution", Gray and Ductile News, Gray and
Ductile Iron Founders' Society, pp. 5-11 (December 1967).
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
-------
D-l
SUMMARY AND CONCLUSIONS
In this appendix, the cost of air pollution control equipment is estimated
for various processes of the integrated iron and steel industry. Factors
affecting the performance of control equipment and the effect of performance
level on cost are discussed.
The purpose of the cost estimating presented here is to. provide average
data on the emission control cost per production unit, as an initial step in
the formulation of a model for calculating the national cost of air pollution
control in the industry. A technique of estimating costs for this purpose is
described and used. The more detailed technique for estimating the cost of a
specific installation to suit the particular conditions and specifications is
not called for here. Without that detail, as outlined below, these estimates
cannot and should not be used to determine the cost of control for a specific
installation. An engineering analysis for that purpose, using classical esti-
mating techniques, is available from a number of engineering firms, and should
be used if the cost of control is desired for, e.g., company A's furnace B in
ci ty C.
The costs presented in the tabulations here are based on process parameters
representative of average modern practice. .Costs will vary from plant to plant
depending on such specifics as raw material properties, details of the process
as applied, product properties desired, unusual materials of construction or
unusual combinations of equipment occasioned by special corrosion or abrasion
conditions, details of integration of equipment into plant lay-out, etc. Typical
methods of control which are, or could be, applied to the average process are
cost estimated. Typical options for gas cooling are incorporated in each system;
this choice, together with the average process effluent level (both gas quantity
and particle concentration), determines the capacity and power ratings of
equipment.
The effect on cost of unusual arrangements for combined control systems
and for area ventilation is discussed. The effects on cost of unusual adapta-
tions of equipment to existing plants and facilities are discussed.
Also presented here is an indication of the theoretically determined
difference in cost for a more effective control system of the types typically
used. This is not the cost of altering an existing installation, with its
specific needs. These cost differences also are intended as input to the
model for national costs. The development of the model with cost data from
systems as used today, or as designed, will yield the ultimate tool for
determining national cost, and provide an average .comparison to the initial
input data presented here.
Certain nominal unit costs have been established as bases for calculating
operating costs. These costs will vary with monetary fluctuations over the
life-span of equipment. In adjusting field data as input to the model devel-
opment calculations, these costs would be normalized. Electrical energy is
standardized at $50/installed HP per year (based on a standardized 330 oper-
ating days per year x 24 hours/operating day = 7,920 operating hours/year).
This corresponds generally to 3/4~/KWH for large motors. Labor cost is set
at $5.00 per manhour including all welfare and fringe costs. Ratios of real,
local costs to these standardized values may be used as factors for adjusting
reported operating costs.
-------
D-2
PRIMARY CATEGORIZATION OF COSTS
Many costs arise during the life span of an industrial project, from
the earliest planning to the final demolition of the obsolete plant.
These costs are commonly assembled into three categories:
1.
Capital Costs: Cash outlays associated with
planning, engineering, purchasing, construction
and startup of the installation. Such costs occur
only once during the life of the installation.
2.
Operating Costs: Charges associated with the
operation, maintenance and financing of the plant
during its period of productivity. These costs
are repetitive in nature, constituting a continual
flow of cash away from the operating organization.
3.
Demolition and Salvage Costs: Cash transactions
arising while the facility is being dismantled and
sold off. Some of the cash flows are expenses and
some are income. The algebraic sum constitutes
either the Demolition Cost or the Salvage Value
depending upon the direction of the net cash flow.
These items occur only once during the life of the
plant, in which respect they are related to the
Capital Costs in (1) above.
THE GENERAL STRUCTURE OF CAPITAL COSTS
The capital cost of any plant or facility is the sum of many separate
cash payments made to suppliers of component parts, to workers of many kinds
for their labor, to consultants and contractors for their services, to
shippers for transportation of components to the construction site, etc.
This can be expressed as
CT = cl + c2 + c3 + ...... + cn
where CT is the total capital cost and cl' c2' c3'
cash payments as described above.
. . . .. c
n
are individual
In most industrial projects, the total number of these payments, n, is
very large. To simplify accounting and cost analysis, they are usually
grouped into a relatively small number of categories. Each category contains
the amounts of all payments related to some recognizable sub-division or
functional aspect of the total project. If these categories are called
Cl' C2' C3' ...... CM'
Cl = cl + c2 + c3
C2 = c4 + Cs + c6 + c7
C = C + C + C
M e m n
-------
D-3
The sum of
individual cash
the categories.
the categories equals the total cost, provided
payment appears once and only once, in one and
That is,
CT = Cl + C2 + C3 + ......+ CM
that each
only one of
In a typical situation, the total number of payments, n,
the range of thousands, but M, the number of cost categories,
10 to 20. Various categorizing schemes may be, and have been
categories may be functional in nature:
might be in
might be only
used. Some
Cl =
materials cost
C -
2 -
field labor cost
C3 =
C4 =
engineering labor cost
freight cost
etc., etc.
In such a scheme, Cl is the sum of all materials costs on the project
while C2 is the total cost of all labor required for field erection of all
materials.
Another system of categorization can be based on major component parts
of the installation. Each category would include all costs (materials,
labor, freight, engineering, etc.) associated with one part of the plant.
C =
1
cost.of furnaces
C =
2
cost of gas cleaning equipment
C =
3
cost of water supply system
C =
4
cost of buildings
etc., etc.
The choice of categorization scheme is usually a matter of tradition
or convenience, as long as each cost item appears once and only once in the
array.
As was already noted, these cost items occur only once in the life of
the project. Moreover, these costs are clearly defined cash transactions,
determined by market action. Unlike certain components of operating cost
(discussed elsewhere), they are generally unaffected by accounting practices,
tax procedures and other constraints arising from policy de~isions.
The planning, engineering and construction of an industrial plant
often encompasses a time span of two to four years. During these years,
CUrreDG~ inflation may be great enough to have a significant effect on
costs arising in the later phases of the overall program.
-------
D-4
CATEGORIZATION SCHEMES
A typical categorization scheme for industrial projects divides
capital costs into the following ten categories:
1.
Material: This includes the purchase cost of all materials,
machines, component parts, cement, structural steel, etc.
which are required in the field to make up the complete
operating installation. Net purchase costs are often F.O.B.
point of origin.
2.
Erection Labor and Supervision: This item includes wages
and salaries, payroll taxes, welfare benefits, etc. for all
persons employed at the construction site in the installation
of the material items listed above. Some estimating pro-
cedures involve the preparation of separate estimates for
the labor force and for a supervisory group of engineers.
This procedure may be needed when supervision is supplied by
an organization that is not responsible for employing the
general construction personnel.
3.
Freight: This covers the cost of transporting all of the
materials from their respective points of origin to the
construction site.
4.
Special Tools: This cost category includes rental and
transportation charges for special tools or equipment that
may be needed at the construction site. Excavating equipment
and large hoisting machinery are often rented for brief periods
of time during a construction project because the amount of
work to be done by them does not warrant their outright
purchase for one job.
5.
Taxes and Insurance: This covers the payment of necessary
sales taxes, permits and charges for insurance protection
as required during the course of the project.
6.
Engineering: In this category are collected all of the costs
associated with the central engineering and design aspects of
the project. This includes the services of engineers and
other personnel involved in the design, purchasing, and
general management of the project. The costs of these
services are usually calculated at a standard rate which
provides for salaries, fringe benefits, occupancy, and
departmental overhead. In addition, this category can
logically include general overhead and fee to the central
engineering organization.
7.
Client Engineering and Coordination: During the design and
construction of a new plant installation, the company or other
organization which will operate the finished plant usually
participates in the engineering and general management of
design and construction. This may involve the preparation of
specifications, review of design drawings, review of purchase
orders, participation in development of field schedulea, etc.
Charges in this category should include those directly
associated with the personnel involved together with an
appropriate share of overhead.
-------
D-5
8.
Startup:
operation
it into a
This covers costs associated with preliminary
and test of the new equipment in order to bring
cQndition of reasonable operating efficiency.
9.
Inventory:. This working capital item covers those moneys
which must be tied up in inventory of raw materials, goods in
process, maintenance supplies, etc.
10.
Land: This covers the fair value for the land area assigned
to the installation. In general, it includes not only the
land occupied by processing or manufacturing equipment but
also land utilized for the storage of materials and products
immediately preceding and following the processing unit. In
building a new installation, there is often a cost arising
from the preparation of the land site. This may involve
grading, filling, removal of old structures, etc., and must
not be omitted from the calculation of total investment.
The total investment is the sum of all of the foregoing items. This
total will not be the total payment to contractors, suppliers, and construct1on
labor because of the presence of items 7 through 10 above. Nevertheless, all
of these items (1 - 10) make up the total investment required to erect the
new industrial installation and bring it into normal working condition. It
is this total investment cost which enters into the cost computations else-
where in this project.
The detailed arrangement and tabulation of individual cost items in
the total as defined above can be handled in a variety of ways. The
particular method selected is largely a matter of individual preferences.
This may be based upon accounting practices well established in earlier
projects or upon cost classifications needed for tax or operating control
purposes.
In one method of tabulating costs, the costs are arranged according to
the piece of equipment which is concerned; the cost of a single item of
equipment would include material, erection labor and supervision, freight,
engineering, etc. This leads to an estimate composed of a group of cost
figures which are the total installed costs of individual pieces of equip-
ment or of operating subsections.
Another approach to the. problem is based upon functional 1ines~ Costs
are arranged in accordance with the ten categories given above. This
method does not display the total installed cost of a single piece of
equipment but does reveal the total cost of each function. In particular,
it discloses the total cost of field labor and the total cost of engineering
services. Control of these two .functions is often considered to be an
important matter by the managers of engineering contracts. In general, this
latter scheme will be used in the present investigation, with some reduction
in the number of categories, because of the generalized nature of the
estimates involved.
-------
D-6
THE GENERAL PROBLEM OF CAPITAL COST ESTIMATING
The estimator of capital costs wishes to predict the total cost, ~,
of a new installation to be designed and constructed at some later time.
His principal working technique is that of extrapolation into the future,
using design data on the new plant and past cost experience with similar
facilities. Various estimating procedures are available, differing in
the amount of work they entail, and in the accuracy of the resulting
estimate. In general, the more laborious methods are needed to achieve
the more accurate results. As a result, a choice must be made in any given
case between estimating precision and estimating cost.
The most economical estimating procedures generally involve a direct
estimate of CT, total capital cost of a new plant, derived from a historical
record of the cost of similar plants. In practice, however, such methods
give very rough, imprecise estimates because the design of the new plant
usually differs in major respects from its predecessors. Since many com-
ponent parts of plants undergo only small changes with time, it is usually
possible to obtain more precise results by making separate estimates of
component costs.
Component cost estimates may be taken from historical records, or from
recent market quotations. Both of these sources unavoidably contain poten-
tial errors which pass along into the total cost, CT' The probable error
in the total is a weighted average of the probable errors in all of the
components. The weighting factors are based on the relative costs. of the
individual components, and the average is calculated as a root-mean square.
As might be expected, this procedure assigns greatest importance to the
most expensive components, with proportionately less emphasis on the cheaper
items. In practical estimating, therefore, high-priced components must
receive a great deal of attention if the final estimate .is to be accurate.
Smaller, inexpensive parts of the plant may be treated in a more approximate
manner without seriously distrubing the total cost.
One consequence of this situation is a trend toward increasing the
number of components to be estimated separately. It is argued that this
will eliminate high-priced components and improve estimating accuracy
because no single error will have a la~ge weighting factor associated
with it. Experienced estimators know tnis to be true, to some extent, but
that a limit exists which cannot be passed. Even though the number of
components separately estimated becomes very large, each component price
estimate (however small it may be) still contains its error. The percent
error in the total is always a weighted average of the percentage errors
in all the component prices.
The cost of preparing an estimate increases as the number of estimated
components increases. This acts to restrain the tendency toward enlargement
of the number of estimated components, which is reinforced by the unavoidable
total error even with a large number of components, as described in the
previous paragraph.
-------
D-7
It is important to realize that some design work must be done before
the actual estimating can begin. This design work, which is expensive,
must generate enough data about each component in the estimate to permit
the setting of a component price. For a fully detailed estimate, the
design cost may be almost as great as that needed for actual construction.
For example, a detailed estimate of the cost of foundations in an
industrial plant might cover the following items:
Earthwork:
Machine Excavation
Trench Excavation
Hand Excavation
Trucking and Hauling
Backfill
Deep Foundations:
Bearing piles
Sheet piles
Walers
Formwork:
Buildings, mats and piers
Spread footers
Grade beams
Footings
Walls - below grade
Walls - above grade
Heavy equipment
Heavy mats
Elevated slabs
Shored slabs and columns
Earth slab paving
Reinforcing:
Bar
Mesh
Miscellaneous
Steel:
Anchor bolts
Embedded steel
Embedded railroad tracks
Miscellaneous steel
Checker plate and grating
Concrete:
Buildings, mats and piers
Spread footers
Grade beams
Continuous footings
Walls - above grade
Walls - below grade
Heavy equipment
Heavy mats
-------
D-8
Concrete: (continued)
Elevated slabs
Shored slabs and columns
Earth slab paving
Fine grading for paving
Batch plant
Waterproof walls and piers
Vapor barrier and waterstop
Joint materials
Color, sealer and grout
Finishes:
Steel Trowel
Screed finish
Brick
Floor hardener
Separate estimates are to be made of labor and materials for each
of the above items. It is evident that this array requires the making of
a very detailed design together with an equally detailed compilation of
historical data.
The design and estimating of a complete plan on this basis is very
expensive, and will be done only under extremely competitive conditions.
Moreover, this kind of analysis is not possible until a specific plant
site has been selected and its characteristics have been determined.
The objectives of the present study can better be served by using a
small number of components. In fact, the elaborate detail set forth above
would be inappropriate because this work is not directed toward any single
plant location.
THE GENERAL STRUCTURE OF OPERATING COSTS
Two classes of transactions enter into the operating cost of a
manufacturing plant. One is composed of direct cash expenditures for
labor, raw materials, fuel, electric power, maintenance supplies, etc.
The other is a group of costs whose magnitudes are determined in part
by managerial policy decisions of various kinds. Depreciation charges and
gene~al overhead burden are of this latter type.
Like capital costs, the multitude of individual operating cost items
is usually arranged into a small number of categories. Some of these have
been mentioned in the preceding paragraph. Since operating costs, unlike
capital cost, arise continually over the many operating years of the plant's
life span, they are usually collected and reported for comparatively short
periods of time. Most often, the year is the time interval chosen for
steady-state analysis, in order to eliminate season effects and daily
fluctuations caused by minor events in plant operation. The total
operating cost over this period can then be related to the total production
during the same time to arrive at a useful value for the unit production
cost.
-------
D-9
Thus, if 0* is the unit production cost, ° is the total operating
cost, and W is the total number of units of proauction, all during a
given year,
°T
0* =-
W
0T = 01 + 02 + 03 + 04 + 05 + ...... + ON
where 01 =
operating labor cost
° =
2
raw materials cost
maintenance labor and materials
°3 =
°4 =
electric power cost
° =
5
depreciation charges
° =
6
working capital charges
etc., etc.
During the operating life-span of a typical plant, there will be
substantial changes in technology, administrative techniques, social
practices, markets, state regulations, interest rates, currency values, etc.
These evolutionary changes may lead to substantial modifications in the
unit production cost.
THE GENERAL PROBLEMS OF ESTIMATING OPERATING COSTS
The estimation of the cost of the two types of transactions described
on the previous page brings two different problems to the estimator. The
first of these, the direct cash outlays for labor, maintenance, power, etc.,
can best be handled with the help of historical records of actual plant
experience. Cost records are generally kept by operating companies for
organizational cost centers based on considerations of product management,
administration structure, etc. When these cost centers cover the equipment
of interest to the estimator, plant records can supply directly the historical
basis needed for close estimating. When the cost centers do not correspond
to the operating area being examined by the estimator, the direct con-
struction of an historical base is not possible. In that case the required
cost components must be arrived at by theoretical calculations, personal
recollections, intuition, etc.
Unfortunately for the present study, steel companies do not generally
maintain cost centers around their pollution control activities. It is
the general practice to use cost centers which include both production units
and control facilities within a single perimeter. At this time there are
only scattered cost data available on operating labor and maintenance for
air pollution control installations in the steel industry.
-------
D-IO
The second type of component in operating cost is that whose
magnitude is established by policy decisions relating to depreciation,
overhead, capital charges, etc. In principle, depreciation should be
based in a simple, non-controversial way on the actual life of the
equipment and its ultimate salvage value. In practice, the prediction
of the life and salvage value of pollution control equipment is uncertain.
Operating conditions are usually severe, maintenance practices vary, and
the danger of obsolescence is great. Depreciation rates therefore are
strongly influenced by policy.
The allocation of corporate overhead involves even more difficult
policy questions. Charges for the use of capital in control equipment
require predictions of interest rates, profitability of alternative
investments, and future credit rating. The estimator is clearly working
in a very imprecise area when he considers these problems.
As a result, operating cost estimating is inherently uncertain and
must not be expected to lead to results of high precision. The selection
of optimum or preferred pollution control equipment or processes should
not be based upon small differences between the operating cost of
alternative designs.
COSTS OF CONTkOL SYSTEMS
A control system is considered to be made up of all the items of
equipment and their auxiliaries which are used solely for the general
abatement of atmospheric pollution in the neighborhood of the steel
works. Typically this will include a collecting hood or gas collecting
pipe at the furnace, ductwork, spray cooler, dust collector, fan and
motor, and stack. Included also will be structural steel, foundations,
control instruments, insulation, piping, water treatment, and electric
power supply facilities for the entire gas cleaning system. (Water treat-
ment includes all those items required for gas cleaning water uses and
sufficient for avoiding a water pollution problem.) Excluded are those
equipment items which, while they may contribute to the functioning of
pollution abatement equipment, would be used for process or economic
reasons even if there were no pollution abatement requirements.
The cost of land occupied by pollution abatement equipment has not
been included. It is recognized that such land has a real value but a
satisfactory method for estimating it has not been established. Costs
associated with preparation of the site, start-up operations and working
capital are also not included. Certain portions of a control system
occupy or utilize parts of steel plant buildings and, therefore, might be
charged with a share of general building costs. This item has not been
estimated here. In calculating operating costs no attempt has been made to
allocate a portion of general overhead to control systems.
Capital and operating costs in the following tabulations are based upon
collectors whose efficiency can be relied upon to produce an outlet dust
loading of 0.05 grains/SCF of gas. A later section of the report contains a
discussion of the relationship between cost and collection efficiency. In
general, higher efficiency is more costly.
-------
D-ll
EFFECT ON COST OF MULTIPLE FURNACE INSTALLATIONS
Pollution abatement equipment becomes increasingly cheaper as the
number of furnaces that can be connected to one common control system
increases. The largest saving in multiple furnace situations can be
realized from the alternate scheduling sequence of furnace operation.
Furnace shops using arc furnaces and BOF vessels which have high and
low gas emission periods seldom reach peak conditions at the same time.
This is generally due to material handling which limits the furnaces in
a common shop to being charged and teemed in succession. These furnaces
can be lanced or blown alternately which will permit designing the
multiple furnace control system for a reduced thermal capacity and reduced
air volume. A single collecting system sized on this basis for two furnaces
will handle approx. 1~7 times more fume than for a single furnace appli-
cation and a three furnace system could be designed to handle 2.5 times
more. This results in a large reduction in capital cost, although there
is a loss in operating flexibility. Where flexibility is important, a
design compromise is possible by using a single collector whose size will
permit peak operations on all furnaces at the same time. When more than
three units are to be served by a common control system, it is best to
assume that several furnaces will have to be at peak load together.
The following tabulation is suggested as a rough guide for estimating
the effect of combinations on capital cost; for all processes and types
of control equipment.
1.
Separate collectors
on each furnace
1 2 3
100% 200% 300%
100% 170% 250%
Number of Furnaces
2.
One collector to handle
peak loads on all
furnaces at once.
3.
One collector to handle
only one peak load at a
time.
100%
140%
200%
For example, the capital cost for control equipment serving two
furnaces together, with both able to run at peak loads, would be 170%
of the cost for a single furnace installation.
SPECIAL PROBLEMS ENCOUNTERED WHEN INSTALLING
NEW CONTROL EQUIPMENT IN EXISTING PLANTS
Existing plants can and have been revamped to accommodate modern dust
collecting equipment. A grass roots plant affords the flexibility of selecting
and installing cleaning equipment and duct work for maximum efficiency and
minimum capital cost. Providing and adapting fume abatement equipment for an
existing plant can in many cases be very expensive, especially if satisfactory
land is not available for locating this new equipment. In many cases the fume
collecting equipment must be located on top of the building roof which requires
-------
D-12
strenghtening all the supporting columns and trusses. A more serious situation
would require placing the fans and motors at roof level. For wet scrubbers in
such a case the weight due to the much higher scrubber horsepower requirements
can become so costly that it might be necessary to use an inferior type of
collector. The second-best collector may ultimately cost the customer more in
capital expenditure, maintenance, operating cost and efficiency.
The most costly aspects of designing a new fume collection system for
an operating facility often involves unusual and unorthodox arrangements of
fume pickup at the furnace. This is in part due to lack of space for supports
and interference with existing structures or obstructing personnel, vehicle
and crane approaches. To avoid these conditions may require an alternate type
of pickup at the furnaces, which, in turn, could dictate the type of apparatus
used for separation of the fume and particulate matter from the gases. In one
existing arc furnace shop, for example, it was most desirable to employ a direct
shell tap extraction from the arc furnaces, but in this shop very little free
area would then have been left for water cooled ducts, spark boxes, or cooling
chambers. Consideration was given to running ducts under the teeming building
but this alternative would have been extremely expensive and would have caused
considerable shut-down time. Moreover, the added furnace roof loading would
have required expensive revamping to support the extra weight of the water-
cooled elbow. The only practical solution was to install roof-truss hoods over
the furnaces. This system had the inherent disadvantage of moving 4 to 5 times
the air volumes required by the direct shell tap. The additional air volume
resulted in much more capital investment on fans and cleaning equipment. This
fume collection system was a compromise design forced by the limitations of
existing facilities. A newly designed plant could include direct shell taps on
the furnaces, resulting in much less expensive equipment and in a considerable
reduction in operating expense.
Limited space around an existing plant may require locating new control
equipment on the roof, as previously discussed, or possibly in a location so
remote as to make the duct runs much longer than would otherwise be needed. The
capital cost increases because of the extra ducting and the added air friction
losses. The increased static pressure requires a larger fan of greater horse-
power. The result will be increased investment and higher operating costs for
the life of the system.
Older control equipment at existing plants has often influenced the
selection of a new type of collector system. A plant already operating with
wet scrubbers will, if at all possible, try to adapt the new system to similar
cleaning equipment especially if there is enough reserve built into the plant
slurry system to handle the additional loading. Maintaining the same pattern
of equipment will not necessarily be the best selection for the process involved
and may increase the capital outlay as well as operating costs of the new
installation. A grass-roots plant is not generally as much influenced by such
continuity factors.
The amount of shut-down time required to install the fume collectors
will add to the cost of the installation by the amount of lost production. The
amount of down time that can be tolerated will influence the type, location and
duct routeingof the system. Any deviation from the most direct design such as
would be practical for a grass roots plant will therefore add cost to the install-
ation.
-------
,
D-13
Detailed design and estimate studies will usually be needed if cost
estimates are desired for new control equipment in existing plants. The
general cost data given in this report will not be reliable in such cases.
DATA SOURCES. PROCEDURE. PRECISION OF RESULTS
The primary data used in the preparation of cost estimates were
taken from a number of sources. These were:
1.
Estimate files of the Swindell-Dressler Company.
These files ranged in age from the immediately
current back to 1962. All costs were adjusted for
price escalation to bring them to present levels.
2.
Cost information supplied by certain steel companies
relating to their own plant facilities. Such data were
generally in the form of total costs for complete
installations. These were adjusted for inflation
by means of the same factors used on the estimate
data from Swindell-Dressler files.
3.
Cost information supplied by certain manufacturers
of pollution control equipment. This information
was presented in response to specific requests by
Swindell-Dressler and came in the form of budget
figures.
The Swindell-Dressler file data are detailed estimates prepared to meet
the requirements of particular competitive situations. Each estimate
assembled the costs for a specific location at a specific moment in time.
and this was done in much detail based upon a combination of firm price
quotations from suppliers and historical data about labor productivity at
the location in question. Each estimate, therefore, contains cost elements
which are influenced by local conditions not necessarily applicable to
other plants and geographical locations. A simple compilation of these
estimates would not have been adequate for the needs of the present investi-
gation.
The data from these many particular cases have been rearranged into
a more generalized form. The primary tactic used here has been that generally
followed by other government agencies and students of cost estimating.
This is to segregate the capital costs of the principal items of equipment
in the installation and to prepare smoothed, adjusted values for these
equipment items over a wide range of operating capacities. As is
well known. such smoothed values usually form a straight line graph on log-
log paper, with the slope of the line being related to certain characteristics
of the type of equipment involved. In this study smoothed material cost
data did give satisfactory linear graphs with slopes that were reasonably
related to the type of equipment.
-------
D-14
Following the practice of others in this field, the minor and bulk
materials were estimated on the basis of ratios applied to the costs of
the principal items. Labor costs were estimated for each principal item
and bulk category by the use of standard Swindell-Dressler factors
relating labor to material costs. The resulting labor figures apply to
the Pittsburgh area but may be adjusted for other locations through the
use of regional labor indices.
The capital costs included only facilities for loading the collected
dust or sludge into trucks for transportation elsewhere. No other disposal
costs or by-product values have been assigned. Central engineering costs,
overheads and fees were based upon a standard sliding scale generally used
by contract engineers.
It is believed that the general precision of the capital cost estimates
is such that most specific plant situations will fall within %15% of the
tabulation values. In more statistical terminology it might be suggested
that the standard deviation is about %10 - 12%. It is to be expected that
any specific plant location which presents unusual cost problems associated
with layout, structure, power supply, etc. might fall outside these limits.
In such cases a detailed plant design and estimate should be prepared if
accurate capital cost data are required. As previously noted, the accuracy
of operating cost values is influenced by many factors which may vary
considerably from one company to another. The selection of control equip-
ment should not be based upon small differences in operating cost estimates.
Operating cost estimates present tabulated costs for the following
items:
1.
Electric Power
2.
Maintenance
3.
Operating Labor
Direct Operating Cost - The sum of above three items.
4.
Depreciation
5.
Capital Charges
The items included in Direct Operating Cost are those cost elements
which are the direct cash outlays discussed on page C-8. They are, to some
extent, under the control of the plant operating management. The costs
assigned for Depreciation and Capital are, as noted on page C-lO based upon
policy decisions not generally under the control of management at the plant
operating level.
Electric energy is calculated at a standardized rate of 3/4~ per kilowatt
hour. The cost of make-up water is not included as such. Operating labor
cost was calculated at the nominal value of $5.00/man hour including all
welfare and fringe costs.
-------
D-15
Maintenance is taken at a nominal cost of 4% of the total investment
shown in the capital cost tables. This figure has been the subject of
considerable discussion and has been retained in this analysis because it
is believed to represent a reasonable value over the total life of the
equipment. Most steel plant maintenance cost records do not make a
separate accounting for each pollution control installation. It is,
therefore, difficult to arrive at an exact, numerical evaluation of total
maintenance during the life of a piece of control equipment. In actual
practice, maintenance expenditures are not uniform from year to year.
In many cases major maintenance outlays occur only after the passage of
several years of operation. Moreover, as might be expected, maintenance
costs usually increase during the service life of an item of control
equipment. There have been some cases reported where major costs were
experienced early in the life of a control installation. It is believed
that these cases should properly be attributed to inadequate engineering
rather than standard maintenance. These incidents were more common some
years ago when knowledge of pollution control engineering was not as .
extensive as it is today. There have also been cases reported in which
major modifications were made to control equipment after it had been in
service for some years. Some times these episodes were caused by changes
in the operating practice of the process segment which placed greater
burdens on the control equipment. This type of cost is not considered to
be a part of maintenance. The 4% figure is retained in this study be-
cause it is believed to represent a reasonable value for good maintenance
in well designed equipment when calculated over the entire life of the
installation.
In those cases where filter bag replacement represents a major
maintenance cost item, the system, less bags, is given the 4% maintenance
charge; and the cost in material and labor for bag replacement at a reason-
able average rate (18 months for Sinter plants, 2 years for steelmaking
shops) is added for the net maintenance cost listed.
Depreciation is calculated on a straight line method using total
investment with an expected life of ten years. Other studies of depreciation
have suggested longer service life times, but these are considered to be
greater than average plant experience will confirm. Advancing technology
and rising standards give importance to the factor of technical obsolescence.
Capital charges are taken at 10% per annum. It is believed that this
will be reasonable in the light of rising interest rates and local taxes.
Annual calculations are based on 330 operating days per year, 24 hours
per day. This gives a total of 7,920 operating hours per year.
ALTERNATE SYSTEMS
In the tabulations which follow, several alternate control systems are
included, for most processes. Aside from cost, other factors enter into the
selection of a system. The nature of the process to some extent dictates or
precludes the use of a particular type of control. For example, some collected
dusts can be reused directly in the process or used as burden in the plant
-------
D-16
following agglomeration; the wet or dry state of the collected dust may afford
a convenience to disposition of the dust according to the current practice of
a particular plant.
The gas may be reusable as a process material in the plant, where changes
in its temperature and humidity by the cooling system would have to be consid-
ered in the total plant energy economy. The local cost of treated water, space
requirements for retreatment facilities, and possible difficulties in using
water near the process vessel may affect the choice of wet or dry systems.
The particle size and concentration of the effluent determine whether
high efficiency gas cleaning equipment (high-energy wet scrubber, electrostatic
precipitator, or baghouse) is needed, based on particle size vs. efficiency
experience data for different types of collectors. This data is largely in
the form of proprietary design curves in the files of equipment manufacturers.
For the purposes of this study, the dividing line between low-and high-energy
wet scrubbers is 12 inches of water pressure drop across the collector, with
high-energy applications generally using several times this pressure drop.
The nature of the dust collector equipment to be used largely determines
the extent and method of cooling the process gas. The process effluent may
vary in temperature from IOO°F for material transfer point ventilation to
3000°F or higher for furnace exhaust. This gas may be quenched by air dilution
or water sprays to a lower temperature, or undergo a heat exchange to cool
without adding material to the effluent stream. If the gas is combustible,
it may initially be burned, with an excess of air to insure completness of
combustion, where a fire or explosion hazard would exist. A wet-type cleaner
may treat water-quenched or hot gases directly. The electrostatic precipi-
tator used on highly (electrically) resistive particles requires a degree of
cooling and humidification control to be effective and of economical construc-
tion. On the other hand, over-cooling or over-quenching can result in conden-
sation with resultant corrosion, collector surface fouling, and dust handling
problems in dry precipitators, and baghouses as well. Thus, in general
- excess air is added to the effluent stream where combustion occurs,
in a water-cooled or other heat-exchange vessel;
- water addition completes the cooling for a wet system;
- indirect cooling by heat exchange will provide the most economical
cooling to about 500°F in dry systems;
- added humidification by water sprays usually completes the treatment
of gases prior to electrostatic precipitation;
- air dilution for bag temperature control usually completes pre-
baghouse cooling.
Finally, to achieve economical fan power levels, the gas volume is kept
low by gas cooling especially with high-energy wet scrubbers. This ultimate
effluent gas, if sulfur oxides persist in significant degree to this point in
the system, must have sufficient lift in the form of thermal or mechanical
energy, or stack height to disperse in the atmosphere.
-------
D-17
SINTER PLANTS
The following tables contain capital and operating cost data for two
siz~~ of sinter plants. Sinter plant control systems are usually designed so
that one control unit handles gases coming from the windbox while a separate
control unit receives dust collected at several points in the material'
handling system.
The attached estimate gives separate figures for the windbox and
materials handling operation. Various combinations of types of collection
equipment are used on sinter plants and it is therefore necessary to offer
separate values for these two zones of collection. The total cost for a
given sinter plant will be the sum of the cost for the windbox and the cost
for the materials handling.
The tables do not include system components through recovery cyclones
or windbox fans. Booster fans are included. Modifications only are in-
cluded in the windbox stack item cost.
The capacity of the sintering machine for the tabulated gas volume
and dust collection cost is based on the average nominal capacity, making
normal sinter. Variations will occur with differences in the burden. For
example, self-fluxing sinter capacity may be as much as 35 percent higher
than a machine's normal 'capacity. (Symposium on Sinter Plants, Discussion,
Iron and Steel Engineer, June, 1959.) In other reports no such change is
noted. With self-fluxing sinter, more particulate matter passes through
the cleaner.
The addition of oily turnings and borings to the burden generates oil mist
in the windbox gas. One solution to this is the use of.a very high energy
wet scrubber system.
The temperature of the windbox gases is determined
process used on the machine. The gases are moist. Too
can result in corrosion -causing condensation and tacky
problems in dry collectors.
by the sintering
low a temperature
dust handling
-------
D-18
SINTER PLANT (WINDBOX) - WET SCRUBBER (LOW ENERGY)
Gas Volume - ACFM @ 325°F
Plant Capacity - TPD
1. Materia1*
2. Labor
3. Central Engineering
4. Client Engineering
TOTAL
105,000
1,000
CAPITAL COST
$193,000
100,000
72 ,000
18,000
$383,000
OPERATING COST ($/Yr.)
1.
Electric Power
$ 20,000
2.
Maintenance
3.
Operating Labor
Direct Operating Cost
4.
Depreciation
5.
Capital Charges
TOTAL
15,000
30,000
$ 65,000
38,000
38,000
$141,000
Note:
1)
Prices: 1969 base.
For items included in materials, see pages C-10 and C-17.
*
630,000
6,000
$880,000
440,000
245,000
61,000
$1,626,000
$110,000
66,000
80,000
$256,000
163,000
163,000
$582,000
-------
D-19
SINTER PLANT (WINDBOX) - ELECTROSTATIC PRECIPITATOR
Gas Volume - ACFM @ 3250F
Plant Capacity - TPD
1.
Materia1*
105,000
1,000
CAPITAL COST
$180,000
88,000
72 , 000
18,000
$358,000
OPERATING COST ($/Yr.)
$ 12,500
14,500
20,000
$ 47,000
Direct Operating Cost
Note: 1) Prices: 1969 base.
36,000
36,000
$119 , 000
* For items included in materials, see pages C-10 and C-17.
2.
Labor
3.
Central Engineering
4.
Client Engineering
TOTAL
1.
Electric Power
2.
Maintenance
3.
Operating Labor
4.
Depreciation
5.
Capital Charges
TOTAL
630,000
6,000
$800,000
385,000
225,000
56,000
$1,466,000
$ 77,000
59,000
30,000
$ 166,000
147,000
147,000
$ 460,000
-------
D-20
SINTER PLANT (WINDBOX) - FABRIC FILTER
Gas Volume - ACFM @ 3250F
105,000
1,000
Plant Capacity - TPD
CAPITAL COST
1. Materia1*
2. Labor
3. Central Engineering
4. Client Engineering
TOTAL
$154,000
72,000
58,000
14,000
$298,000
OPERATING COST ($/Yr.)
1.
$
8,500
Electric Power
2.
Maintenance
18,000
3.
Operating Labor
20,000
Direct Operating Cost
46,500
4.
Depreciation
30,000
5.
Capital Charges
30,000
$ 106,500
* For items included in materials, see pages C-IO and C-17.
Note :l)This system is rarely used.
2) Prices: 1969 base.
630,000
6,000
$800,000
340,000
222,000
55,000
$1,417,000
$ 45,000
87,000
30,000
162,000
142,000
142,000
$ 446,000
-------
D-21
SINTER PLANT (MATERIAL HANDLING) - WET SCRUBBER (LOW ENERGY)
Gas Volume - ACFM @ 1350F
Plant Capacity - Tons/Day
CAPITAL COST
1.
Materia1*
48,000
1,000
$194,000
146,000
81,000
20,000
$441,000
OPERATING COST ($/Yr.)
$ 36,000
Note:
1) Prices: 1969 base.
2.
Labor
17,600
15,000
$ 68,600
44,000
44,000
$156,600
* For items included in materials, see pages C-lO and C-l7.
3.
Central Engineering
4.
Client Engineering
TOTAL
1.
Electric Power
2.
Maintenance
3.
Operating Labor
Direct Operating Cost
4.
Depreciation
5.
Capital Charges
TOTAL
250,000
6,000
$420,000
480,000
184,000
46,000
$1,130,000
$
95,000
45,500
40,000
$
180,500
113 , 000
$
113 , 000
406,500
-------
D-22
SINTER PLANT (MATERIAL HANDLING)-ELECTROSTATIC PRECIPITATOR
Gas Volume - ACFM @ 1350F
Plant Capacity - Tons/Day
1,000
250,000
6,000
48,000
CAPITAL COST
1. Material'\-
2. Labor
3. Central Engineering
4. Client Engineering
TOTAL
$149,000
$420,000
91 , 000
195,000
60,000
133,000
15,000
$315,000
33,000
$781,000
OPERATING COST ($/Yr.)
1.
Electric Power
$ 7,000
$ 41,000
2. Maintenance 12,500
3. Operating Labor 15,000
Direct Operating Cost $34,500
4. Depreciation 31,500
5. Capital Charge 31,500
TOTAL $97,500
*For items included in materials, see pages C-IO and C-17.
Note: 1) Prices: 1969 base.
31,000
20,000
$92,000
78,000
78,000
$248,000
-------
D-23
SINTER PLANT (MATERIAL HANDLING) - FABRIC FILTER
Gas Volume - ACFM @ 1350F
48.000
1.000
Plant Capacity - Tons/Day
CAPITAL COST
L.
Materia1*
$120.000
2.
Labor
68,000
3.
Central Engineering
49.500
4.
Client Engineering
12.500
TOTAL
$250,000
OPERATING COST ($/Yr.)
1.
$
9,000
Electric Power
2. Maintenance 12,800
3. Operating Labor 15.000
Direct Operating Cost 36 , 800
4. Depreciation 25,000
5. Capital Charges 25.000
TOTAL $ 86,800
*For items included in material, see pages C-IO and C-17.
Note:
1)
Prices: 1969 base.
250.000
6.000
$350,000
166,000
116,000
29.000
$661,000
$ 38,000
36,500
20.000
94,500
66,000
66.000
$ 226,500
-------
D-24
PELLETIZING PLANTS
The following tables contain cost data for a pelletizing plant
of 1,500,000 tons per year. This is a commonly used plant
capacity with larger output being achieved through use of parallel
units. It is not likely that many pelletizing plants will be
built whose capacity is less than that shown here. It is believed
that the costs presented are reasonably typical of the several
types of moving grate equipment now in use.
Like the sinter plant, several control systems are used at different
points on the unit. The total cost is the sum of the cost at the
dryer exhaust and the materials handling dust points.
Many pelletizing plants hold to the shaft furnace design, using
multiples of the 60 ton/hr. furnace. A system of cyclones is
included in this section for the cleaning of the process gas
leaving the furnace. And also, air from the cooling unit and material
handling points at the discharge station is cleaned separately.
-------
D-25
PELLETIZING PLANT (MOVING GRATE - DRYER EXHAUST) - CYCLONE
Gas Volume - ACFM @ 2500F
320,000
1,500,000
Plant Capcity - Tons/Year
CAPITAL COST
1.
$ 180,000
95,000
Materia1*
2.
Labor
3.
69,000
Central Engineering
4.
Client Engineering
17,000
$ 361,000
TOTAL
OPERATING COST ($/Yr.)
1.
$ 20,000
14,000
30,000
$ 64,000
36,000
36,000
$ 136,000
Electric Power
2.
Mainten~lnce
3.
Operating Labor
Direct Operating Cost
4.
Depreciation
5.
Capital Charges
TOTAL
*
For items included in materials, see pages C-10 and C-24.
Note:
1)
Prices: 1969 base.
-------
D-26
PELLETIZING PLANT (MOVING GRATE ~ MATERIAL HANDLING)
- WET SCRUBBER (LOW ENERGY)
o
Gas Volume - ACFM @ 70 F
55,000
Plant Capacity - Tons/Year
1,500,000
CAPITAL COST
1.
Material*
$
80,000
2.
Labor
54,000
3.
Central Engineering
36,000
4.
Client Engineering
9,000
TOTAL
$ 179,000
OPERATING COST ($/Yr.)
1.
Electric Power
$ 22,000
7,000
10,000
$ 39,000
18,000
18,000
$ 75,000
2.
Maintenance
3.
Operating Labor
Direct Operating Cost
4.
Depreciation
5.
Capital Charges
TOTAL
*
For items included in materials, see pages C-IO and C-24.
Note:
1)
Prices: 1969 base.
-------
D-27
PELLETIZING PLANT (SHAFT FURNACE - PROCESS EXHAUST)-CYCLONES
Gas Volume - ACFM @ 4600F
125,000
60
Plant Capacity - Tons/Hr.
CAPITAL COST
1.
$135,000
Materia1*
2.
Labor
82,000
3.
Central Engineering
55,500
4.
Client Engineering
13,500
$286,000
TOTAL
OPERATING COST ($/YR.)
1.
Electric Power
$ 23,000
2.
Maintenance
11,300
3.
Operating Labor
15,000
$ 49,300
Direct Operating Cost
4.
Depreciation
28,600
5.
Capital Charges
28,600
.$106,500
TOTAL
*
For items included in materials, see pages C-IO and C-24.
Note:
1)
Prices: 1969 base.
-------
D-28
PELLETIZING PLANT (SHAFT FURNACE - MATERIAL HANDLING)
CYCLONES AND WET SCRUBBER
(LOW ENERGY)
Gas Volume - ACFM @ 700F 30,000 19,000
Plant Capacity - Tons/Hr. 60
CAPITAL COST
1. Materia1* $ 45,000 $35,000
2. Labor 32,000 22,000
3. Central Engineering 23,000 19,000
/
4. Client Engineering 6,000 5,000
$106,000 $81,000
" TOTAL $187,000
OPERATING COST ($/Yr.)
1. Electric Power $ 4,000 $ 7,500
2. Maintenance 4,000 3,500
3. Operating Labor 5,000 5,000
$ 13,000 $16,000
Direct Operating Cost $ 29,000
4. Depreciation 10,500 8,000
5. Capital Charges 10,500 8,000
$ 34,000 $32,000
$ 66,000
* For items included in material, see pages C-IO and C-24.
Note: 1)
15000ACFM capacity for once a week cleaning routine.
2)
Prices: 1969 base.
-------
D-29
COKE OVEN
Cost data are not presented for control of emissions from coke ovens.
The engineering problems involved are still being investigated, both in
the U.S.A. and abroad. Satisfactory control equipment, with proven
industrial performance, has not yet been developed.
-------
D-30
BIAST FURNACE
The attached table presents cost information for a typical modern large
blast furnace. It is anticipated that most future blast furnaces in the
United States will be of this size or greater. They will probably have
the type of wet scrubbing system shown here. Older units with combinations
of several types of control equipment are not likely to be copied in the
future.
Blast furnace gas cleaning costs should be divided between emission control
and normal plant operation. In the absence of an industry consensus, it is
suggested that an equal share be allocated to each of these accounts. The
portion of the top gas which is fine-cleaned for the blast furnace stoves
is shown for a number of plants on pages C-3l to C-37 of the technical
counterpart of this report, entitled, "Final Technological Report on a
Systems Analysis Study of the Integrated Iron and Steel Industry," May 15, 1969.
-------
D-31
BLAST FURNACE - WET SCRUBBER (TWO STAGE. HIGH ENERGY)
Wind Rate - SCFM
150,000
210,000
Gas Volume - SCFM
CAPITAL COST
1.
$1,427,000
Materia1*
2.
Labor
636,000
3.
Central Engineering
360,000
4.
Client Engineering
90,000
$2,513,000
TOTAL
OPERATING COST ($/Yr.)
1.
$ 20,000
100,000
40,000
$ 160,000
251,000
251,000
$ 662,000
C-IO and C- 30.
Electric Power
2.
Maintenance
3.
Operating Labor
Direct Operating Cost
4.
Depreciation
5.
Capital Charges
TOTAL
*
For items included in materials, see pages
Note:1)These are total costs of cleaning the furnace top gas
of particulate matter. Since this operation serves
the ends of both emission control and plant operational
requirements (material recovery and fuel conditioning
for re-use), a share of the cost should be apportioned
to each account. It is suggested, in the absence of an
industry consensus, that the shares be equal.
2)Prices: 1969 base.
-------
D-32
BASIC OXYGEN FURNACE
The following pages contain cost data on several sizes of basic oxygen furnaces.
They assume that a new plant is being designed and that the pollution control
equipment is included in the original design. The figures cover a single furnace
only. It is recognized that various combinations of multiple units are used
in actual practice. The influence of this is discussed on page C-ll.
Heat extracting hoods ar~ included. These are total combustion systems
for typical oxygen-blow rates, with excesa air used for a portion of the
cooling. Water additions for saturation in wet scrubber systems and for
humidification (considerably less than for saturation) in electrostatic pre-
cipitator systems completes the cooling typically. For baghouse systems, the
gas would be kept dry, using air dilution at the hood and before the baghouse
with heat exchange means between.
-------
D-33
BASIC OXYGEN FURNACE - WET SCRUBBER (HIGH ENERGY)
Gas Volume - ACFM @ 1800F 220,000 440,000 660,000
Furnace Size - Tons 100 200 300
CAPITAL COST
1. Materia1* $910,000 $1,460,000 $1,960,000
2. Labor 490,000 790,000 1,060,000
3. Central Engineering 250,000 390,000 470,000
4. Client Engineering 60,000 100,000 120,000
TOTAL $1,710,000 $2,740,000 $3,610,000
OPERATING COST ($!Yr.)
1. Electric Power $ 207,000 $ 432,000 $ 664,000
2. Maintenance 68,000 110,000 145,000
3. Operating Labor 40,000 60,000 80,000
Direct Operating Cost $ 315,000 602,000 889,000
4. Depreciation 171,000 274,000 361,000
5. Capital Charges 171,000 274,000 361,000
TOTAL $ 657,000 $1,150,000 $1,611,000
* For items included in material, see pages C-IO and C-32.
Note:
1)
One Furnace System. For effect on cost of combined cleaning
systems on multiple furnace shops, see page C-ll.
2)
These estimates cover full combustion systems in which all of
the gas leaving the converter is mixed with an excess quantity
of air. All of the carbon monoxide is therefore burned to carbon
dioxide. This is the common industry practice in this country.
Systems have been designed which collect this gas in a substantially
unburned state. Such non-combustion systems may offer certain
economies. The exact extent of these economies has not yet been
generally recognized in the industry.
3)
Prices: 1969 base.
-------
D-34
BASIC OXYGEN FURNACE - ELECTROSTATIC PRECIPITATOR
a
Gas Volume - ACFM @ 500 F
375,000
100
785,000
Furnace Size - Tons
200
1,200,000
300
CAPITAL COST
3.
Central Engineering
250,000
$1,600,000 $2,250,000
800,000 1,100,000
410,000 550,000
100,000 140,000
$2,910,000 $4,040,000
1.
Material'\'
$ 900,000
2.
Labor
450,000
TOTAL
60,000
$1,660,000
4.
Client Engineering
OPERATING COST ($/Yr.)
1. Electric Power $ 90,000
2. Maintenance 66,000
3. Operating Labor 20,000
Direct Operating Cost $ 176,000
4. Depreciation 166,000
5. Capital Charges 166,000
TOTAL $ 508,000
$ 210,000 $ 310,000
116,000 162,000
30,000 40,000
$ 356,000 $ 512,000
291,000 404,000
291,000 404,000
$ 938,000 $1,320,000
* For items included in material, see pages C-IO and C-32.
Note:
1) One Furnace System. For effect on cost of combined cleaning systems
on multiple furnace shops, see page C-ll.
2) Prices: 1969 base.
-------
D-35
BASIC OXYGEN FURNACE - FABRIC FILTER
Gas Volume - ACFM @ 2750F 288,000 600,000 892,000
Furnace Size - Tons 100 200 300
CAPITAL COST
1. Materia1* $ 660,000 $1,280,000 $1,840,000
2. Labor 360,000 690,000 990,000
3. Central Engineering 200,000 340,000 470,000
4. Client Engineering 50,000 90,000 120,000
TOTAL $1,270,000 $2,400,000 $3,420,000
OPERATING COST ($/Yr.)
1. Electric Power $ 43,000 $ 89,000 $ 130,000
2. Maintenance 59,000 112,000 160,000
3. Operating Labor 20,000 30,000 40,000
Direct Operating Cost $. 122,000 $ 231,000 $ 330,000
4. Depreciation 127,000 240,000 342,000
5. Capital Charges 127,000 240,000 342,000
TOTAL $ 376,000 $ 711,000 $ 1,014,000
* For items included in material, see pages C-IO and C-32.
Note:
1)
One Furnace System.
on multiple furnace
For effect on cost of combined cleaning systems
shops, see page C-ll.
2)
This system is used in Europe, but so far has not had an American
application.
3)
Prices: 1969 base.
-------
D-36
OPEN HEARTH FURNACE
The following tables contain cost data for open hearth furnaces. They are
based upon the addition of gas cleaning equipment to an existing furnace
shop. It is not likely that many new open hearths will be built in the
future. The figures shown are for a single furnace. The effect upon cost
of multiple furnace combinations are discussed on pageC-llof this report.
It is assumed that waste heat boilers and boiler fans are existing at the
furnaces, and needed stack modifications are included in the estimates,
along with booster fans. Waste heat boilers, while they contribute to
pollution control by cooling the gases (without adding additional material
to the gas stream which would increase size and cost of subsequent equipment),
also serve the plant energy economy, and have been in general use on open
hearth furnaces having no abatement equipment. Thus, they are not included
in the cost of air pollution control and no credit is assigned for steam
produced.
The estimates are based on averaged data for current oxygen-blown furnaces
of different sizes, charged typically with 50% hot metal, 50% cold scrap.
The typical gas cleaning equipment begins with boiler exhaust gas at 5000F
and 18% mOisturQ.(Steam augmentation is assumed during the dry gas period
after hot metal addition when fuel and atomizing steam rates are low, and
during low-rate initial oxygen lancing when gas temperature is low and the
checker water cooling sprays are not used.) Thus, temperature and humidity
control are minimized for dry gas cleaning systems. This gas volume per ton
furnace capacity diminishes on the average with increasing furnace capacity,
and is cleaned directly in an electrostatic precipitator system. The gas
is cooled by air dilution before a baghouse collector. The wet scrubber
saturates the gas.
-------
D-37
OPEN HEARTH FURNACE - WET SCRUBBER
(HIGH ENERGY)
Gas Volume - ACFM @ 1800F
30,000
90,000
240,000
Furnace Size - Tons
60
200
600
CAPITAL. COST
1. Materia1~'c' $160,000 $430,000 $1,000,000
2. Labor 85,000 230,000 540,000
3. Central Engineering 60,000 .140,000 280,000
4. Client Engineering 15,000 35,000 70,000
TOTAL $320,000 $835,000 $1,890,000
OPERATING COST ($/Yr.)
1. Electric Power $ 24,000 $ 77,000 $ 210,000
2. Maintenance 13 , 000 33,000 76,000
3. Operating Labor 40,000 60,000 80,000
Direct Operating Cost $ 77,000 $ 170,000 $ 366,000
4. Depreciation 32,000 83,500 189,000
5. Capital Charges 32,000 83,500 189,000
TOTAL $.141,000 $ 337,000 $ 744,000
* For items included in materials, see pages C-IO.and C-36.
Note: 1) One Furnace System. For effect on cost of combined gas cleaning
systems on multiple furnace shops, see page C-ll.
2)
A variance from this design and cost is noted for a tar-fired
furnace with no waste heat boiler. Gas volume at higher temp-
eratures before and after saturation, and other factors lead
to a 60% higher cost.
3)
For a discussion of unusual problems encountered when installing
new collecting equipment at existing furnace shops, and an in-
dication of cost variances, see page C-ll.
4)
Prices: 1969 base.
-------
D-38
OPEN HEARTH FURNACE - ELECTROSTATIC PRECIPITATOR
o
Gas Volume - ACFM @ 500 F
29,000
85,000
225,000
Furnace Size - Tons
60
200
600
CAPITAL COST
70,000
$320,000 $700,000
170,000 380,000
110,000 200,000
30,000 50,000
$630,000 $1,330,000
1. Materia1*
2. Labor
3. Central Engineering
4. Client Engineering
TOTAL
$130,000
52,000
13,000
$265,000
OPERATING COST ($/Yr.)
1.
Electric Power
$
5,000
$ 15,000 $ 45,000
25,000 54,000
30,000 40,000
$ 70,000 $ 139,000
63,000 133,000
63,000 133 , 000
$196,000 $ 405,000
2. Maintenance 11 , 000
3. Operating Labor 20,000
Direct Operating Cost $ 36,000
4. Depreciation 26,500
5. Capital Charges 26,500
TOTAL $ 89,000
* For items included in material, see pages C-IO and C-36.
Note:
1)
One Furnace System. For effect of combined gas cleaning systems
on a multiple furnace shop, see page C-ll.
2)
For a discussion of unusual problems encountered when installing
new collectors at existing furnace shops, and an indication of
cost variances, see page C-l1.
3)
Prices: 1969 base.
-------
D-39
OPEN HEARTH FURNACE - F ABRI C FILTER
Gas Volume - ACFM @ 2750F 45,000 135,000 350,000
Furnace Size - Tons 60 200 600
CAPITAL COST
1. Materia1* $75,000 $210,000 $530,000
2. Labor 40,000 120,000 300,000
3. Central Engineering 36,000 80,000 180,000
4. Client Engineering 9,000 20,000 45,000
TOTAL $160,000 $430,000 $1,055,000
1.
Electric Power
OPERATING COST ($/Yr.)
$ 8,000
2.
Maintenance
7,700
$ 22,000 $ 54,000
21,000 51,000
30,000 40,000
$ 73,000 $ 145,000
43,000 105,500
43,000 105,500
$159,000 $ 356,000
3.
Operating Labor
20,000
Direct Operating Cost
$ 35.700
4.
Depreciation
16,000
TOTAL
16,000
$ 67,700
5.
Capital Charges
* For items included in material, see pages C-IO and C-36.
Note:
1)
One Furnace System. For effect on cost of combined gas cleaning
systems on a multiple furnace shop, see page C-ll.
2)
This system is not currently in general use, but it has been
successfully applied in the U.S.
3)
For a discussion of unusual problems encountered in installing new
collecting equipment at existing furnace shops, and an indication
of cost variances, see page C-ll.
4)
Prices: 1969 base.
-------
1----
I
D-40
ELECTRIC ARC FURNACES
The following pages contain cost data relating to electric arc furnaces de-
signed for production of carbon steel. The figures are for completely new
installations. The special problems encountered when installing new control
equipment in existing plants were discussed on page C-l1. Each cost value
applies to a system of two furnaces with a common gas cleaner capable of
handling only one furnace at peak loads at any given time. For effect on
cost of a different system of multiple furnace control see page C-ll.
The volumes listed are based on typical oxygen blowing rates used in furnaces
making carbon steel from cold scrap. Oxygen and exhaust rates may be con-
siderably higher when making stainless heats. An excess of air would typ-
ically be added to the furnace gases for complete combustion of carbon
monoxide and hydrocarbons (the latter, during the melt-down of oily scrap),
and for cooling. These mixed gases would then be water quenched in wet
scrubbing, and also in pre-conditioning of the particles before electrostatic
precipitation, though less water would be used in the latter system to avoid
condensation and to optimize the temperature and humidity conditions for
effective precipitation. Although, for baghouse collection, these gases also
could be water quenched to some extent, effecting an economy in collector
size,it is more typical to cool by use of a radiating exchanger, and to.
finish the cooling with controlled air dilution just before the baghouse.
.~
-------
D-41
ELECTRIC ARC FURNACE - WET SCRUBBER (HIGH ENERGY)
o
Gas Volume - ACFM @ 180 F
36.000
25
137.000
150
210.000
250
Furnace Size - Tons (each)
CAPITAL COST
1. Materia1* $173,000 $511,000 $723,000
2. Labor 93,000 277,000 388,000
3. Central Engineering 67,000 162,000 215,000
4. Client Engineering 17,000 40.000 54.000
TOTAL $350,000 $990,000 $1,380,000
OPERATING COST ($/Yr.)
1. Electric Power $ 40,000 $174,000 $265,000
2. Maintenance 14,000 40,000 55,000
3. Operating Labor 40,000 60,000 80.000
Direct Operating Cost $ 94,000 $274,000 $400,000
4. Depreciation 35,000 99,000 138,000
5. Capital Charges 35,000 99,000 138 , 000
TOTAL $164,000 $472,000 $676,000
* For items included in materials, see pages C-IO and C-40.
Note:
1)
Two Furnace System, alternating peak loads. For effect on cost of dif-
ferent combinations of furnaces per cleaning system see page C-ll.
2)
See also the section on Special Problems Encountered When Installing
New Control Equipment in Existing Plants, page C-ll.
3) A variation from these costs is noted in a case of a single furnace
cleaning system where, after correction for the savings in a 2-
furnace system, the cost would be 40% higher than indicated here.
Remote placement of the scrubber is one factor in this variation.
4)
Prices:
1969 base.
-------
D-42
ELECTRIC ARC FURNACE - ELECTROSTATIC PRECIPITATOR
Gas Volume - ACFM @ 5000F
48,000
25
185,000
150
280,0()()
Furnace Size - Tons (each)
250
CAPITAL COST
1.
Materia1~'(
$159,000
2.
Labor
85,000
$465,000 $652,000
251,000 352,000
151,000 197,000
38,000 49,000
$905,000 $1,250,000
3.
Central Engineering
61,000
TOTAL
15,000
$320,000
4.
Client Engineering
OPERATING COST ($/Yr.)
1. Electric Power $ 8,000 $ 30,000 $ 60,000
2. Maintenance 13 , 000 36,000 50,000
3. Operating Labor 20,000 30,000 40,000
Direct Operating Cost $ 41,000 $ 96,000 $ 150,000
4. Depreciation 32,000 90,500 125,000
5. Capital Charges 32,000 90,500 125,000
TOTAL $105,000 $277 ,000 $ 400,000
~'( For items included in materials, see pages C-10 and C-40.
Note:
1)
Two Furnace System, alternating peak loads. For effect on cost of
different combinations of furnaces per cleaning system see page C-ll.
2)
See also section on Special Problems Encountered When Installing New
Control Equipment in Existing Plants, page C-ll.
3)
Prices: 1969 base.
-------
D-43
ELECTRIC ARC FURNACE - FABRIC FILTER
Gas Volume - ACFM @ 2750F
60,000
25
230,000
150
350,000
250
Furnace Size - Tons (each)
CAPITAL COST
1.
$120,000
$441,000 $654,000
209,000 321,000
140,000 196,000
35,000 49,000
$825,000 $1,220,000
$ 40,000 $ 52,000
39,000 57,000
30,000 40,000
$109,000 $ 149,000
82,500 122,000
82,500 122,000
$274,000 $ 393,000
* For items included in materials, see pages C-IO and C-40.
Material*
2.
Labor
60,000
3.
44,000
Central Engineering
4.
Client Engineering
11 , 000
$235,000
TOTAL
1.
Electric Power
OPERATING COST ($/Yr.)
$ 8,000
2.
11 , 000
Maintenance
3.
Operating Labor
20,000
$ 39,000
Direct Operating Cost
4.
23,500
Depreciation
5.
23,500
$ 86,000
Capital Charges
TOTAL
Note:
1)
Two Furnace System, alternating peak loads. For effect on cost of
different furnace c9mbinations per cleaning system see page C-ll.
2)
See also section on Special Problems Encountered When Installing
New Control Equipment in Existing Plants, page C-ll.
3)
Prices: 1969 base.
-------
D-44
ELECTRIC ARC FURNACE
Combination Direct Evacuation Control and Furnace Canopy-
Type Area Ventilation System - Fabric Filter
Gas Volume - ACFM @ 1400F
Shop Size - 2 Furnaces @ Tons (each)
125.000
20
750.000
120
CAPITAL COST
1.
Materia1*
$240,000
$1,200,000
2.
Labor
102,000
480,000
3.
Central Engineering
96,000
353,000
4.
Client Engineering
TOTAL
24,000
$462,000
88.000
$2,121,000
1.
Electric Power
OPERATING COST ($/Yr.)
$ 18,500
$ 100,000
2.
Maintenance
21,000
98,000
3.
Operating Labor
Direct Operating Cost
30,000
$ 69,500
40,000
$ 238,000
4.
Depreciation
46,000
212,000
5.
Capital Charges
46,000
212.000
$ 662,000
TOTAL
$161,500
* For items included in material see pages C-IO and C-40.
Note:
1)
Two Furnace System, alternating peak loads. For effect on cost
of different furnace combinations per cleaning system see page C-ll.
2)
See also the section on Special Problems Encountered When Installing
New Control Equipment in Existing Plants, page C-ll.
3)
Prices: 1969 base.
-------
D-45
S~ING
The following table presents cost data on scarfing units. These
units are of two different sizes. The smaller size is usually
employed when the billets to be handled are n~ver larger than about
50 inches. Larger billets will require the larger gas cleaning
equipment. The material cost excludes the cost of the Smoke
Tunnel. In wet cleaning systems on a scarfer, the water circuit
is normally coupled to an existing slab mill water treatment system,
so that slurry treatment is excluded in this case.
-------
D-46
SCARFING - WET SCRUBBER (HIGH ENERGY)
Gas Volume - ACFM @ 1000F
50,000
100,000
CAPITAL COST
TOTAL
$114,000 $176,000
66,000 96,000
48,000 68,000
12,000 17,000
$240,000 $357,000
1.
Material'\-
2.
Labor
3.
Central Engineering
4.
Client Engineering
OPERATING COST ($/Yr.)
1. Electric Power $ 38,000 $ 75,000
2. Maintenance 10,000 14,000
3. Operating Labor 5,000 7,000
Direct Operating Cost $ 53,000 $ 96,000
4. Depreciation 24,000 36,000
5. Capital Charges 24,000 36,000
TOTAL $101,000 $168,000
,\-
For items included in materials, see pages C-IO and C-45.
Note:
1)
Prices: 1969 base.
-------
D-47
SCARFING - ELECTROSTATIC PRECIPITATOR
Gas Volume - ACFM @ 100°F
CAPITAL COST
1.
Materia1*
2.
Labor
3.
Central Engineering
4.
Client Engineering
TOTAL
OPERATING COST ( $ /Yr . )
1.
Electric Power
2.
Maintenance
3.
Operating Labor
Direct Operating Cost
4.
Depreciation
5.
Capital Charges
TOTAL
50,000
$135,000
85,000
57,000
14,000
$291,000
$ 8,000
12,000
5,000
$ 25,000
29,000
29,000
$ 83,000
*
For items included in materials, see pages C-IO and C-45.
Note:
1)
Prices: 1969 base.
100,000
$204,000
112,000
76,000
19,000
$411,000
$ 18,000
16,000
7,000
$ 41,000
41,000
41,000
$123,000
-------
D-48
HCL PICKLING LINE - WET WASHER
The following table presents cost data on a spray washing system for an
HCL Pickling Line acid fume removal system. Most modern lines are now
sized for 80 inch strip. Fiberglass material is used for all duct and
stack work. Fume is scrubbed by successive spray and eliminator units.
For optional acid brick lined tunnel (6 ft. sq.) to outside fume collectors,
add $184 per foot of length to capital cost total, and $44 per foot of
length to annual operating cost total.
-------
D-49
HCL PICKLING LINE - WET WASHER
Gas Volume - ACFM @ 100°F
130,000
Line Capacity
80 inch at 1,000 FPM
CAPITAL COST
1. Materia1*
2. Labor
3. Central Engineering
4. Client Engineering
TOTAL
$ 81,000
30,000
23,000
6,000
"$140,000
OPERATING COST ($/Yr.)
1.
Electric Power
$ 10,000
2.
Maintenance
6,000
3.
Operating Labor
Direct Operating Cost
5,000
$ 21,000
4.
Depreciation
14,000
5.
Capital Charges
TOTAL
14,000
$ 49,000
*
For items included in materials, see pages C-IO and C-48.
Note:
1)
Prices:
1969 base.
-------
D-50
COLD ROLLING MILL - MIST ELIMINATOR
The following table presents cost data for an eliminator system to remove
the palm oil and water mist emission at roll stands of a typical, large,
five stand tandem cold rolling mill. The suction of the system picks up
mist from closure plate enclosed areas at each stand, carries it through
a tunnel to two mist eliminators and fans. The ventilation air thus
cleaned is disqharged up a stack. The treatment of collected oil for re-
use or disposal is not included.
-------
D-51
COLD ROLLING MILL - OIL MIST ELIMINATION
Gas VA lume - ACFM @ 1l0oF
200,000
Mill Size
80 inch, 5 stand tandem
CAPITAL COST
1. Material*
$ 85,000
2.
Labor
62,000
3.
Central Engineering
29,000
4.
Client Engineering
TOTAL
7,000
, $183,000
OPERATING COST ($/Yr.)
1.
Electric Power
$ 18,000
2.
Maintenance
7,000
3.
Operating Labor
7,000
Direct Operating Cost
$ 32,000
4.
Depreciation
18,000
5.
Capital Charges
TOTAL
18,000
$ 68,000
*
For items included in materials, see pages C-IO and C-50.
Note:
1)
Prices: 1969 base.
-------
D-52
POWER PLANT BOILERS
The following pages contain cost data on several sizes of in-plant boiler
houses. They assume that smoke and fly ash control equipment is being in-
stalled on an existing coal-fired boiler. The figures cover a single boiler
only. Various combinations of multiple boiler-collector units are used in
actual practice, with savings in larger sizes dissipated in additona1 duct,
dampers and complicated setup. The stack is considered to be already existing.
Multiple cyclones, when used for primary collecting, are included as they
yield no process advantage to the boiler. Booster fans are included.
Mechanically fed coal-fired boilers may achieve acceptable fly ash control
with multi-cyclones alone. However, large modern boiler houses. in integrated
steel plants would usually use pulverized coal firing for efficiency and quick
regulation of firing rate as well as ease of combined or auxiliary firing with
blast furnace or coke oven gas. Pulverized coal's higher percentage of fly ash with
a finer size grading requires the use of high efficiency control equipment,
of which the electrostatic precipitator is almost solely used (often in con-
junction with a mechanical primary co1lector~ as it is more economical than
wet scrubbing. The exhaust gas usually contains a significant amount of sulfur
dioxide, which promotes effective cleaning with a smaller precipitator than
would be required without it. The hot, buoyant gases leaving the precipitator
disperse more readily than if cooled by scrubbing or for baghouse cleaning.
Sulfur dioxide emission suppression, using limestone injection with baghouse
collection or absorbtive solution scrubbing, currently undergoing tests for
public utility application, may eventually displace electrostatic precipitation
of fly ash. But the trend in steel plant boilers is toward relatively po11ution-
free fuels, particularly gas and oil. Combustion devices to prevent carbon
monoxide emissions are considered 100% process beneficial, and not funded as
pollution control equipment. The formation mechanism and control techniques
for nitrogen oxides emissions are currently under study; a preventive method will
likely be sought for their limitation. The development of acceptable soot
build-up removal means remains a problem.
-------
D-53
POWER PLANT BOILER
Mechanically Fed, Coal Fired Boiler-Multicyclone Collector
o
Volume, ACFM @ 600 F
Boiler Size, pounds steam/hr.
CAPITAL COST
1.
Material~'(
2.
Labor
3.
Central Engineering
4.
Client Engineering
TOTAL
32,000
50,000
$20,000
10,000
10,000
2,500
$42,500
OPERATING COST ($/Yr.)
$ 2,300
1.
Electric Power
2.
Maintenance
3.
Operating Labor
Direct Operating Cost
4.
Depreciation
5.
Capital Charges
TOTAL
*
1,700
7,000
$11,000
4,300
4,300
$19,600
For items included in materials, see pages C-IO and C-52.
Note:
1)
2)
Prices: 1969 base.
One Boiler System.
96,000
150,000
$ 60,000
30,000
24,000
6,000
$120,000
$
7,000
5,000
15,000
$ 27,000
12,000
12,000
$ 51,000
-------
D-54
POWER PLANT BOILER
Pulverized Coal Fired Boiler - Electrostatic Precipitator
Volume, ACFM @ 3000F
100,000
200,000
CAPITAL COST
/
1.
Material'"
$260,000
$440,000
2.
Labor
140,000
230,000
3.
Central Engineering
100,000
170,000
4.
Client Engineering
TOTAL
25,000
$525,000
45,000
$885,000
1.
Electric Power
OPERATING COST ($!Yr.)
$ 28,000
$ 55,000
2. Maintenance 21,000
3. Operating Labor 30,000
Direct Operating Cost $ 79,000
4. Depreciation 52,500
5. Capital Charges. 52,500
TOTAL $184,000
36,000
40,000
$131,000
88,500
88,500
$308,000
*
For items included in materials, see pages
C-IO and C-52.
Note:
1)
2)
One Boiler, Two Precipitator System.
Prices: 1969 base.
-------
D-55
SAMPLE CALCULATION - OPERATING COST ($/Yr.)
The sample illustrates the calculations performed in arriving at the
operating cost for a fabric filter installation on a 150 ton Electric Arc
Furnace. Electric power costs (@ 3/4~ per kwh) is obtained by calculating
the total horsepower of all motors (plus power to lights and instruments)
and multiplying by a cost per horsepower factor. Power to a fan motor is
calculated by applying an efficiency to the power required for reversible
adiabatic compression. This latter quantity is called "Air H.P." Power
to a water pump motor is similarly calculated, the reversible pumping power
requirement being called "Water H.P."
1.
Electric Power (*)
Basis: $50/HP/Yr 800 HP Motor
$50/HP/Yr x 800 HP =
$ 40,000/Yr
2.
Maintenance
Basis:
4% of Capital Cost
Capital Cost $825,000
0.04 x $825,000 =
$ 33,000/Yr
6,000/Yr**
3.
Depreciation
Basis:
10% of Capital Cost
Capital Cost $825,000
0.10 x $825,000 =
$ 82,500/Yr
4.
Capital Charges
Basis: 10% of Capital Cost
Capital Cost $825,000
0.10 x $825,000 =
$ 82,500/Yr
5.
Operating Labor
Basis: 3/4 Man/Shift or 18 Manhours/Day
$5.00/Manhour
18 MH/Day x $5.00/MH x
330 Opr.Day/Yr =
$ 30,000/Yr
TOTAL
$274,000/Yr
*
See following page for notes.
** The difference from the 4% standard maintenance
cost with bag replacement cost figured as
described on page C-15.
-------
D-56
*1.
Electric Power Cost @ $0.0075/KWH
Operating Days = 330 Days/Yr
1 HP = 0.746 KW
$0.0075/KWH x 330 Days/Yr x 24 Hr/Day x 0.746 KW/HP
x .89 Mo~or Eff. $50/HP per Yr
*2.
Air HP = 0.0001575 PQ
P = Static Pressure, in. water
Q = Volume, CFM
Motor HP = Air HP
Eff.
Eff. = Efficiency - Range 60 to 70%
GPM x H
Water HP = 3,960
*3.
GPM = Gallons per Minute
H = Head, in Ft.
Motor HP = Water HP
Eff.
Eff. = Efficiency - Range 75 to 85%
METHOD OF DETERMINING EXHAUST GAS VOLUMES IN SIZING
COLLECTING SYSTEMS FOR PRICING
The following sample illustrates the method of calculating the capacity"
of the collector in each estimated system. In general, the exhaust gases are
cooled in transit through the system, so that successive items of equipment
in the system will have different volumetric capacities due to gas volume
changes with temperature and with the material additions (dilution air or water
vapor) added to effect cooling.
The starting point is to determine a typical process exhaust gas composition
and volume rate per unit of process throughput (as SCFM/ingot ton). In some
cases this is determined solely by the oxygen lancing rate which generates the
maximum exhaust volume during a steelmaking heat. In the open hearth case,
since fuel and air are customarily added to the furnace during lancing, and
waste heat boilers are generally used for cooling the exhaust gases, typical
volumes of gas At the boiler outlet condition we~e selected as a starting volume
for the gas cleaning system.
-------
D-57
In the blast furnace,scarfing, sinter plant windbox, pelletizing (dryer or
process), and power plant cases typical modern practice was used as a basis
for determining the process exhaust volume. In materials handling and mist
pick-up cases, where in-drawn ventilation air entrains particles, mist and
vapors to be controlled, typical modern systems were studied to determine
ventilation rates for adequate emission containment and to ensure the inclusion
of sufficient pick-up points to contain a plant's effluent according to the extent
that current technology can meet current standards.
Sample of volume determinati~n method:
Peak oxygen rate to process = 1500 SCFM at 32°F
1.)
2. )
Combustion with air of carbon monoxide
Carbon monoxide (maximum) = 3000SCFM
2CO + 02 + ?1. N2 = 2C02 + 79 N2
21 21'
produced.
+ excess air
Combustion products
CO 3000 SCFM
2
N2
+ Excess air
1 x 79 x 3000 SCFM = 5640 SCFM
2 21 '
Excess air = 500% in a typic~l case
= 5 x 100 x 5640 SCFM = 35,700 SCFM
79
Total 44,300 SCFM
44,300 SCFM x 1.7 (=Factor for two furnaces with alternating peak loads.:
"" 75,500 SCFM
3.)
Cooling the gases
The combustion occurs in a water-cooled, double-wall duet where
cooling occurs by radiation and convection of heat to the ~alls. The
gases leaving this section will typically be at about 1200 F. The
size of such indirect heat exchanger will be determined by combining
heat transfer and heat balance equations in an iter~tive calculation,
based on certain reasonable assumptions of water ~emperatures, gas
velocity, and water circuit capacity. Optimizing the total cooling
and gas cleaning system is an extensive design task, so that typical
equipment for each system has been selected for this study's estimates.
The cooling by air or water additions to the gases at l2000F involve:
a heat balance for calculating resultant volume.
, ' .
;Mml\n + MwHw '"" ~~~
M = pound moles of each component.
H ;::: enthalpy of each component at conditions.
m = each component of uncooled gas.
n = each component of cooled gas. ,
w ;::: water at spray water temperature.
-------
D-58
For a final temperature of 500°F, suitable for an electrostatic precip-
itator, about 20% moisture is required by this analysis.
75,500 SCFM f .8 = 94,500 SCFM
94,500 SCFM x (560 + 460)OR -
492°R -
185,000 ACFM @ 500°F
CAPITAL COST BREAKDOWN
The following tables illustrate the relative importance of various
components in total material costs.
This is a very rough breakdown, and variations occur due to capacity
and type of system. However, the relative orders of magnitude are well
maintained. Certain conclusions can be drawn from this tabulation con-
cerning the sensitivity of the total to local conditions. Foundations and
structure may change considerably without having a marked effect on the total.
Very often, a local requirement which tends to increase structure will simul-
taneously reduce foundations. The figures used for these two components are
based upon simple structures supporting the collector near grade, and a soil
bearing value of 4,000 1bs. per square foot.
The stack and fan components are rather closely related to gas volume
and collector type. They are therefore relatively well defined. Electrical,
while an important component, is predicted with comparative certainty from
horsepower.
The key cost element is the collector itself, and it is to this item
that the estimator gives the greatest attention. Generally this will involve
obtaining a price quotation from a reliable manufacturer, although the published
literature also contains useful information.
The second category, labor et a1, is estimated on the basis of
anticipated labor costs for each of the components in Total Material.
Typical rules for this calculation are:
(a)
(b)
Collector:
Labor is about 35% of Material
Fan, motor and starter:
Material
Labor is about 15% of
(c)
(d)
Stack:
Labor is about 100% of Material
Ductwork:
Labor is about 100% of Material
(e)
(f)
Steel:
Labor is about 30% of Material
Foundations:
Labor is about 130% of Material
(g)
Electrical:
Labor is about 150% of Material
-------
D-59
MATERIAL BREAKDOWN
"
Sinter Plant - Windbox Gas Cleaning
Wet Electrostatic Fabric
Scrubber Precipitator Filter
1. Foundations 4 3 4
2. Ductwork and Stack 2 5 4
Modifications
3. Collector 30 68 71
4. Fan and Motor 7 7 10
5. Structural 2 3 3
6. Electrical 7 9 5
7. Water Treatment & Piping 46 1 1
8. Controls 2 4 2
Total 100% 100% 100%
Sinter Plant - Material Handling -
Dust Collection
1. Foundations 4 3 4
2. Ductwork and Stack 12 18 21
3. Co llec tor 28 47 46
4. Fan and Motor 7 5 7
5. Structural 4 7 10
6. Electrical 8 17 9
7. Water Treatment & Piping 35 1 1
8. Controls 2 2 2
Total 100'70 100% 100%
-------
D-60
MATERIAL BREAKDOWN
Pelletizing Plant (Moving Grate) - Dust Collection
1. Foundations
2. Ductwork and Stack
3. Collector
4. Fan and Motor
5. Structural
6. Electrical
7. Water Treatment and Piping
8. Control
Total
Pelletizing Plant (Shaft Furnace)
Cyclone
Wet
Scrubber
5
2
15
5
45
40
14
20
7
5
11
8
1
18
2
2
100%
100%
Process Exhaust Material Handling
Cyclones Cyclones Wet Scrubbe...
1. Foundation 2 4 2
2. Ductwork and Stack 10 20 12
3. Collector 41 34 39
4. Fan and Motor 18 13 18
5. Structural 11 12 7
6. Electrical 12 13 8
7. Water Treatment & Piping 2 1 12
8. Controls 4 -2. 2
Total 100% 100% 100%
-------
D-61
MATERIAL BREAKDOWN
Blast Furnace
1. Foundations
2. Ductwork and Stack
3. Collector
4. Fan and Motor
s. Structural
6. Electrical
7. Water Treatment and Piping
8. Control
Total
Two Stage Venturi
Scrubber System
3
10
46
7
6
25
3
100%
-------
D-62
MATERIAL BREAKDOWN
Basic Oxygen Furnace
Wet Scrubber Electrostatic Fabric
(High Energy) Precipitator Filter
1. Foundations 4 3 2
2. Ductwork and Stack 30 36 37
3. Collector 10 31 32
4. Fan and Motor 9 5 6
5. Structural 6 6 5
6. Electrical 7 8 7
7. Water Treatment and Piping 31 7 7
8. Controls 3 4 4
Total 100% 100% 100%
Open Hearth ,Furnace
Wet Scrubber Electrostatic Fabric
(High Energy) Precipitator Filter
1. Foundations 3 2 2
2. Ductwork and Stack 21 25 25
Modifications
3. Collector 15 40 42
4. Fan and Motor 6 2 4
5. Structural 9 9 9
6. Electrical 11 10 8
7. Water Treatment and Piping 33 9 7
8. Controls 2 3 3
Total 100% 100% 100%
-------
D-63
MATERIAL BREAKDOWN
Electric Arc Furnace (Direct
Extraction Fume System)
Wet Scrubber Electrostatic Fabric
(High Energy) Precipitator Filter
1. Foundations 4 3 2
2. Ductwork and Stack 22 .31 35
3. Collector 15 35 34
4. Fan and Motor 9 6 7
5. Structural 7 6 7
6. Electrical 10 10 9
7. Water Treatment and Piping 30 5 2
8. Con t ro1s .2 4 4
Total 100% 100% 100%
Electric Arc Furnace (Combination Direct
Evacuation Control and Furnace Canopy- Type
Area Ventilation System)
Fabric Filter
1. Foundations 3
2. Ductwork and Stack 37
3. Collector 27
4. Fan and Motor 10
5. Structural 10
6. Electrical 7
7. Water Treatment and Piping 1
8. Controls 5
Total 100%
-------
D-64
MATERIAL BREAKDOWN
Scarfing
Wet Scrubber Electrostatic
(High Energy) Precipitator
1. Foundations 2 2
2. Ductwork and Stack 12 20
3. Collector 30 55
4. Fan and Motor 24 7
5. Structural 5 3
6. Electrical 16 7
7. Water Circuit 7 2
8. Controls 4 4
Total 100% 100%
Power Plant Boiler
Electrostatic
Cyclone Precipitator
1. Foundations 6 3
2. Ductwork 13 33
3. Collector 40 20
4. Fan and Motor 19 2
5. Structural 8 19
6. Electrical 12 17
7. Water Treatment and Piping 0 1
8. Controls 2 5
Total 100% 100%
-------
, D-65
GASEOUS POLLUTANTS
The present report does not present cost data on equipment for the
control of gaseous pollutants. Methods for the chemical treatment of gases
for the removal of sulfur.' and nitrogen. oxides are s till under development.
Reliable plant cost data will ~ot be available for some time.
Volatiles emitted during the processing of coke oven by-products can
generally be controlled by careful operating control of leaks, drips, drains,
and vents. Any waste gases from flare stacks will probably contain sulfur
oxides, for which treatment methods are not commercially available.
AREA VENTILATION AND EMISSION CONTROL
While the technology for cleaning of effluent material contained in
exhaust ducts from enclosed processes has reached a state of development
where clearly defined practices and equipment can be specified, the means
to clean areas where process materials enter or leave'the process enclosure
and to clean the ventilated air from shop structures and outside handling
areas is only now developing. Until the sizing and alternate methods have
been tested by sufficient application, and competitive pricing has evolved,
a definitive estimate of. the cost and performance of truly adequate control
means is premature. .
The ventilation air vol~mes may be ~any times the volume of the gases
cleaned in ducted exhaust circuits from the process; and explosion hazards
at times occur with the influx of air. An example estimated here at the
current level of development is the electric arc furnace melt shop with a
combination of direct evacuation control at the furnace and a canopy above
the furnace. This system provides containment and control during all phases
of the heat cycle when the furnace roof is in place, and, in addition, good
control in preventing fume from escaping from the building during those
operations' when there is no local containment at the furnace, such as
charging, teeming,andslagging. 'The added volume to the collector is 1 to
2.5 times gre'ater than with furnace flue gas treatment alone, or 2 to 3.5
times greater for the total system. In a case where the canopies are
installed higher, at the roof truss, with no direct furnace evacuation,
the volume is 4 to Stimes greater than it would be if shell evacuation'
alone were to be uSed. The basic oxygen furnace fume system, sized for
peak volumes during oxygen lancing, could be fitted with auxiliary hoods
and dampers to accommodate the hot metal charging and teeming area at low
level, utilizing this peak evacuation capacity 'for area ventilation during
off-blow periods. The same external operations at the open hearth would
require added exhaust capacity, used in turn on each furnace of the shop.
Drafts in the shop seriously effect the '''catch'' of open hoods, especially
high-lofted canopies as applied to arc furnaces. .
Alternate means ,are being applied for exhausting and fume removal.
These include various pickup devices:
-------
D-66
1.
Close fitting hoods (with relatively low volume
required) applied to pickling tanks and roll stands
for mist pickup.
2.
Low auxiliary hoods and partial enclosures applied
to pouring operations of hot iron or steel, or the
crushing, screening, loading and discharging of dry
materials (sinter, ore, coal, coke, fluxes and other
chemicals).
3.
Tunnels as applied to scarfing units and conveying
lines.
4.
High canopies with isolation dampers for selective
ventilation of high concentration dust areas, and
total building air-change systems are currently
being evaluated at a few melt shops. Buildings
to enclose extensive areas of material handling
and open processing with many dust generation
points or discharges that are difficult to control
at the source, are used to some extent now (at
crushing and screening stations, for example).
In principle, the enclosure of such an area with cleaning and possibly
recycling of the ventilation air therefrom could effect a reduction in volume
and system complexity compared to that for many high pickup canopies. In
practice, however, while emissions to the atmosphere could be significantly
reduced, hazards would in many cases accompany returning air from the collector
discharge to the workspace, limiting application of this principle. The
magnitude of the task suggests the need for less costly, more effective,
close-to-source control means. The volume required for. adequate entrainment
of emissions varies greatly, becoming much larger and less effective when
pickup devices are farther removed from the source.
And while concentrations of pollutant material can be measured at points,
the open-air distribution of concentration cannot be adequately profiled. The
concentration of an air borne material beyond the plant area is subject to the
weather and fall-out variables. Therefore, research is needed to quantitatively
evaluate an area atmosphere by means that could be used for design criteria by
equipment manufacturers and would give correlated information in performance
guarantee tests and the abatement inspector's spot check.
With enough application of engineering design, less expensive means of
controlling presently uncontained volumes will evolve; Plant design can
accomplish some grouping of.high dust areas to reduce ventilation requirements.
Process change and new equipment design will increasingly consider pollution
problems as a factor. Building design, currently based on natural ventilation
means, could undergo changes to reduce the extent and facilitate the means of
ventilation. And with optimization of means, a more realistic cost level will
in time evolve.
-------
D-67 and D-68
Some prior cost tables give estimates of costs for ventilating dust
and mist areas and cleaning the captured air around several processes. The
estimates represent the most adequate systems currently being applied or
quoted for process ventilation needs to supplement ducted process gas
exhausting and cleaning:
Sinter plant material handling,
Arc furnace canopy-type area ventilation,
Scarfing tunnel evacuation,
Pickling line mist removal,
Rolling mill mist pickup.
Note: The report by Swindell-Dressler Company appears in its entirety
as Appendix C of the companion "Final Report on a Cost Analysis of
Air-Pollution Controls in the Integrated Iron and Steel Industry", dated
May 15, 1969; and an adaptation of that portion not included in this
Appendix D is presented in pages VI-36 to end of Section VI of this
technological report.
-------