Smarajit Sarkar

Department of Metallurgical and Materials Engineering

NIT Rourkela

Ahindra Ghosh and Amit Chatterjee: Ironmaking and Steelmaking Theory and Practice, PrenticeHall of India Private Limited, 2008
Anil K. Biswas: Principles of Blast Furnace Ironmaking, SBA Publication,1999
R.H.Tupkary and V.R.Tupkary: An Introduction to Modern Iron Making, Khanna Publishers.
R.H.Tupkary and V.R.Tupkary: An Introduction to Modern Steel Making, Khanna Publishers.
David H. Wakelin (ed.): The Making, Shaping and Treating of Steel (Ironmaking Volume), The
AISE Steel Foundation, 2004.
Richard J.Fruehan (ed.): The Making, Shaping and Treating of Steel (Steeelmaking Volume), The
AISE Steel Foundation, 2004.
A.Ghosh, Secondary Steel Making Principle & Applications, CRC Press 2001.
R.G.Ward: Physical Chemistry of iron & steel making, ELBS and Edward Arnold, 1962.
F.P.Edneral: Electrometallurgy of Steel and Ferro-Alloys, Vol.1 Mir Publishers,1979
B. Ozturk and R. J. Fruehan,: "Kinetics of the Reaction of SiO(g) with Carbon Saturated Iron":
Metall. Trans. B, Vol. 16B, 1985, p. 121.
B. Ozturk and R. J. Fruehan: "The Reaction of SiO(g) with Liquid Slags, Metall. Trans.B,
Volume 17B, 1986, p. 397.
B. Ozturk and R. J. Fruehan:.Transfer of Silicon in Blast Furnace": , Proceedings of the fifth
International Iron and Steel Congress, Washington D.C., 1986, p. 959.
P. F. Nogueira and R. J. Fruehan: Blast Furnace Softening and Melting Phenomena - Melting
Onset in Acid and Basic Pellets", , ISS-AIME lronmaking Conference, 2002, pp. 585.

INTRODUCTION
INTRODUCTION

The Blast Furnace route is the dominant route for the

production of iron for steel making.

India produces around 67 million tonnes of crude steel per

anum out of which 57% is from the pig iron produced
through the blast furnace.

The respective figures are 1400 million tonnes and 72.5%

world wide.
Average coke consumption in Indian blast furnace is around
550-600kg/THM
Respective figure for advanced countries is around 350400kg/THM
Average Si content in pig iron is 0.8-1.0% in India.
Respective figure in advanced countries is 0.4-0.6%.

 B.F.

process is the first step in Producing Steel

From Iron Oxide.

This

Would remain so probably at least for the first

quarter of the century despite
Speedy depletion of Coking coal reserves
Enhanced adoption of alternate routes for iron making for
ultimate conversion to steel.

 The

B.F. works on a counter current principle

Ascending hot gases meet Descending solid
charge
The charge includes Iron bearing materials (ore,
sinter, pellets), coke & flux (Lime stone, Dolomite)
The ascending gases cause reduction of Iron
oxide in the Iron bearing materials while
progressively heating it.
The result is Production of
Liquid slag
Liquid Metal
B.F. Gas of considerable calorific value

0.8-0.6t
0.5-0.6t
1.7-1.8t

3200m3
+
80kg dust

Fuel
Reducing agent supply
Permeable bed
(spacer)

2500 m3
0.6t

1t
7

An iron blast furnace produces pig iron of the following

composition:
Fe-92.8 %, C-3.8%,.Si-2.1%, P-0.9% and Mn-0.4%
The ore smelted analyzes as follows:
Fe2O3-78%, SiO2-8.4%, Al2O3-5% and rest is
MnO ,P2O5 etc.
Assume that 99.5% of the iron ore is reduced and 0.5%
slagged. Calculate the weight of ore used to produce 1
ton of pig iron.

10

11

12

13

All the reduced elements join the metal. A typical

composition of the Metal (Iron) produced in Blast
Furnace is presented below.

14

The Slag is a low melting chemical compound formed by

the chemical reaction of the gangue and the flux in the
charge.

All unreduced ones join the slag

The major constituents of the slag include the following

Al2O3
20.45%
CaO
32.23%
SiO2
33.02%
MgO
9.95%
S
0.89%
MnO
0.54%
TiO2
1.01%
FeO
0.41%
K2O+Na20 1%
Trace Oxides 0.5%

(Curtsey TATA STEEL)

15

16

Smarajit Sarkar
Department of Metallurgical and Materials Engineering
NIT Rourkela

17

Blast furnace productivity depends upon an optimum gas

through flow as well as smooth and rapid burden descent.

The character of the gas and stock movements is intimately

associated with the furnace lines.

The solid materials expand due to heating as they descend

and their volume contracts when they begin to soften and
ultimately melt at high temperatures in the lower furnace.

18

 A further volume contraction occurs when the solid coke burns

before the tuyeres.

An enormous volume of the combustion gas has to bubble

through the coke grid irrigated with a mass of liquid metal and
slag.

An optimum furnace profile should cater to the physical and

chemical requirements of counter flow of the descending solid,
viscous pasty or liquid stock and the ascending gases at all
places from the hearth to the top

cont
19

Only then, an optimum utilization of the

chemical and thermal energies of the
gases as well as a smooth, uniform and
maximum iron production with minimum
coke rate will be realized.

20

In an integrated steel works the capacity of the

Blast Furnace depends upon
The capacity of the works.
The process of steelmaking adopted.
The ratio of hot metal and steel scrap in the
charge.
Consumption of foundry iron in the works.
Losses of iron in the ladle and the casting
machine.
The number of furnaces to be installed
o

21

22

Stock line: The distribution pattern at the top.

Charge or stock level in the furnace throat
The materials or the stock or the burden should
be properly distributed for uniform distribution of
the ascending gas.
Zero stock line: Horizontal plane formed by
bottom of big bell when closed. 6ft stock level for
instance located 6ft below zero stock line.

23

It is the volume of Blast Furnace occupied by the charge

materials and the products , i.e. the volume of furnace
from the stock line to the tap hole.
Useful volume = the furnace capacity C.U.U.V.
C.U.U.V = coefficient of utilization of useful volume.
The value of C.U.U.V. varies in a wide range from 0.481.50 m3/ton of pig iron

24

V =k D2H
V=Useful volume
H=Total height
D=Diameter at the bottom of the shaft
K=A coefficient usually lies with in the range of 0.47
to 0.53. High value is for slim profile.

25

Total height = useful height +distance between stock line

and the charging platform (it is governed by the
construction of gas off-take and charging platform, this
dimensions varies from 3 to 4m.)
Useful height= height from the tapping hole to the stock
line.
The height of the blast furnace is mainly governed by the
strength of the raw materials, particularly that of coke.
cont

26

The strength of the coke charged to the

furnace should be sufficient to withstand the
load of raw materials without getting
crushed. Coke provides permeability(in the
dry as well as wet zones )and also
mechanical support to the large charge
column, permitting the gases to ascend
through the voids.
Total height (H)= 5.55V0.24
Useful height (H0) =0.88H

27

Diameter:
The belly /bosh parallel is the cylinder that
connects the tapers of the shaft and the bosh.
Its diameter, dbll, and the ratio of this diameter to
the useful or inner height of the furnace as well
as to the diameter of the hearth play an
important role in the operation of the furnace.
The correct descent of the stock, ascent of the
gas and efficient utilization of the chemical and
thermal energies of the gas depend greatly upon
these ratios.
28

The importance of an adequate belly diameter lies in the

fact that softening and melting of the gangue and
formation of the slag occurs in this region.
An increase in the diameter facilitates gas passage
through the sticky mass and also slows down stock
movement, thus increasing the residence time for indirect
reduction.
However, the belly diameter cannot be increased
arbitrarily as it is directly related to bosh angle, bosh
height, hearth and throat diameters and useful height.

29

The belly height depends upon the softenability of the

ferrous burden and also on the shaft angle desired.
If the slag fusion occurs at higher temperatures and in a
narrow temperature range as in the case of pre-fluxed
burden, the hydraulic resistance decreases in the
vertical cross-section and the belly height can be
correspondingly reduced.

dbelly =0.59 (V)0.38

HbelIy = 0.07H

30

The hearth is designed such that its volume

between the iron notch and tuyeres is sufficient
to hold the molten metal and the slag.
The dia of hearth depends upon:
The intensity of coke consumption.
The quality of burden.
The type of iron being produced.
D hearth =0.32 V0.45

31

A very approximate relationship between the

coke burning rate and hearth diameter is
given by the following equation:
D = c Q 0.5
D = hearth diameter, m
Q = coke throughput, tonnes/24h
c = throughput coefficient which varies
between 0.2-0.3 depending upon burden
preparation.

32

For highly prepared burden, the value of

c = 0.2 has been achieved in modern large
furnaces .
Therefore, for a furnace planned to produce
10,000 THM per day with a coke rate of
500 kg/THM, i.e., a coke throughput of
5,000 tonnes per day, the hearth diameter
should be about 14.1 m.
The value will be 21.2 m if the value of
c=0.3.
33

With increasing diameter of the hearth,

the gas penetration must be ensured
by providing adequate bed
permeability with the use of
mechanically strong, rich, pre-fluxed
burden of uniform size and low slag
bulk as well as strong lumpy coke.
The Hearth height should be 10% of the
total height of the furnace

34

The shaft height must be sufficient to allow the

heating, preparation and reduction of ore before
the burden reaches the bosh. In the upper
regions of the shaft , volume changes due to
increase in temperature and carbon deposition.
These demand an outward batter for smooth
flow of materials. In the lower region of the
shaft , the material starts fusing and tends to
stick to the furnace wall. So to counteract the
wall drag an outward butter is necessary.
35

Stack height Hstack = 0.63 H- 3.2 m

Stack angle
The stack angle usually ranges from 850 to 870
(i) 850 for weak and powdery ores;
(ii) 860 for mixture of strong and weak, lumpy or
fine ores;
(iii) 870 for strong, lumpy ore and coke.

36

The

variations in the angles are necessary

for obtaining an adequate peripheral flow
which is an essential pre-requisite for
forcing of the blast furnace.
Since the ore hump is located in the
intermediate zone and it moves almost
vertically downwards pushing the lighter
coke towards the wall and the axis.
A smaller shaft angle in the case of weak
and powdery ore helps to loosen the
periphery.

37

Stack angle can be calculated from the formula

Stack angle ()= Cot-1(D-d1/2xStack Height)
Where, D= Bosh parallel Diameter
d1= Throat Diameter
Bosh angle can be calculated from the formula
Bosh angle ()= Cot-1(D-d/2xBosh Height)
Where, D= Bosh parallel Diameter
d= Hearth Diameter

38

When the raw materials are charged into the

blast furnace, little volume change takes place
for a few meters of their descent and hence the
walls of the throat are generally parallel
Throat diameter can not be too small as it has to
allow the enormous volume of the gas to pass
through at a reasonably low velocity to maintain
adequate solid gas contact and to decrease the
dust emission, throat hanging and channeling.

Cont..

39

 Throat

diameter can not be too wide as it

may compact the charge. A certain
velocity and lifting power of gas is
necessary for losening the charge at top.
Throat Diameter d throat =0.59 V0.35
Where, V= useful volume

40

A considerable amount of slag and iron descends to

the hearth through the inter-tuyere zones. If they do
so without having been adequately heated, the
thermal state of the hearth may be disturbed with
attendant high sulphur in iron, sluggish slag
movement, erratic metal analysis, frequent tuyere
burning, etc.

41

The distance between the adjacent tuyeres

around the hearth circumference should be such
as to obtain, as far as possible, a merging of the
individual combustion zones of each tuyere into
a continuous ring.

42

43

The number of tuyeres mainly depend upon the

diameter of the hearth. The diameter of the
tuyeres depend upon the blast volume.
The following formulae can be used to determine
the number of tuyeres
Pavlov: n = 2d +1
Rice:
n = 2.6d-0.3
Tikhomirov et al : n = 3d-8
Where n= Number of tuyeres,
d=hearth diameter
44

Capacity

2000

3000

5000

Useful Volume (m3)

1700

2550

4250

Total Height (m)

33.08

36.46

41.22

Useful Height (m)

29.11

32.08

36.27

Bosh Parallel Dia (m)

9.96

11.62

14.11

Bosh Parallel Height (m)

2.32

2.55

2.89

Bosh Height (m)

4.37

4.81

5.44

9.1

10.92

13.74

Hearth Area (m2)

65.04

93.66

148.27

Hearth Height (m)

3.308

3.646

4.122

Stack/Shaft Height (m)

17.64

19.77

22.77

Throat Dia (m)

6.87

7.85

9.29

Bosh Angle (0)

84.32

85.84

88.05

Stack Angle (0)

85

84.55

83.96

Nos. of Tuyeres

20

25

34

(THM/Day)
Parameter

Hearth Dia (m)

45

46

 This is a unique design in which

large bell is replaced by a distributor
chute with 2 hoppers
A rotating chute is provided inside
the furnace top cone
Advantages:
Greater charge distribution
flexibility
more operational safety and
easy control over varying
charging particles
Less wearing parts: easy
maintenance
47

48

The advantages accruing from improved distribution

control can be summarised as follows:
Increased productivity, decreased coke rate, improved
furnace life .
Reduced refractory erosion
Improved wind acceptance and reduced hanging as well
as slips
Improved efficiency of gas utilisation and its indirect
reduction
Lower silicon content in hot metal and consistency in the
hot metal quality
Reduced tuyere losses and minimisation of scaffold
formation
Lower dust emission owing to uniform distribution of fines.

49

50

As has been made clear that even the most efficient of the
modern blast furnace would produce an effluent gas containing a
significant proportion of CO which could not be used for iron
oxide reduction. The actual CO content may vary around 20-30%
by volume. This has a calorific value of nearly 900 kcal/m 3. The
quantity of gas produced depends upon the amount of fuel burnt.
For one tonne of coke burnt nearly 4000 m 3 of effluent gas may
be produced. Hence a blast furnace requiring 1000 t of coke per
day would generate nearly 4 x 106 m3 of gas with a total energy
content of 3600 x 106 kcal which is nearly equivalent to 500 t of
coke.
51

The effluent gas from the furnace cannot directly be

used as a fuel since a substantial quantity of dust from
the burden is also discharged along with. It may lead
to accumulation of dust and wear in the equipment
using the gas. The gas is, therefore, cleaned before its
use and in so doing the sensible heat of the gas is
invariably lost. The chemical heat of the cleaned gas
is what is utilised.

52

The average dust content may vary in the range of 7-30 g/m3. In general
cleaning is carried out in three stages viz. coarse, semi-fine and fine
cleaning. The coarse cleaning is done in dust catchers and cyclones in
dry condition. The dust content of the coarse cleaned gas is nearly 5-10 g/m3.
The semi-fine cleaning is carried out in scrubbers, ventury washers,
cyclone separators, centrifugal disintegrators, feld washers or even in
electrostatic precipitators. The dust content is thereby reduced to 05-15
g/m3. Fine cleaning is carried out mainly by electrostatic precipitators or at
times by high speed rotary disintegrators, The dust content is thereby
reduced down to 0.01 g/m3 The semi-fine and fine cleaning is carried out
either in wet or dry condition. Wet methods are generally preferred to dry
methods for their better efficiency and smooth working.

53

54

Two adjacent uptakes are joined together to form one single duct
and the two such ducts, thus formed, are connected to form only
one duct which carries the gas downwards into the dust catcher.
The downcoming pipe or duct is called downcomer.

A bleeder valve is a safety device, which opens automatically or is

opened, to release extra pressure developed inside the furnace and
thereby eliminate the danger of explosion.

The uptakes and the downcomers are steel pipes and are lined
from inside with firebricks. The sizes of the uptakes and
downcomers and the angle of their joints are such that gas flows out
of the furnace smoothly without any hindrance.

55

The uptakes should be located on the furnace-top

periphery at those points which are not directly vertically
above the iron-notch, slag notch, blast main entrance to
the bustle pipe, etc. These are active points of the
furnace and if the uptakes are located right above these
points it may cause uneven distribution of the gas
through the burden. The entire design should also
ensure that minimum of dust is carried form the furnace
with the gases.

56

57

58

It essentially consists of a tall cylindrical structure

comprising of a combustion chamber and heat
regenerator unit of checker bricks. The clean blast
furnace gas is burnt in the combustion chamber
and the hot products of combustion later heat up
the checker bricks. In this case the stove is said to
be on 'on-gas' and is maintained on gas until the
checker bricks are heated to a certain
temperature.
59

Firing is stopped and cold blast is passed through

checkers which impart the heat stored in them and
there by produce preheated blast. The stove is
said to be 'on blast'. It can continue heating the
blast till a certain minimum temperature of the
blast is obtainable. The stove is again put on gas
and the cycle is repeated.

60

The stove design and the number of stoves, employed

should ensure a steady supply of preheated blast to the
furnace. This duty demands that the amount of heat
generated by way of combustion of gas per unit time
should be adequate to heat up the required amount of
blast to the required temperature per unit time, taking
into account the usual efficiency of heat transfer via
checker system and the usual heat losses from the
system.

61

The thermal efficiency of the stove varies between

75-90%. The checker work cools more rapidly
whereas it takes longer time to heat it up. In practice
a stove may be on gas for 2-4 hours and on blast for
1-2 hours. For an uninterrupted steady supply of
blast at specified temperature therefore a battery of
at least three stoves is necessary. A two stove
system is quite unsatisfactory and hence three or
four stove system is preferred.
62

The checkerwork has to absorb maximum heat at faster rate while

heating and should desorb heat equally rapidly to the incoming cold
blast. The larger the weight of bricks the more will be its heat storing
capacity. The larger is the surface area exposed as flues the faster is
the heat exchange with gas. The bricks should have maximum weight
with maximum surface area of flues i.e. maximum openings to allow free
passage of gases. It has been found that a ratio of weight of bricks
in kilogram to heating surface in square metres of about 5-6 in
minimum. Below this structural difficulties may arise.

63

The checker bricks are supported on steel grids which in turn

are supported by cast iron or steel columns. Since the
maximum temperature during combustion is generated near
the dome and since the top portion of checker bricks have to
stand higher temperatures, with progressively decreasing
value downwards, the quality of checker bricks used also very
accordingly. Heavy duty firebricks are essential for dome
construction. The top 3-6 m height of the checkers is made up
of higher alumina bricks or semi-silica bricks while the
remainder as of good quality firebricks.

64

65

0.8-0.6t
0.5-0.6t
1.7-1.8t

3200m3
+
80kg dust

Fuel
Reducing agent supply
Permeable bed
(spacer)

2500 m3
0.6t

1t
66

67

68

69

Richness: Richness means the percentage of

metallic iron in the ore. e.g. In order to produce a

tonne of pig iron about1.5tonnes of ore is required
in Australia (68% Fe), about 2 tonnes are required
in India (55-60%) and nearly 3 tonnes are required
in U.K. (30-35%)
Composition of the gangue : The
composition of gangue associated with an ore
may reduce the value of an otherwise rich ore or
in some case may even enhance that of a lean
ore.
70

e.g. Value of an ore is drastically reduced by the

presence of alkali oxides , reduced to some extent
by the presence of alumina and is in fact
enhanced by the presence of lime and/or
magnesia.
Location:

The location of an ore, both

geographical and geological, is very important

Treatment

and preparation needed

before smelting

71

Cold strength
Porosity
Decrepitation
Low-temperature breakdown under reducing
conditions (LTB)
Hot compression strength
Softening temperature and range
Swelling and volume change
High-temperature bed permeability under
compressive load and reducing conditions.

72

Cold strength measurement comprises of tumbler or

drum test for abradibility, shatter test for impact and
compression test for load during storage.

Tumbler or drum test: It measures the susceptibility of

ferrous materials (coke as well) to breakage due to
abrasion during handling, transportation, charging on to
the blast furnace bells as well as inside the furnace itself.
In this test, a certain weight of the material within a
selected size range is rotated in a drum of given size for
a given time with certain number of revolutions.
73

The abrasion strength is given by the percentage

weight of + 6.3 mm surviving the test and dust
index by the percentage of - 0.6 mm. For good
pellets the respective percentages are 85-95 and
3-7, for sinters 60-80 and 5-10 and for ores they
vary greatly, 60-95 and 2-25.

74

The abrasion strength is given by the percentage

weight of + 6.3 mm surviving the test and dust
index by the percentage of - 0.6 mm. For good
pellets the respective percentages are 85-95 and
3-7, for sinters 60-80 and 5-10 and for ores they
vary greatly, 60-95 and 2-25.

75

In order to minimize the amount of fines delivered to the

furnace, a practice attracting an interest is to deliberately
subject the materials, especially coke and sinter, to
mechanical breakdown and stabilize the charge, e.g., by
means of vibrating screens. They break where the bonds are
weak and the undersize screened out.

However, it cannot be helped if any fines are generated

between charging into the skip car and then into the furnace.

76

In order to minimize the amount of fines delivered to the

furnace, a practice attracting an interest is to deliberately
subject the materials, especially coke and sinter, to
mechanical breakdown and stabilize the charge, e.g., by
means of vibrating screens. They break where the bonds are
weak and the undersize screened out.

However, it cannot be helped if any fines are generated

between charging into the skip car and then into the furnace.

77

Shatter test: It measures the susceptibility to breakdown due to

impact during loading, unloading and charging into the furnace.

In this test a certain weight of material is allowed to fall on a steel

plate from a certain height for a pre-determined number of times
and the amount of undersize measured. For strong sinters the
percentage +10mm surviving is above 80.
Compression test: It is used mainly for pellets. Pellets, unreduced
or reduced to various degrees, are subjected to compressive load at
ambient or high temperatures and the percentage of + 5 mm yield
measured and correlated with blast furnace performance.

78

Porosity: While ores and pellets possess mostly open pores, in

sinters there are macro- and micro-pores as well as open and
closed pores (cut off from outside and cannot be reached by
gas).
True porosity and hence closed porosity can be determined from
open porosity which can be measured from the true and bulk
densities.
Although reducibility increases with increasing open porosity, the
latter changes continuously during reduction on load. Generally,
a high initial porosity results in earlier softening of the material.

79

Decrepitation : When iron bearing materials are suddenly

exposed to the exhaust gas temperature at the stock level on
charging, breakdown may occur due to thermal shock. This is
known as decrepitation.
Experimentally it is measured by dropping a known weight of
material in a furnace previously heated to a temperature level
of 400600C, under normal atmosphere, inert atmosphere or
under mildly reducing conditions. After the charge attains the
temperature it is removed, cooled and sieved to measure the
breakdown.

80

 In a typical test 500 g of 20-40 mm size undried ore is

dropped in a furnace previously heated to a temperature
level of 400C and retained there for 30 min under a flow
rate of 5000 litres of nitrogen per hour. The sample is
then removed, cooled and the percentage of 05 mm and
-56 + 05 mm material in the product is determined by
sieving.
It is believed that ores with more than 10% porosity will
not decrepitate.
81

 In a typical test 500 g of 20-40 mm size undried ore is

dropped in a furnace previously heated to a temperature
level of 400C and retained there for 30 min under a flow
rate of 5000 litres of nitrogen per hour. The sample is
then removed, cooled and the percentage of 05 mm and
-56 + 05 mm material in the product is determined by
sieving.
It is believed that ores with more than 10% porosity will
not decrepitate.
82

Low-Temperature Breakdown Test (L.T.B.T.)

It has been observed in the experimental blast furnace that the iron
bearing materials do disintegrate at low temperatures under mildly
reducing conditions, that is in the upper part of the stack, affecting
the furnace permeability and consequently the output adversely. It is
believed that deposition of carbon in this region of the stack is also a
contributory factor although with sinters the breakdown has been
associated with the presence of micro-cracks.
In essence the test consists of subjecting the charge to static bed
reduction at low temperatures in a rotating furnace for a fixed dura
tion. The percentage of fines generated is quoted as the
L. T.B. T. index.

83

84

Lump ores, sinter and pellets disintegrate into smaller pieces during their
downward travel through the blast furnace owing to the weight of the
overlying burden, as well as abrasion and impact between the burden
materials.
It has been found that this tendency gets aggravated when the oxides are in
a reduced state. Reduction of hematite into magnetite occurs in the upper
stack at 500-600C, and this is accompanied by volume expansion even to
the extent of 25%.
This results in compressive stresses being developed and contributes
significantly to breakdown of the iron oxides.
Blast furnace operators prefer a low RDI (below 28 or so) since the adverse
effect of high RDI has been clearly demonstrated in practice.

85

86

Reducibility is the ease with which the oxygen

combined with iron can be removed indirectly.
A higher reducibility means a greater extent of
indirect reduction that may be obtained in the
blast furnace resulting in a lowered coke rate
and higher productivity.

87

Reducibility of ferrous materials is characterized by their

fractional oxygen removal rates in gaseous reducing
atmosphere. The percent degree of reduction or
percent fractional oxygen removal is given by

Where
n0 = number of moles of oxygen originally combined with iron only;
n = number of moles of oxygen left combined with iron after experi
mental time, t.

88

A schematic representation of relationship between reduction at

40% degree of reduction and 60% degree of oxidation levels,

89

90

The reduction of the iron oxides takes place in a series of

sequential steps. The overall rate will be determined by the
slowest of the process or processes in the series. The possible
consecutive steps are:

transport of gaseous reductant from the bulk gas phase to the

particle surface through a boundary gas film;

molecular diffusion of the gaseous reductant through the

product layer to the reaction interface;

adsorption of the gaseous reductant at the interface;

reaction at the interface (reaction between adsorbed

reductant and oxygen of the lattice);
91

desorption of gaseous products from the interface;

mass transport of iron and oxygen ions and transformations in the

solid phase; formation and growth of the reaction products, viz.,
magnetite, wustite and iron; .

molecular diffusion of gaseous products through the product layer

to the particle surface;

transport of the gaseous products from the particle surface

through the boundary gas film to the bulk gas phase.
From the above possibilities, the rate limiting cases are:
chemical control = steps (iii) to (vi)
diffusion control = steps (i) and (viii); (ii); (vi) and (vii)
92

particle size
porosity
crystal structure
pore size
volume change
impurities

93

Reduction of natural hematite ores by CO or H2 starts between 2005000C, depending upon the physical characteristics and
mineralogical composition. However, the rate below 5000C is
sluggish.
Hematite is more reducible than magnetite although the amount of
oxygen to be removed per unit weight of iron is about 12 percent
higher in the former.
The better reducibility of hematite may be due to:

formation of porous wustite from hematite, easily accessible to

reducer gas whereas magnetite forms dense wustite during
reduction;

94

tendency of hematite to break down and expose larger

surface due to expansion in volume during reduction to
magnetite ;

pores in hematite are more elongated and the microporosity

larger; magnetite has larger grain size and is more closely
packed;

a higher value of overall rate constant for wustite reduction

since the wustite lattice formed during reduction of hematite
exhibits a higher degree of disorder than that formed from
magnetite.
95

96

Since the blast furnace is a counter-current process, the tuyere gas

with high CO-content comes into contact with the wustite which
needs a very high reduction potential for conversion to Fe. The
resulting lower potential gas, as it ascends, encounters magnetite
and hematite which require much smaller equilibrium CO/C0 2 ratios
for reduction to lower oxides. Hence, the reduction of wustite is of
prime importance in iron oxide reduction, especially because about
70 percent of hematite oxygen is present as wustite.

97

98

The efficiency of operation of a blast furnace may be

measured in terms of coke rate which should of course
be as low as possible. The achievement of a satisfactory
coke rate depends on optimising the extent to which the
carbon deposition reaction proceeds. If the top gas is
high in C02 sensible heat is carried from the furnace as a
result of the exothermic reaction.
2CO=CO2+C
If on the other hand the top gas is high in CO, chemical
heat leaves the furnace.

99

10
0

10
1

The combustion of carbon to its monoxide yields only about 28

percent of the total heat possible to be obtained and only oxidation to
the dioxide gives the largest amount of heat.

This has a great significance in the blast furnace process where about
30-40 percent of the coke carbon is oxidised to CO2 and the rest into
CO.

In the blast furnace about half of the heating value of carbon is

obtained for a top gas CO/CO2 ratio of 2. The rest is carried away in the
outgoing gas.

Although a part of this heat is redeemed by burning the gas out of the
furnace, it means conversion of expensive metallurgical coke to a gas
which could be obtained from a less valuable fuel.
10
2

The heat values in the given equations are for pure carbon. For
a good quality coke with 86% C, the respective values will be
about 2000 and 7000 kcal/kg.coke.

Carbon dioxide, in the absence of carbon, is stable at high

temperatures.

Up to about 2000C, its thermal dissociation is small. At

atmospheric pressure, it amounts to about 2 percent.

At high temperatures, carbon dioxide is unstable in the

presence of carbon and is reduced to carbon monoxide
according to the given equation.

10
3

This is generally known as Boudouard or carbon gasification

reaction and is highly endothermic; the reverse reaction is
exothermic and is called carbon deposition reaction. The
former reaction is of extreme importance for the economy and
smooth running of the blast furnace. It occurs at high
temperatures, above about 10000C, and therefore consumes an
enormous amount of heat in the lower third of the blast furnace.
The carbon deposition reaction is most pronounced at
temperatures 500-550C and consequently takes place in the
upper portion of the shaft.

10
4

10
5

Since 2 moles of CO are produced from 1 mole of CO 2

(carbon being solid), the Boudouard reaction is
dependent on pressure.

The isobars in the previous figure show that at any given

temperature the equilibrium %C02 increases with
increasing pressure.

The stability of CO decreases with decreasing

temperature and increasing pressure. The maximum
instability is between 600 to 800C as shown by the steep
slope of the curves.

10
6

The reaction velocity becomes so slow below about

400C that a negligible amount of CO is decomposed
into CO2 and C.

The reaction is accelerated in the presence of catalysts

which provide nucleation sites for deposition of carbon.

In the presence of catalysts, the reverse of reaction

accelerates above 400C and reaches a maximum
between 500-5500C.

10
7

Chemical Influence
It is well known that the reduction rate of wustite is critical in the
overall kinetics of iron oxide reduction.
The equilibrium partial pressure or concentration of CO2 would
decrease if aFeO is lowered by solution and/or compound
formation. Hence, the reduction rate would also decrease.

10
8

Natural ores can contain iron oxides as compounds with gang materials, such as,
2FeO.Si02, FeO.AI203, FeO.Cr203, FeO.TiO2 etc where wustite exists in a state of low
activity. The activity of wustite can also decrease when it undergoes sintering with the
impurities present, such as SiO2, Al2O3 etc.

10
9

The reduction rate of ore increases with increase in linear velocity of

the reducing gas due to the reduction of the boundary layer
thickness at the bulk-gas/particle interface. After a critical gas
velocity is reached, there is no further increase in the rate with
increasing gas velocity since the overall rate becomes controlled or
limited by other processes. The figure shows that the limit is only
0.4 m/s. The figure also shows that the critical velocity is
independent of the degree of oxidation. In blast furnace, the
linear gas velocity does not affect the reduction rate since it ranges
between 1-20 m/s and is often exceeded.

11
0

11
1

For the reduction of iron ores the reducing gas has to diffuse
into the interior of the body where transformations can occur.
In general, the reduction rate increases with temperature but
the degree depends upon the mechanism of the reaction .
The overall reduction rate depends upon the relative
contributions of chemical control and gaseous mass
transport and hence depends upon the particular reactions
occurring and the reaction temperature. Since chemical
reaction has higher activation energy than gaseous diffusion,
the former will increase at a much greater rate with increase
in temperature than the latter.
11
2

Hence, a stage will arrive where diffusion will become ratecontrolling. Depending upon the degree of reduction, at lower
temperatures of about 500-600C, the chemical reaction
rate controls the reduction rate forming what is known as
the kinetic region in the blast furnace. At temperatures
above 600C, gaseous diffusion becomes the dominant
rate controlling mechanism. The temperature regime in the
blast furnace shaft is such that it can be assumed a zone of
mixed-control exists.

11
3

 Reactions

with high activation energies are very

temperature sensitive. Conversely, low activation

energies imply lower temperature sensitivity of rate.
Temperature

sensitivity of rate is more pronounced at

lower temperatures.
The

pre-exponential factor in Arrhenius law has little

effect on temperature sensitivity of reaction rate.

11
4

11
5

In the blast furnace , the reducing gas is

predominantly CO with varying amounts of
hydrogen depending upon the moisture content of
the blast and other blast additives like fuel oil or
natural gas. Study shows that a mixture of CO and
hydrogen appears to be a more efficient reductant
than either of them.

11
6

RAW MATERIALS for BF

COKE

Why COKE, not COAL?

1.

Natural COAL is too dense and fragile to be used in the modern BF.
(a) The blast will not be able to penetrate it quickly for burning.
(b) It is not strong enough to stand nearly 25 m high burden lying over
it.
(c) The volatiles present will be released in the BF
However, we need C for giving heat and to reduce iron ore into
iron metal

Can COAL properties be improved for this purpose?

1.

Yes, fortunately certain coals when heated out of contact with air
result in a carbonaceous mass, which is strong, porous, volatile free,
just suitable for use in BF.
This carbonaceous mass is called COKE and this coal is called
METALLURGICAL COAL or COKING COAL.

11
7

Reserves of metallurgical coals in

INDIA
Coal

Fields

quality
Prime
Coking

Jharia
(IX seam & above)

Medium Jharia, RamCoking

Semi to

garh, Raniganj
Raniganj, Jha-

weakly

ria, Jhilimili-

Coking

Sonhat
Total

Gross
Net Reserve
Reserve (Mt)
(Mt)
5288

2312

20388

2752

2067

47728

27733

5569

11
8

Types of typical metallurgical coals

Type

%
F.C.

% H2

Prime coking

88-91

Medium coking (HV)

86-88

Medium coking (LV)

%
V.M.

Cal.
Value

Kcal/kg

Moist
ure

4.8-5.2 23-32

8800

8600

1-2

90-91

5.1-5.4 22 max
4.7-4.9 22-32

8800

1-2

Semi-coking

83-85

5.4-5.8 37-44

8300

2-3

Weakly coking

83-84

5.1-5.6 38-46

8300

3-5

11
9

FUNCTIONS OF COKE
Coke fulfils the following major roles in a BF:
1. It is a fuel provides heat for endothermic chemical reactions and
2.
3.
4.
5.

melting of slag and metal.

It produces and regenerates the reducing gases for the reduction of
iron oxides.
It provides an open permeable bed through which the slag and metal
move down to hearth and hot reducing gases move upwards.
It supports the overlying burden load specifically at the lower
levels/regions.
It carburizes iron thereby lowering its melting pointand
consequentlythe working temperature of the BF is reduced.

12
0

QUALITY REQUIREMENTS
OF COKE
Coke must possess the following properties:
1. Proper chemical composition w.r.t. Fixed Carbon, ash, S, P etc.
2.
3.
4.

5.

Proper reactivity- (a) area exposed to blast, (b) temperature and

pressure of the blast, (c) affinity of the particular type of carbon for
oxygen.
Proper size range.
Thermal stability at high temperature: (a) absence of large lumps, (b)
uniform texture, (c) minimum inert inclusions, (d) high carbonisation
temperature and heat soak, (e) prior mechanical conditioning and (f)
low chemical reactivity.
Proper strength and abrasion resistance.

12
1

QUALITY REQUIREMENTS
OF COKE
Value of coke is assessed in terms of its:
1.

2.
3.
4.
5.

Room temperature strength.

High temperature strength.
Reactivity (RI).
Chemistry and
Strength after reduction (CSR)

12
2

INDIAN STANDARD SPECIFICATION

FOR BF COKE
Characteristics

Requirements
Grade I of Grade II
Grade III

Proximate analysis (dry) %

Ash (max)

20

24

28

V.M. (max)

0.70

0.70

0.70

P
Shatter Index:

0.30

0.30

Over 38 mm (% max)

85

0.30
-

Over 13 mm (%max)

97

85
97

Micum Index:
75
14

over 40 mm (% min)

75

Through 10 mm (% max0
Stability factor:

14

On 25 mm (% min)

40

40

Porosity, %

35-48

35-48

12
3

MODIFICATIONS IN COKE
MAKING PRACTICE
The modifications include:
1.

2.
3.
4.
5.
6.

Pre-heating.
Briquette blending.
Stamp charging.
Admixing oil in the charge.
Admixing water in the charge.
Using a better coal blend.

12
4

COMPARISON OF EFFECT OF
SOME PARAMETERS
Parameters

Conventi
onal
Charging

Briquette
.
blending

Binder-

Preheating

Stamp
charging

800850

11001200

8-9

5.5-6.5

45-50

60-65

105-110

110115

112115

100-105

150175

110115

less briquetting

Bulk Density of
coal charge(kg/m2)

700-750 750-800 750-800

M 10 of coke

10-11

CSR of coke

30-40

Oven throughput I

100

8.5-9.5
35-45

9-9.5
35-45

105-110
Maintenance
requirement index

100

105

12
5

The function of coke in the blast furnace is five-fold, namely,

(i) it acts as a fuel by providing for the thermal requirements in the furnace, the
reaction being,
2C + O2 = 2CO: H0 = - 2300 kcal/kg.C
On complete combustion to CQ2 the heat evolved is 8150 kcal/kg.C. Thus only
about 28 percent of the obtainable heat is supplied by coke;
(ii) it provides CO for the reduction of iron oxides;
(iii) it reduces the oxides of metalloids, such as, Mn, Si, P and others if present;
(iv) it carburizes the iron and lowers its melting point;
(v) it provides permeability (in the dry as well as the wet zones) and also
mechanical support to the large charge column, permitting the gases to ascend
through the voids.

12
6

Coke is the universal fuel used in the blast furnace. It acts

both as a reductant as well as a supplier of heat. It also
comprises the major portion of iron production cost. Now-adays other fuels are also being used as part replacement of
coke. These fuels cannot be charged from the top and as
such they are injected into the furnace through the tuyeres
along with the blast. In some countries, especially in Brazil,
charcoal is used as a blast furnace fuel.

12
7

Coke size: Coke comprises about 50-60 percent of the volume of the
charge material. The coke size is important as it provides
permeability in the dry as well as in the wet bosh zone The coke size
is always 3-4 times larger than the ore size, since coke is partially
burnt as it descends. It also has a lower density, and hence a greater
tendency for fluidisation. Of course, in the lower bosh region of a
blast furnace, coke is the only solid that remains, and which helps to
support the burden. The optimum size range for lump ore is 10-30
mm and for coke is 40-80 mm. Since the coke size becomes
smaller as it descends through the blast furnace due to mechanical
breakdown, gasification, attrition, etc., the factor of prime importance
is the strength of coke.
12
8

Coke strength: Mechanically considered, it is the quality cohesion that

prevents the coke from collapsing and tends to avoid the formation of
small particles. High cohesion or strength is related to several coke
making properties. On the basis of breakage by impact, compression
or abrasion, the coke strength should be assessed both at ambient as
well as high temperatures. Studies of the structure of different coke
samples show that the best varieties have a regular distribution of pores:
with adequate thickness and hardness of the walls between the pores
and are free from cracks generated internally. Such a structure ensures
withstanding of high compressive forces and high temperatures in the
all-important lower furnace.

12
9

The strength of coke produced in the coke-ovens is

influenced by:

blending ratio of coals of varying caking components and

proportion of the fibrous portion;

particle size and distribution of charging coal;

coke-oven temperature and combustion conditions;

moisture and addition of oil;

soaking time;

width, height and method of heating.

13
0

Ash and moisture content in coke. Low moisture & ash content
are desirable in metallurgical coke.
Low moisture in coke can be achieved by suitable control of the
water supply during quenching. That is why dry quenching using
inert gas has been suggested but dry coke is dusty and can create
handling nuisance. Metallurgical coke should contain < 1.5%
moisture. Each additional 1% H20 in blast furnace coke
increases fuel consumption by 1.2%.
Maximum desirable ash in metallurgical coke is 8-10%. Each
additional 1% ash in blast furnace coke increases fuel consumption
by 2%. In addition, high ash content becomes very undesirable, if it
causes trouble in fusion. Also, volume of slag produced in blast
furnace becomes more with the use of high ash coke resulting in
reduction of the useful volume of the blast furnace and hence
production of pig iron from blast furnace is reduced. High ash coke
has high resistance to abrasion and has high strength.

13
1

Density & Porosity. The more completely the coal is devolatalised, the
more closely the density of coke approach that of the graphite (i.e. 2.3).
Coking follows formation of a porous structure which increases in
density by shrinkage during further heating at a high temperature. Since
all cokes contain traces of hydrogen and mineral matter of the coal, its
absolute density never reaches that of graphite. The higher the rank of
the coal, the higher is the density of coke made (Varies from 1. 7 to 1.95
for C% in coal from 81 to 89%).
With increase in time and temperature of carbonization (with faster
heating rate) the density of the coke increases while the porosity
decreases. Denser coke are stronger and harder. An important property
of coke for its storage and transport is its bulk density.
High porosity is desirable in furnace cokes to obtain high rates of
combustion.
In blast furnace, high strength is considered more important
because coke has to be dropped in the blast furnace from a great height
and it should not break before reaching the hearth for better
performance and reducibility in the furnace.

13
2

It measures the resistance of coke to breakage by impact i.e. of its

strength. 50 lb. of 2 inches size coke is placed in a rectangular box
of dimension 28" x 18" x 15" placed 6 ft above a steel base plate
inch thick. The hinged base of the box is released suddenly, when
the content drop onto the base plate.

Box is dropped several times and coke is then screened through a

series of sieves made of square stamped sheet. The percentage of
coke retained on 2", 1.5" and 0.5" sieves are recorded and called
shatter index. Desirable values of shatter index for blast furnace
coke are 80% on 2" screen, 90% on 1.5" screen and 97% on 0.5"
screen.

13
3

This is a measure of both hardness and strength of the coke. 50 kg of

coke of 50 mm size is rotated in the micum drum for 4 minutes at the
rate of 25 rotation per minute (rpm). Micum drum is a cylinderical steel
drum (whose length and diameter both are 1 metre each) fitted
lengthwise with four angle irons (100 x 50 x 10 mm) 90 apart inside the
drum. After rotating the coke, it is taken out and screened through 60
mm, 40 mm, 20 mm and 10 mm round hole screen.

The percentage of coke retained on a 40 mm screen is called M40 index

(Micum forty index) whereas, the percentage of coke that passes
through a 10 mm screen is called M10 index (Micum ten index)

13
4

M40 gives the resistance of the coke to breakage by impact i.e. it is a

measure of the strength of the coke. M 10 gives the resistance of the coke to
breakage by abrasion (i.e. rubbing) and it is a measure of hardness of the
coke.

High M40 and low M10 values are desirable for metallurgical coke. As
per ISI minimum M40 index should be 75% and maximum M10 index should
be 14% for metallurgical coke.

For use in 2000 m3 volume blast furnace, coke M40 should be more
than 78 and M10 should be less .than 10.

13
5

It is defined as the ability of coke to react with O2, CO2 or steam

(H2O).

More reactive cokes have higher thermal values of their volatile matter.
Coke of high reactivity ignites easily and gives rapid pick up of fuel bed
temperature. However, low reactivity coke gives a higher fuel bed
temperature than a highly reactive coke

Reactivity is inversely proportional to the absolute density. It is affected

by the presence of easily reducible iron compounds in ash.

Coke of high reactivity is obtained from weakly caking coals or blends.

Strongly coking, high rank coals produce coke with low reactivity.
13
6

From a chemical standpoint, the coke should be of

low reactivity. The vertical distance of the indirect
reduction zone of 800-1000C, i.e., the residence time of
ore in this zone can be increased if the coke gasification
temperature be raised which is possible by the use of
less reactive coke. As for example, it is reported that an
increase of reactivity by 100% results in an increase in
the coke rate between 30-70kg/THM

13
7

For blast furnace coke, size and hardness are more

important than reactivity. Satisfactory hearth temperature is
obtained with unreactive coke containing little breeze.

Reactivity of coke is measured by Critical Air Blast method and is

reported as Critical air blast (CAB) value of coke. The CAB
value of coke is the minimum rate of flow of air in ft 3/minute
necessary to maintain combustion in a column of closely graded
material (14 to 25 B.S.) which is 25 mm deep and 40 mm in
diameter. The typical CAB value for oven coke is 0.065
ft3/minute. More reactive coke has got lower CAB value.

13
8

Another modern and current method of expressing the reactivity and

strength of coke is Coke Reactivity Index (CRI) and Coke Strength
After Reaction (CSR) which is being followed in Indian steel plants.
Coke Reactivity Index (CRI).
To determine CRI, 200 gm of coke sample (size + 20 - 25 mm) is taken in
a stainless steel tube and heated in electric furnace to 1100C. CO 2 gas at
5 kg/cm2 pressure is passed through the coke bed for two hours. CO
formed (by reaction C + CO2 = 2CO) is burnt in a burner and is exhausted
out. Carbon of coke reacts with CO2 (depending upon the reactivity level
of the coke) and there is a loss of weight of coke depending upon its
reactivity. More is the loss in weight of the coke, reactivity is more. % loss
in weight of coke is reported as coke reactivity index (CRl). Ideal CRI
value of a good blast furnace coke should be about 20%. Typically
CRI of Indian blast furnace coke is about 25%.

13
9

Coke Strength after Reaction (CSR). The left out coke

from the CRI determination test is rotated for 60 rotation
in a micum drum. And the % of coke retained on a 10
mm size screen is reported as coke strength after
reaction (CSR). Stronger the coke, more is its CSR
value. Ideal value of CSR for blast furnace coke is a
minimum of about 55%. Typically CSR of Indian blast
furnace coke is about 60-65.
14
0

Agglomeration of Iron Ore Fines

About 65 75 % of iron ore gets converted into fines
( - 5 mm ) during various operations from mining to conversion
into CLO. Majority of these fines are exported to other countries
at throwaway price resulting in greater financial loss to the
nation. Most widely used methods for the agglomeration of these
fines to render them useful for BF are Sintering and Pelletization.

Sintering Sintering is essentially a process of heating

of mass of fine particles to the stage of incipient fusion
for the purpose of agglomerating them into lumps.

14
1

To increase the size of ore fines to a level acceptable

to the BF

To form a strong and porous agglomerate

To remove volatiles like CO2 from carbonates, S from

sulphide ores etc

To incorporate flux in the sinter

To increase the BF output and decrease the coke rate

14
2

Iron ore sintering is carried out by putting a mixture Iron

bearing fines mixed with solid fuels on a permeable bed. The
top layer of sinter bed is heated up to the temperature of 1200
- 13000C by a gas or oil burner. The combustion zone initially
develops at the top layer and travels through the bed raising
its temperature layer by layer to the sintering label. The cold
blast drawn through the bed cools the already sintered layer
and gets itself heated.

14
3

In the combustion zone, bonding takes place

between the grains and a strong and porous
aggregate is formed. The process is over when
the combustion zone reaches the lowest layer of
the bed. The screened under size sinter is
recycled and over size is sent to B.F.

14
4

14
5

Two types of bonds may be formed during sintering.

Diffusion or Recrystallization or Solid State Bond : It is formed as a result of
recrystallization of the parent phase at the point of contact of two particles in solid
state and hence the name.
Slag or Glass Bond: It is formed as a result of formation of low melting slag or glass
at the point of contact of two particles, depending upon the mineral constitution, flux
addition, etc.
As a result the sinter can have three different types of constituents:

Original mineral which has not undergone any chemical or physical change during
sintering.

Original mineral constituents which have undergone changes in their physical

structure without any change in their chemistry. Recrystallization is the only change
at some of the particle surfaces.

Secondary constituents formed due to dissolution or reactions between two or more

of the original constituents
14
6

The proportion of each of the physical and chemical change during

sintering depends upon the time-temperature cycle of the process.
The higher is the temperature more will be the proportion of new
constituents by way of solutions and interactions whereas lower is
the temperature and longer is the duration more is the process of
recrystallization in solid state.
The more is the slag bonding, stronger is the sinter but with less
reducibility and, more is the diffusion bonding, more is the
reducibility but less is the strength. Since ores are fairly impure
slag bond predominates. On the other hand in rich sinters slag
bond is of minor importance.

14
7

14
8

The area under the time-temperature curves

essentially determines the nature and
strength of the bonds developed during
sintering of a given mix. For a given mix it is
most unlikely the bonds of sufficient
strength will be formed below a certain
temperature level within a reasonably short
time. Hence the area under the curve above a
certain temperature, which may be around
1000C for iron ores, is the effective factor in
deciding the extent of sintering
14
9

rather than the whole area under the curve from

room temperature to the combustion temperature
level. The nature of the time-temperature graph will
depend upon the rate of heating and cooling of a
given mix. The nature of this graph is of paramount
importance in assessing the sintering response. The
factors that affect this curve are then the variables of
the process and which should be adjusted properly
for obtaining effective sintering.

15
0

Bed permeability
Total volume of air blast drawn through the bed

Particle size of iron ore

Thickness of the bed

Rate of blast drawn through the bed

Amount and quality of solid fuel incorporated in the sinter

mixture

Chemical composition of ore fines

Moisture content in the charge

15
1

During sintering, heat exchange takes place between the solid charge
and air drawn. At any time, the air takes the heat from combustion zone
and then transfers to the lower layer of the bed. For faster rate of heat
exchange, the volume of air drawn should be more. If suction rate of
air is too high, transfer of heat may become less efficient. On the other
hand, the flame front will not move down the bed properly if suction is
less. Higher the bed permeability, more will be the air drawn. But,
higher permeability leads to loss of strength in the resulting sinter due
to reduction in bond strength. Hence a compromise is made between
these two factors. It is usual practice to draw about 700 1100 m 3 of
air/ton of charge.

15
2

An increase in particle size increases bed permeability and the

volume of air drawn.

Strength of sinter gets reduced with an increase in particle size of

the ore due to reduction in contact area.

For effective sintering, the use of larger ore lumps is undesirable.

Iron ore size > 10mm is rarely preferred.

Higher proportion of 100 mesh size fines adversely affects the bed
permeability. Better is that 100 mesh size fraction should be
screened off and used for pelletization. Ideal size of iron ore for
sintering is 0.07 10 mm.

15
3

During mining and ore dressing operations, especially

where very fine grinding is necessary for wet
concentration, a large amount of - 0.05 mm fines is
generated which are not amenable to sintering
because of very low permeability of the bed. They can,
however, be agglomerated by balling them up in the
presence of moisture and suitable additives like bentonite,
lime, etc. into 8-20 mm or larger size. These green pellets
are subsequently hardened for handling and transport by
firing or indurating at temperatures of 1200-1350C.

15
4

Pelletisation essentially consists of formation of green

balls by rolling a fine iron bearing material with a critical
amount of water and to which an external binder or any
other additive may be added if required. These green
balls of nearly 8-20 mm size are then dried, preheated
and fired, all under oxidising conditions, to a temperature
of around 1250-1350C. Bonds of good strength are
developed between the particles at such high
temperatures.

15
5

The pelletisation process consists of the following

steps:

Feed preparation.

Green ball production and sizing.

Green ball induration:

(a) Drying
(b) Pre-heating
(c) Firing

Cooling of hardened pellets.

15
6

The observations on ball formation that eventually led to the

development of the theory of balling are as follows:

Dry material does not pelletise and presence of moisture is essential

to roll the powder into balls. Excessive water is also detrimental.

Surface tension of water in contact with the particles plays a

dominant role in binding the particles together.

Rolling of moist material leads to the formation of balls of very high

densities which otherwise is attainable by compacting powder only
under the application of a very high pressure:

The ease with which material can be rolled into balls is almost
directly proportional to the surface area of particles, i.e. its fineness.

15
7

The capillary action of water in the interstices of the grains causes a

contracting effect on them. The pressure of water in the pores of the
ball is sufficiently high so as to compact the constituent grains into a
dense mass. The compressive force is directly proportional to
fineness of the grains since the capillary action rises with the
decrease in pore radius and the latter decreases with increasing
fineness. An optimum moisture is important since too little of
water introduces air inclusions in the pores and too much of water
would cause flooding and destruction of capillary action. The
optimum moisture content usually lies between 5-10 percent or
more, the finer the grains the larger the requirement.

15
8

Besides the bonds formed due to surface tension mechanical

interlocking of particles also pays a significant role in developing the
ball strength.

Maximum strength of a green ball produced from a given material

will be obtained by compacting the material to the minimum porosity
and with just sufficient water to saturate the voids. The rolling action
during pelletisation is beneficial in reducing the internal pore space
by effecting compaction and mechanical interlocking of the particles.

15
9

From fundamental studies it has been concluded that there are three
different water-particle systems:

The pendular state, when water is present just at the point of contact of
the particles and surface tension holds the particles together.

The funnicular state, when some pores are fully occupied by water in
an aggregate system.

The capillary state, when all the pores are filled with water but there is
no coherent film covering the entire surface of the particles.

16
0

16
1

The ball formation is a two stage process, i.e. nucleation or seed

formation and their growth. The formation of balls on a pelletiser depends
primarily on the moisture content. Seeds are formed only if critical
moisture level is maintained and without which the process cannot
proceed properly. Growth takes place by either layering or assimilation. It
has been observed that the size of the balls produced in a pelletiser from a
charge containing right amount of moisture depends on the time and speed
of the pelletiser, i.e. number of revolution.. Three regions can be clearly
observed, during ball formation. :
o

Nucleii formation region

Transition region

Ball growth region.

16
2

16
3

When a wet particle comes in contact with another wet

or dry particle a bond is immediately formed between the
two. Similarly several such particles initially join during
rolling to form a highly porous loosely held aggregate
and crumbs which undergo re-arrangement and partial
packing in short duration to form small spherical, stable
nucleii. This is the nucleation period, a pre-requisite for
ball formation since these very nucleii later grow into
balls.

16
4

After nucleii are formed they pass through a transition period

in which the plastic nucleii further re-arrange and get
compacted to eliminate the air voids present in them. The
system moves from a pendular state through funicular state
to the capillary state of bonding. Rolling action causes the
granules to densify further. The granules are still plastic with a
water film on the surface and capable of coalescing with other
granules. The size range of granules in this region is fairly
wide.

16
5

The plastic and relatively wet granules grow if they are

favorably oriented. In this process some granules may even
break because of impacts, abrasion, etc. Growth takes place
by two alternative modes.

growth by assimilation is possible when balling proceeds

without the addition of fresh feed material.

growth by layering is possible when balling proceeds with

the addition of fresh feed material.

16
6

Growth by Assimilation

If no fresh feed material is added for balling the rolling action may break
some of the granules, particularly the small ones, and the material
coalesces with those which grow. The bigger the ball the larger it will grow
under these conditions. Since smaller granules are weaker they are the first
victim and growth of the bigger balls takes place at their expense.
Growth by Layering

Growth of the seeds is said to be taking place by layering when the balls
pick up material while rolling on a layer of fresh feed, The amount of
material picked up by the balls is directly proportional to its exposed
surface, i.e. the increase in the size of the balls is independent of their
actual size.
Growth by layering is more predominant in the disc pelletisers and
growth by assimilation is more predominant in drum pelletisers, at
least beyond the feed zone.
16
7

In general natural lumpy ore or sinter or pellets or a suitable combination of

two or more of these form the burden.. The modern large capacity furnaces
necessarily need fully prepared burden to maintain their productivity since the
required blast furnace properties cannot just be met by natural lumpy ore. The
selection of the process of agglomeration, whether sintering or pelletising, will
depend upon the type of ore fines available, the location of the plant and other
related economic factors involved.
Sintering is preferred if the ore size is -10 mm to + 100 mesh and if it is
-100 mesh pelletising is generally adopted. Pelletising in fact requires
ultrafines of over 75% of -325 mesh. These processes are therefore not
competitive.

16
8

Minimum closure of pores by fusion or slagging; open pore

system; very good reducibility due to high microporosity .
Porosity of sinter is 10-18% and that of pellets is 20-30%.

The shape of pellets is near spherical and hence bulk

permeability of the burden is much better than that obtained
from sinter which is non-uniform in shape.

The shape, size and low angle of repose give minimal

segregation and an even charge distribution in the furnace.

16
9

More accessible surface per unit weight and more iron per unit of furnace volume
because of high bulk density, 3-3.5 tonnes/m3 .Larger surface and increased time
of residence per unit weight of iron give better and longer gas/solid contact and
improved heat exchange;

Degradation of sinter during its transit is much more than that of pellets. The
sinter therefore has to be produced nearby the blast furnace plant while pellets
can be carried over a long distance without appreciable degradation. Ease in
handling

It should also be noted that If high rates of productivity demand elimination of

fines and since sinter happens to contribute more to the generation of fines than
that of pelllets, the later will have to be chosen as the burden in preference to
sinter.
17
0

The installation cost of a pelletising plant will be 30-40% more than

that of sintering plant of an equal size.

o
o

The operating cost of sintering is slightly less than that of pelletising.

Difficulty of producing fluxed pellets.

Swelling and loss of strength inside the furnace

Fluxed pellets break down under reducing conditions much more

than acid and basic sinters and acid pellets.

Strong highly fluxed sinters, especially containing MgO, are being

increasingly preferred to pellets.

17
1

The life of lining, under the conditions prevailing inside the furnace, decides the
furnace campaign which should not be less than a few years. The chief causes of
failure of the lining are:

Carbon monoxide attack.

Action of alkali vapours.

Action of limy and alkaline slags.

Action of other volatile matters.

Abrasion by, solids, liquids and gases.

Temperature.

Action of molten metal.

Conditions of operation and design.

Blowing-in procedure.
17
2

All these factors may not be operative at all the areas in

a furnace. One or a few factors, at best, may by
dominant at anyone area in the furnace. For example, in
the stack the lining has to withstand predominantly
abrasion by solid burden and attack of carbon monoxide,
whilst in the bosh region the lining has to stand high
temperature, erosion by ascending gas and attack of
molten lime and alkali slags. Similarly the hearth has. to
stand action of molten slag and metal without breakouts.

17
3

The chief causes of failure in the furnace lining are:

Carbon Monoxide Attack. Probably the most common failure is due

to the disintegration of the bricks by carbon deposition, produced by
dissociation of the carbon monoxide in the blast furnace gas. This
takes place in the upper portions of the stack according to the
reaction :2COC02 +C.

If the bricks are porous, carbon penetration takes place, causing the
bricks to disintegrate.

Some authorities are inclined to select firebricks for this portion of

the furnace on the basis of iron oxide content

17
4

Rigby and Green, however, state that the total iron oxide content of
the firebrick is no criterion of the resistance of the material to carbon
monoxide attack, as the ease with which the iron nodules can be
reduced to metallic iron appears to be the determining factor. This
view is supported by many authorities.

The usual method of assessing the resistance of firebricks to

carbon monoxide attack is to expose the material to a current of
pure carbon monoxide at a temperature of 4500 C. and to note the
time necessary to cause disintegration. If a brick will resist this
treatment for two hundred hours it is considered sufficiently resistant

17
5

more modern method is to pass a stream of carbon monoxide through a

column of crushed firebrick for a period of four hours. The weight of carbon
dioxide present in the exit gases, determined by a CO2 absorption train, is
taken as an indication of the material's resistance to carbon monoxide attack.

Scientists conducted a series of investigations to determine the sensitiveness

of refractory bricks to carbon monoxide attack at temperature ranges of 4200
to 5000 C. The results of these experiments seemed to indicate that the
manufacturers of blast furnace linings should ascertain that:

The original clay should be free from such ferruginous substances as

pyrites and siderite;

It will also be obvious that carbon monoxide attack will be reduced by the use
of a dense brick of low permeability.

17
6

. The blast furnace burden consists of small amounts of alkalis and some
cyanogen is generated by the interaction of nitrogen and carbon. These alkalis
and cyanides are volatilized in the hotter parts of the furnace, and tend to collect
in the cooler places.

Kraner shows that the alkalis react with the brickwork forming nephelite (Na20 .
Al203 2Si02), whilst Rigby, Booth and Green quote microscopic examinations of
brickwork taken from furnace linings 20 ft. down the stack, which shows the
formation of kaliophilite (K20. A1203 2Si02) and leucite (K20. AI20a 4Si02).

Green and Hugill state that experience demonstrates that these vapours,
particularly cyanide vapours, can lead to general corrosion and modification
of the surface of the lining with consequent loss of strength and
refractoriness.

17
7

. The action of blast furnace slag on the refractory lining is a function of the
basicity of the slag. Since the ash of the coke is not released until the coke
is burnt in the tuyere zone, it follows that the CaO/Si02 ratio must be high in
the bosh. Therefore there will be a tendency for these high-lime slags to
attack the brickwork.

Other Volatile Materials

. There are other volatile materials, such as zinc and lead, which may be
included in the blast furnace burden. These metals volatilize in the hotter
zones and condense in the pores of the brickwork in the cooler parts. Their
subsequent oxidation causes a swelling of the brick, which results in
disintegration.

17
8

Probably the chief abrasion occurs on the stockline as the burden

drops from the large bell. To some extent the severity of the
abrasion in this zone will depend on the nature and size of the raw
materials. Further abrasion in the form of wall drag takes place
lower down the furnace.

Some authorities query the extent to which abrasion is responsible

for disintegration of the bricks in these lower zones. Kraner appears
to suggest that the effect of abrasion or wall drag is to accelerate
the disintegration due to carbon monoxide and alkali attack.

17
9

Unless great care is taken and patience shown during the drying of
the lining and the blowing in of the furnace, considerable damage
may be done to the lining. The importance of drying the lining
course by course as it is built is now appreciated.

When a blast furnace is in normal production, temperature gradients

are generally gradual, but during the blowing-in period, wide
fluctuations are possible. Care must therefore be exercised at this
period to see that the various zones are brought up to the working
temperatures as evenly and as gradually as possible.

18
0

The lining here should have a very good abrasion resistance and
resistance to carbon monoxide attack; refractoriness is relatively of
less significance. A good dense refractory is ideal for this purpose.

The bricks themselves should be true to shape and size so as to

reduce joints between bricks to a minimum thickness.

It is a common practice to use armour plates at the throat to

withstand abrading action of falling burden. Immediately below this,
over a length of nearly 2-3 m, high-fired, super duty firebricks are
used.

..

18
1

The entire stack below the top few metres of height is lined with high
duty firebricks. A 35-40% Al203 firebrick with a close texture is
usually preferred for the stack,

A 60% Al203 brick have been recommended for the lower parts of the
stack.

These bricks are made by machine moulding under high pressures

and de-aired conditions since these lead to a high bulk density and
low permeability.

In order to reduce the number of joints to a minimum, large specially

shaped blocks have been employed.

18
2

In the bosh the brickwork has to withstand:

High temperature conditions;

Erosion by the blast;

Lime and alkali slag attacks.

Considering the severity of temperature and chemical attack in this region the
lining should possess good refractoriness, refractoriness under load, low
after expansion and resistance to action of molten limy and alkali slags.
The majority of bosh linings are of high duty or super duty firebricks with 4565% Al203, laid in the conventional banded bosh construction with copper
cooling plates.

18
3

The extremely successful use of carbon blocks for lining the hearth and its
walls led to its adoption even in the bosh region since carbon refractory
possesses better properties, especially high thermal conductivity, than those
of the conventional high duty firebricks.

Carbon lined walls can be cooled by either spray coolers, or water jackets.
The changeover from firebrick lining to carbon lining, therefore, eliminates
the corrugated pattern of construction produced by rows of coolers inserted
in the lining and permits simpler construction which also results in smoother
and uniform wear.

The development and use of graphite-silicon carbide brick in Japan has

given excellent performance both for bosh and hearth and might find wider
use in near future.
18
4

The lining in hearth should primarily prevent breakouts. The use of

bricks of high alumina to silica ratio, lower permeability and porosity
with well laid joints can minimise breakouts. In spite of this the
earlier firebrick hearth still suffered from frequent breakouts and the
attendant troubles. The occurrence of breakout was believed to be
due to the oxidation of iron by gases that penetrated firebricks and
the consequent effect of iron oxide in lowering the melting point of
the refractory, resulting ultimately in its failure. Ramming of carbon
plus tar mixture at such vulnerable areas exhibited excellent
resistance to such breakouts.

18
5

This finally lead to the development and use of carbon lined

hearths. High refractoriness, high thermal conductivity, high
abrasion resistance, high bulk density coupled with low
porosity, good crushing strength, almost complete inertness to
carbon saturated iron and slag and such other properties make
carbon as an almost ideal material for hearth construction
provided bricks or blocks of these are keyed into position with the
thinnest possible joints. It has been observed that not only the
carbon hearth contour is better maintained during the campaign but
the problem of breakouts is virtually eliminated.

18
6

In the early adaptation carbon blocks were used as only the facing
lining with high duty firebrick for backing. However all carbon hearth
i.e. the whole wall thickness and a considerable bottom thickness,
has almost universally been accepted as a standard method of
preparing the hearth: the remaining bottom thickness is made up by
high duty firebricks. The shape and size of the carbon blocks used
for making the bottom vary considerably but all aim at achieving
keyed joint with thinnest possible joints, preferably without the use
of external jointing material.

18
7

In the modern all carbon hearth construction use of large blocks of

few square meter in size, with length of approximately half the
hearth diameter are increasingly being adopted. These are laid
horizontally and keyed together. Each block is anchored firmly at the
hearth wall to prevent it from floating by molten iron.

Careful control of the manufacture of hearth blocks and its con

struction is of fundamental importance in achieving the desired
campaign life of the hearth.

18
8

18
9

The process of starting a newly lined furnace is called blowing-in. In

general the operation involves four main steps, viz. drying, filling, lighting
and operation until normal production is established. There is no standard
practice of blowing-in and the details of each of the above mentioned
steps adopted depend on local conditions and customs.

Drying : The new lining of a furnace contains a significant proportion of

moisture which must be slowly and completely removed before the
temperature of the furnace is raised. This operation is known as drying in
which the furnace is slowly heated. Any amount of time and trouble taken
in ensuring careful and gradual drying of the furnace is more than repaid
in its subsequent operation.

19
0

At the end of drying, depending upon the method used for drying,
the furnace is cleared off all the things used for filling. The coolers
are turned on and once the inside temperature is tolerable furnace
personnel can get in and prepare for filling the furnace. The
inspection of coolers at this stage is a must since rectification of
faulty coolers is readily possible at this stage. In fact a check list is
prepared and each item is checked off as reports are received of
their satisfactory performance.. The tap holes are prepared and
coolers, which were earlier removed to have access inside, are
packed in position.

19
1

Filling of the furnace usually means filling the hearth with light kindling wood
and shavings saturated with oil up to the tuyere level and laying over this a
scaffold of old timber slippers. Coke is charged above the timber scaffold
from the top upto the bosh level. A quantity of limestone sufficient to flux the
ash in the charged coke, is also charged along with the coke after the initial
coke blanks. A small amount of old blast furnace slag is also incorporated
with coke after the coke level rises beyond the mantle level. The early slag
volume is deliberately maintained at high level to heat up the hearth and
prepare it to receive iron. On the coke blanks are laid light burden charges
of ore, stone and coke i.e. the ratio of iron ore to coke is low, about 05-06.

19
2

After filling the furnace as mentioned above for blowing-in the bells are opened and the
dust catcher dump valve is closed..

The furnace is lighted either by inserting red-hot bars through the tuyeres or slag hole and
iron notch. Alternatively gas torch may also be used. Generally highly combustible material
is kept in front of the tuyeres during filling to light the furnace readily. Burning is allowed
with natural draught alone for the first 2436 hours; a light blast is put on only thereafter.

As soon as good amount of gas emerges from the furnace top the bells are closed and the
dust catcher dump valve is slowly opened to conduct exit gases through the gas cleaning
system.

The blast volume is fairly rapidly increased to normal volume of blowing. Usually more than
three fourth of the standard volume of blast is blown by the end of fourth day and full blast
is on by the end of a week after the furnace is ignited.

19
3

Tap holes are kept open for hot gases to escape out during the early period. Once
coke burning and slag formation starts furnace crew are vigilant in observing the
tap holes. The first indication of a sudden decrease in the out coming gas through
the tap hole is taken as an indication of beginning of slag accumulation in the
hearth and the tap hole is immediately closed thereafter. Nearly six to eight hours
may elapse after this before sufficient slag has accumulated to warrant flushing.

After nearly two days, as the ratio of iron ore to increase in the burden, that first
cast may be due. The amount of slag and metal flowing out of the furnace is
correlated to the charge schedule and proportion in order to assess the progress
of blowing-in operation. After the first cast is over charging and tapping schedules
are established and are strictly adhered to until routine production is established. It
generally takes nearly a week to establish normal routine practice so that metal
and slag of desired composition will be tapped out of furnace at the desired
intervals.
19
4

Smarajit Sarkar
Department of Metallurgical and Materials Engineering
NIT Rourkela

19
5

Burden distribution is one of the key operating

parameters influencing blast furnace
performance, particularly the productivity and
the coke rate.

The proper distribution of burden materials

improves bed permeability, wind acceptance,
and efficiency of gas utilisation.

19
6

In a typical Indian blast furnace equipped with a bellless (Paul Wurth) distribution system, the decrease
in coke rate that is due exclusively to burden
distribution was found to be 1012 kg/thm.

19
7

Design of the blast furnace

Angle and size of the big bell.

and its charging device

Additional mechanical

(effect of these factors is

device(s) used for obtaining

constant).

better distribution.

Speed of lowering of large

bell.

19
8

Inconsistency in
physical properties of
charge materials
(deficiencies caused by
this should be
eliminated by improving
quality of the burden.

Size range of the various

charge materials

Angle of repose of raw

materials and other
physical characteristics of
the charge.

Density of charge
materials.

19
9

 Level,

system and
sequence of
charging, programme
of revolving the
distributor (conditions
determining major
means of blast
furnace process
control from top).

Distribution

of charge
on the big bell
Height of the big bell
from the stock-line i.e.
charge level in the
furnace throat.
Order and proportion
of charging of various
raw materials.

20
0

20
1

 This is a unique design in which

large bell is replaced by a distributor
chute with 2 hoppers
A rotating chute is provided inside
the furnace top cone
Advantages:
Greater charge distribution
flexibility
more operational safety and
easy control over varying
charging particles
Less wearing parts: easy
maintenance
20
2

20
3

The advantages accruing from improved distribution

control can be summarised as follows:
Increased productivity, decreased coke rate, improved
furnace life .
Reduced refractory erosion
Improved wind acceptance and reduced hanging as well
as slips
Improved efficiency of gas utilisation and its indirect
reduction
Lower silicon content in hot metal and consistency in the
hot metal quality
Reduced tuyere losses and minimisation of scaffold
formation
Lower dust emission owing to uniform distribution of fines.

20
4

The density of three important raw materials viz. the ore, the
coke and the limestone are quite different.

The heaviest is iron ore with around 5-6 glcc, the lightest is
coke with density of around 15 glcc and the limestone is
intermediate with-a value of density around 30-35 glcc.

It means that the rolling tendency of coke particles is maxi

mum and that of the ore is minimum. Since the density values
cannot be altered, the sizes may be so chosen that their
differential rolling tendencies are offset to some extent.

20
5

The problem of very dense ores is serious

from the point of view of their sluggish
reduction rates rather than their tendency
towards segregation. Such ores are
therefore invariably crushed and sintered to
obtain more porous agglomerates before
charging these in the furnaces.

20
6

 When

a multi-particle material is allowed to

gently fall on a horizontal plane it tends to form

a conical heap. The base angle of this cone is
known as angle of repose of that material.

This angle depends upon the particle size, its

surface characteristics, moisture content,
shape, size distribution, etc.

20
7

 For

an iron ore of 10-30 mm size, with an

average mean size of 18 mm, the angle of

repose is around 33-35. For coke of 27-75 mm
size, with an average size of 45 mm, the same
is around 35-38. Similarly the angle of repose
for sinter is in the range of 31-34 and for pellets
it is around 26-28.

20
8

The higher is the angle of repose the more it has the tendency to
form ridges on charging in a blast furnace.

The more dried is the ore and the more it is free from fines the
less pronounced is the angle of repose and thus less is the
tendency towards segregation.

The clayey ores tend to form ridges because of their high angle
of repose. The effective way to reduce the angle of repose of any
iron ore is to eliminate the fines, dry the ore if wet and to wash
off clay, if any, adhering the ore.

20
9

21
0

21
1

21
2

 On

dumping, as the materials fall on the stock

surface, they take a parabolic path and mainly

two different profiles of the accumulated mass
emerge depending upon whether the particles
hit the in-wall directly(V- shape) or the stock
surface (M-shape)

21
3

The M-profile itself is generally obtained if the material

strikes the stock surface. This happens when the
bell/throat diameter ratio is small (larger bell-inwall
distance) or the charging distance is small . It is clear
that the peak of the M-contour approaches the inwall
(hence the peripheral permeability decreases) as the
charging distance increases and ultimately the M
changes to V profile.

21
4

21
5

Smarajit Sarkar
Department of Metallurgical and Materials Engineering
NIT Rourkela

21
6

21
7

Right at the top of the furnace is the granular zone that contains
the coke and the iron bearing materials charged, sometimes
along with small quantities of limestone and other fluxes. The
iron-bearing oxides charged get reduced to wustite and metallic
iron towards the lower end of the granular zone.

As the burden descends further, and its temperature rises on

account of contact with the ascending hot gases, softening and
melting of the iron-bearing solids takes place in the so-called
cohesive zone (mushy zone).

21
8

Out of five zones the cohesive zone plays the most

important role in the B.F. operations.
This is the zone where the ferrous burden soften and
melt.
Its shape, position and extent in the B.F. affect the gas
flow pattern.
The burden loses its permeability. Gas flow occurs only
through the coke layers.
Loss of permeability is caused by liquid phase in the
ferrous burden.
The liquid formation causes a pressure drop.

21
9

Also the solid phase gets deformed due to the weight of the
burden.
This deformed solid may occupy the gaps between the solid
ferrous pieces also causing loss of permeability.
The different phenomena simultaneously occuring in the
cohesive zone are
a. softening and melting of the oxide phase.
b. carburisation of the metalic phase
c. softening and melting of the metallic iron phases
Softening and melting of the oxide phase will be affected by :
a. the quantity of non ferrous oxide ( slag formers) present
b. distribution, morphology and chemistry of slag formers
c. degree of pre reduction (this will affect the availability of
FeO as a slag former).

22
0

Softening and melting of the metal phase will depend on :

carbon content of metal phase

cross sectional area of metal phase

These complex processes are simplified as under :

(I) Formation of the first oxide melt.
a. It is formed at the interface of lowest melting, usually

the

interface between an FeO particle and another oxide particle.

b. The formation will depend on the microstructure (phases
present and their distribution)
c. The first liquid tends to wet the ore particles.
d. This wetting resists transport through it and slows down the
reduction kinetics.
e. This is often called reduction retardation.

22
1

(II) Reduction degree of the metallic burden

a. Metallic burden in most B.F.s. reaches the softening zone
with reduction degree higher than 50 pct.
b. The situation in the cohesive zone is a porous solid iron
oxide shell confining solid and liquid oxide.
c. liquid slag covers the solid oxide particles, develops a semi
solid material, acts as a lubricant for other particles.
d. the semi solid core has a reduced mechanical strength
e. resistance to deformation is determined by the porous Iron
shell.
f. volume fraction of liquid slag increases with temperatures.
g. the Iron shell can no longer hold the liquid.
h. This results in the dripping of liquid slag from the burden
material.
i. Iron shell gets carburised, its melting point decreases,
mechanical strength also decreases.

22
2

softening is defined as the moment when the metallic burden can

no longer resist the action of mechanical forces.
this generally considers with dripping of material from the burden
component.
the softening and the dripping depend on
a. increase of the molten slag volume.
b. carbon content in the Iron shell
Both parameters increasing with increase in temperature.
The reduction degree will tend to remain constant during this
period due to reduction retardation.
Lesser the reduction degree lesser is the softening temperatures
(here the thickness of iron layer is low, lowering its mechanical
strength, assisting easy dripping).
Higher the reduction degree higher should be the softening
temperatures (lower is the FeO, the slag former).

22
3

Different configurations of the cohesive zone are as follows :

1. W-type, 2. Flat, 3. Inverted-U, and 4. Inverted-V

(inverted-V configuration is believed to result in the optimum
performance of the B.F.)

This inverted-V configuration is maintained by radial burden

distribution.

Cohesive zone should be formed lower down the bosh (results in

increase in the granular zone, increase in the gas utilization,
decrease in the volume of the dripping Zone, decrease in the
contact time between SiO gas and hence Si pick up by the metal.)
22
4

Lowering of cohesive zone can be achieved by a high

softening temperature of the burden material.

The thickness of the cohesive zone should be low (a

thinner cohesive zone is expected to allow more air
passage for a given pressure deferential).

22
5

Further down the furnace, impure liquid iron and liquid slag are
formed. The absorption of carbon lowers the melting point of iron
drastically. For example, an iron alloy containing 4 wt. % carbon
melts at only 1185C..

In the cohesive zone and below it, coke is the source of carbon for
carburisation of liquid iron. However, carbon directly does not
dissolve in liquid iron at this stage. The possible mechanism of
carburisation of iron entails the formation of CO by gasification of
carbon, followed by the absorption of carbon by the reaction:

2CO(g) = [C]in Fe+ CO2(g)

22
6

Coke is the only material of the blast furnace charge which descends to
the tuyere level in the solid state. It burns with air in front of the tuyeres
in a 1-2 m deep raceway around the hearth periphery.

Beyond the raceway there is a closely packed bed of coke, the central
coke column or dead man's zone.

The continuous consumption of coke and the consequent creation of an

empty space permit the downward flow of the charge materials.

The combustion zone is in the form of a pear shape, called 'raceway' in

which the hot gases rotate at high speeds carrying a small amount of
burning coke in suspension.

22
7

The raceway is a vital part of the blast furnace since it is the heat
source in a gigantic reactor and at the same time a source of
reducing gas.
The salient features of Combustion zone are summarized below:

The force of the blast forms a cavity the roof of which is formed of
loosely packed or suspended coke lumps and the wall more closely
packed.

The CO2 concentration tends to increase gradually from the centre

and reaches a maximum value just before the raceway boundary
where most of the combustion of coke occurs according to:
C+O2 (air) =CO2+94450 cal

22
8

The temperature of the gas rises as the coke consumption

proceeds and reaches a maximum just before the raceway
boundary. Thereafter, it falls sharply as the endothermal
reduction of CO2 by C proceeds;
CO2 +C =2CO-41000 cal

The concentration of CO2 fall; rapidly from the raceway

boundary and the gasification is completed within 200-400
mm from the starting point of the reaction.

22
9

23
0

The primary slag of relatively low melting point which forms in the lower part
of the stack or in the belly consists of FeO-containing silicate and
aluminates with varying amounts of lime which has become incorporated
depending upon the degree of calcination undergone .
As the slag descends, ferrous oxide is rapidly reduced by carbon as well as

by CO. As the lime is continually absorbed, the original FeO-Si02-AI203

system rapidly changes to the CaO-Si02-AI203system with some minor
impurities accompanying the burden. The dissolution of lime and the
approach to the CaO-Si02-Al203 system is more pronounced,

.
23
1

As the liquid primary slag runs down the bosh and loses its fluxing
constituent FeO, the liquidus temperature also increases. If, therefore,
the slag has to remain liquid it must move down to hotter parts of the
furnace as rapidly as its melting point is raised. As the reduction of FeO
is almost complete above the tuyeres the resulting bosh slag, composed
mainly of CaO-Si02-AI203

The hearth slag is formed on dissolution of the lime which was not
incorporated in the bosh and on absorption of the coke ash released
during combustion. The formation is more or less complete in the
combustion zone.

23
2

This slag runs along with the molten iron into the hearth
and accumulates there and forms a pool with the molten
metal underneath. During the passage of iron droplets
through the slag layer, the slag reacts with the metal and
a transference of mainly Si, Mn and S occurs from or to
the metal, tending to attain equilibrium between
themselves as far as possible.

23
3

23
4

 0.81

kg. C is required for indirect reduction

of 1 kg. Fe from Fe203 and about 1790 kcal

of heat is evolved in the process.

for direct reduction of 1 kg. Fe, only 0.23

kg. C is consumed but results in an
absorption of 656 kcal of heat.

23
5

23
6

23
7

23
8

23
9

The different phases of iron oxides in equilibrium with H2/H2O mixtures at various
temperatures are shown in previous fig. by the dotted lines. The reduction of
FeO by H2 is endothermic and therefore the curve inclines downwards with
increasing temperature whereas the corresponding curve for reduction by CO
(full line) inclines upwards because of the exothermic nature of the reaction.
These two curves intersect at about 821C, i.e., at 821C hydrogen and CO
have the same reducing power over FeO. Thermodynamically, below this
temperature, the reducing power of CO is much greater.

24
0

The curve for Eq. (iii) slopes upwards because the reaction is
exothermic, i.e., FeO becomes more stable with increasing
temperature in the presence of CO.

The curve for Eq. (ii) slopes downwards the reaction being
endothermic.

At 900C, the equilibrium concentration of CO for Fe304-FeO

is 20 percent whereas that for FeO-Fe about 70 percent, i.e.,
for the conversion of magnetite entirely to wustite the gaseous
phase must have a CO/ COz ratio greater than 0.25
whereas. that for wustite to iron the ratio should be higher
than 2.3.
24
1

24
2

24
3

Below 600C :
Pre-heating and pre-reduction

600 -950C:
Indirect reduction of iron oxides by CO and H2

9500C to softening temperature:

Direct reduction; gasification of carbon (solution loss
reactions) by CO2 and H2 becomes prominent.

24
4

The formation of cohesive layers or partially

reduced and partially molten iron oxide takes
place.
The coke slits provide passage for gaseous flow.

Dripping or Dropping Zone

Semi fluidized region in which liquids drip and
fragments of cohesive layers drop.
Zone through which liquids trickle down to the
hearth. It is the final stage of iron oxide reduction

24
5

Blast, injectants and coke are converted to hot reducing gas. This
gas reduces the ore as it moves counter currently towards the top of
the furnace.

Hearth

It is a container for liquids and coke where slag/metal! coke/gas

reactions take place.

Metal droplets pass through the slag/coke

layer. Liquid metal/coke layer in which chemical reactions take

place only to a small extent.

24
6

24
7

24
8

fluidization of small particles when the local gas

velocity is excessive;

diminution of void age due to swelling and

softening-melting;

flooding of slag in the bosh zone when the slag

volume and gas velocity are excessive.

24
9

The charge in the blast furnace descends under gravity against the
frictional forces of solids and buoyancy of gas. With increasing gas
velocity, the pressure drop increases approximately quadratically
until the upward thrust of the gas and downward thrust of the solids
are held in balance.

When this critical velocity is exceeded (the point of incipient

fluidization), the packing in the bed becomes loose, the finer
particles begin to teeter and the pressure drop ceases to increase,
i.e., the resistance to gas flow drops (due to increase in void age at
places where the fines become suspended).

25
0

25
1

The mechanism of the softening-melting phenomena

is schematically illustrated in previous Figure. It is
evident that with the onset of softening, the voidage in
the bed decreases and the bed becomes more
compact (origin of the terminology cohesive).
As a consequence, further indirect reduction of iron
oxide by gases becomes increasingly difficult. Upon
melting, dripping of molten FeO-containing slag
through the coke layers increases the flow resistance
through the coke slits and the active (i.e. dripping)
coke zone because of loss of permeability.

25
2

The cohesive zone has the lowest permeability. Hence,

for proper gas flow:
Ts

should be as high as possible

The

thickness of the cohesive zone should be as small

as possible. This thickness depends on the difference

between Ts and T m (Tm - Ts), and therefore, the
difference should be as low as possible.

25
3

25
4

Gas flow through Granular zone:

For resistance to gas flow, more important
than the particle diameter is the relative size of
the materials in the bed.
In a mixed bed of widely varying particle size,
the small particles land in the interstices of the
large ones and decrease the void age .
Starting with large uniform spheres, the void
age decreases as the small ones are introduced
and the bed becomes more and more compact
as the proportion of the latter increases.
The bed is most dense, i.e., the voidage is
minimum when 60-70 percent of the total
volume of the particles consists of the large
ones for about all the cases.
25
5

The m increases on either side of the

minimum, i.e., with increasing or
decreasing volume fraction of the small
particles (approaching more uniformity
of the size distribution).
The voidage decreases greatly as the
ratio ds/ d1 decreases.
This shows that for a good and
uniform permeability and low resistance
to gas flow in a mixed bed, the size
fractions should be as narrow as
possible.
One can easily visualize the adverse
effects of multi-granular bed of particles
of varying diameter on the voidage.
25
6

A narrow size distribution has the following advantages:

charge permeability increases and the gas distribution is
more uniform with better utilization of the chemical and
thermal energies of the gases;
more even material distribution at the stock level and less
material segregation in the shaft during descent;
gas flow is not impeded if the size ratio is within limits but
at the same time gives rise to a tortuous flow of gases with
continuous changing of flow directions, providing a larger
gas/solid contact time.
25
7

The fraction of iron bearing material below the limiting size

is therefore termed as 'fines' by the blast furnace technologists
and is invariably eliminated by screening at every possible
stage.
From the point of view of reduction the maximum top size of
an iron bearing material should be as low as possible, since the
rate of reduction decreases, perhaps exponentially, with
increasing size.
The size range of materials charged in the blast furnace
represents a compromise to give both good stack permeability
and adequate bulk reducibility.
25
8

Gas flow in wet zone:

Wet zones consist of the coke beds in the
bosh and belly regions, i.e. inactive coke zone,
active coke zone, and the coke slits in the
cohesive zone.
Here molten iron and molten slag flow
downwards through the bed of coke. This
reduces the free cross section available for gas
flow, thus offering greater resistance, thereby
increasing the pressure drop.
An extreme situation arises when, at high gas
velocity, the gas prevents the downward flow
of liquid. This is known as loading. With further
increase in gas velocity, the liquid gets carried
upwards mechanically, causing flooding.
25
9

Scientists have tried to estimate pressure

drop in blast furnace. However, they are
approximate. Moreover, they are only for the
granular zone and coke zones.

The situation in the cohesive zone is very

complex, and reliable theoretical estimates
are extremely difficult to come by.
26
0

Therefore, for practical applications in blast

furnaces, an empirical parameter, called Flow
Resistance Coefficient (FRC) has become
popular. The FRC for a bed is given as

where

the gas flow rate is for unit cross section

of the bed, i.e. either mass flow velocity or

volumetric flow velocity .

26
1

FRC=1/ bed permeability

The FRC for a furnace can be empirically determined

from measurements of pressure drop and gas flow rate.
Since it is possible to measure pressures at various
heights within a furnace, the values of FRC for individual
zones can also be determined.

26
2

These measurements have indicated that

FRCs for the granular, cohesive, coke +
tuyere zones are approximately 20%, 50%
and 30% of the overall furnace FRC.

This means that the cohesive zone is

responsible for the maximum flow resistance
and pressure drop, to a very large extent.

26
3

Smarajit Sarkar
Department of Metallurgical and Materials Engineering
NIT Rourkela

26
4

26
5

 Decreasing
o
o
o

o
o

the extent of SiO formation by:

Lowering ash in coke, and the coke rate

Lowering RAFT
Lowering the activity of Si02 in coke ash by lime
injection through the tuyeres.

Decreasing Si absorption by liquid iron in the bosh

by enhancing the absorption of Si02 by the bosh
slag. This can be achieved by:
Increasing the bosh slag basicity.
Lowering the bosh slag viscosity..
26
6

o
o

Removal of Si from metal by slag-metal reaction

at the hearth by:
Lowering the hearth temperature
Producing a slag of optimum basicity and fluidity.

26
7

Desulphurisation of metal droplets through slagmetal reaction in the furnace hearth :

(CaO) + [S] + [C ]= (CaS) + CO (g)

Desulphurisation through the coupled reaction:

(CaO) +[S] +[ Mn] = (CaS) + (MnO)
(CaO) + [S] + [ Si] = (CaS) + 1/2 (SiO2)

26
8

Sulphur pick-up through the vapour-phase

reaction:
CaS( in coke ash) + SiO (g) = SiS(g) + CaO
In the bosh and belly regions, SiS decomposes
as
SiS(g) = [Si] + [S]
o Decreasing the vaporisation of sulphur in the
race way
o Preferential absorption of SiS by the bosh slag

26
9

Reducing slag i.e. FeO content should be low

High basicity

High temperature, since desulphurisation is an

endothermic reaction

Kinetic factor
Contact surface of metal and slag ( by agitation)
Fluidity of slag( by adding MgO , MnO)

Time of desulphurisation

27
0

27
1

27
2

27
3

27
4

The efficiency of operation of a blast furnace may be

measured in terms of coke rate which should of course
be as low as possible. The achievement of a satisfactory
coke rate depends on optimising the extent to which the
carbon deposition reaction proceeds. If the top gas is
high in C02 sensible heat is carried from the furnace as a
result of the exothermic reaction.
2CO=CO2+C
If on the other hand the top gas is high in CO, chemical
heat leaves the furnace.

27
5

27
6

27
7

The combustion of carbon to its monoxide yields only about 28

percent of the total heat possible to be obtained and only oxidation to
the dioxide gives the largest amount of heat.

This has a great significance in the blast furnace process where about
30-40 percent of the coke carbon is oxidised to CO2 and the rest into
CO.

In the blast furnace about half of the heating value of carbon is

obtained for a top gas CO/CO2 ratio of 2. The rest is carried away in the
outgoing gas.

Although a part of this heat is redeemed by burning the gas out of the
furnace, it means conversion of expensive metallurgical coke to a gas
which could be obtained from a less valuable fuel.
27
8

The heat values in the given equations are for pure carbon. For
a good quality coke with 86% C, the respective values will be
about 2000 and 7000 kcal/kg.coke.

Carbon dioxide, in the absence of carbon, is stable at high

temperatures.

Up to about 2000C, its thermal dissociation is small. At

atmospheric pressure, it amounts to about 2 percent.

At high temperatures, carbon dioxide is unstable in the

presence of carbon and is reduced to carbon monoxide
according to the given equation.

27
9

This is generally known as Boudouard or carbon gasification

reaction and is highly endothermic; the reverse reaction is
exothermic and is called carbon deposition reaction. The
former reaction is of extreme importance for the economy and
smooth running of the blast furnace. It occurs at high
temperatures, above about 10000C, and therefore consumes an
enormous amount of heat in the lower third of the blast furnace.
The carbon deposition reaction is most pronounced at
temperatures 500-550C and consequently takes place in the
upper portion of the shaft.

28
0

28
1

Since 2 moles of CO are produced from 1 mole of CO 2

(carbon being solid), the Boudouard reaction is
dependent on pressure.

The isobars in the previous figure show that at any given

temperature the equilibrium %C02 increases with
increasing pressure.

The stability of CO decreases with decreasing

temperature and increasing pressure. The maximum
instability is between 600 to 800C as shown by the steep
slope of the curves.

28
2

The reaction velocity becomes so slow below about

400C that a negligible amount of CO is decomposed
into CO2 and C.

The reaction is accelerated in the presence of catalysts

which provide nucleation sites for deposition of carbon.

In the presence of catalysts, the reverse of reaction

accelerates above 400C and reaches a maximum
between 500-5500C.

28
3

 CO2

emission

Industry
Power
Transport
Steel
other

Contribution %
51
16
10
23

28
4

28
5

The purpose of HTP is to introduce more

oxygen to burn more carbon by blowing more
air and at the same time maintaining the
linear gas velocity (and pressure drop)
identical to that in the conventional practice
without any formation of channels,
maldistribution of gas, increase in coke rate
or flue dust emission

Advantages:
For the same volume flow rate, a greater mass of air
(hence, oxygen) can be blown with HTP; higher output;

28
6

A major benefit that is so obvious is increased

production rate because of increased time of contact of
gas and solid as a result of reduced velocity of gases
through the furnace. Increased pressure also increases
the reduction rate of oxide;

Suppression of Boudouard reaction (C02 + C= 2CO) and

hence savings in fuel;

More uniform distribution of gas velocity and reduction

across furnace cross-section; smoother furnace
operation due to increased permeability;
28
7

less flue dust losses, less variation of coke input, better

maintenance of the thermal state of the hearth, more
uniform iron analysis;

More uniform operation with lower and more consistent

hot metal silicon content have been claimed to be the
benefit of high top pressure;

Bhilai Steel Plant (operative), RSP yet to implement

28
8

SiO2 +C ={SiO} +{CO}

From above equation it can be seen that partial
pressure of SiO can be brought down by increasing
the partial pressure of CO; in other words the SiO2
reduction reaction can be discouraged by application
of top pressure which enables a higher blast pressure
and hence an increase in partial pressure of CO.

28
9

29
0

The blast volume and therefore the coke

throughput can be increased by 30
percent with the maintenance of identical
pressure drop and gas velocity conditions
in the blast furnace by increasing the top
pressure to 2.1 from 1.1 ata and bottom
pressure to 3.5 from 2.5 ata under the
given blowing conditions.

29
1

'raceway adiabatic flame temperature

This is the highest temperature available inside the
furnace. There is temperature gradient in vertical
direction on either side of this zone. This temperature is
critically related to the hearth temperature known as
operating temperature of the furnace. It is equally
related to the top gas temperature such that the hot
raceway gasses have to impart their heat to the
descending burden to the extent expected and leave
the furnace as off-gases at the desired temperature.
29
2

29
3

29
4

The primary purpose of using injectants with the

blast is profitability which depends upon the
relative price of coke and injectants and the
amount of coke that can be saved per unit of the
latter, i.e., upon the replacement ratio:

29
5

29
6

H20 + C = CO + H2 (1)
HO (1200C) = + 2700 kcal/kg C

Presence of moisture in the blast generates double the

volume of reducing gas per mole of carbon burnt. As per
Eq.1 for every carbon burnt one mole of CO and an
additional mole of hydrogen will be available as product
of burning of coke for reduction in bosh and stack.

29
7

The more the moisture the more will be this additional

hydrogen available.

Kinetically hydrogen reduction of iron oxide is faster than

that by CO because of its small size. Presence of
moisture helps to burn coke at a faster rate with its
attendant favorable effects.

Some of the endothermic heat of moisture disintegration

is compensated by way of exothermic reduction of iron
oxide by hydrogen.

29
8

higher gasifying power which intensifies coke

consumption In the raceway;

smoothens the temperature gradient and facilitates stock

descent ;

enlarges the combustion zone and accelerates stock

descent; heats up the axial zone; maintains thermal
state of the hearth;

29
9

even with incomplete temperature compensation, the coke

rate may not rise because of higher reducing power and
higher heat transfer coefficient of hydrogen;

decreases pressure loss due to lower density and viscosity of

hydrogen.

The blast pressure may drop even by 0.1-0.2 atm. which

means the furnace can be blown at a higher blast rate.

30
0

It has been estimated that for an increase of 20 g/Nm3

moisture in the blast the endothermicity can be
compensated by a rise of 200C in the blast preheat.

By increasing moisture and compensating it by

additional rise of preheat means that cheaper heat
energy can be used to feed the furnace and thereby
decrease the coke consumption and economise the
operation.

30
1

Oxygen enrichment of the blast and moisture enrichment

have quite opposite thermal effects. The two can be saddled
together to obtain better inputs.

Hot blast temperature, extent of oxygen enrichment and

humidification of blast have to be adjusted as interrelated
parameters simultaneously to obtain optimum conditions of
operation for maximum benefits such as minimum coke rate,
higher productivity and so on.

30
2

The reasons for the injection of coal have been

economic as well as operational flexibility and include
the following:
After the steep rise in oil prices following the oil crisis,
iron makers were compelled to abandon heavy oil
injection and were looking for a less expensive auxiliary
fuel.
PCI accommodate shortages of coking capacity, by
replacing coke by coal in the blast furnaces. After a
thorough investment analysis, it has been found that a
reliable coal injection system requires much lower
capital cost and involves operating cost than the
extension of coking capacity.
.
30
3

Coal causes a lower reduction in flame

temperature per unit injection than oil or natural
gas. It, therefore, allows more scope for blast
temperature adjustment/oxygen enrichment for
increased rates of injection and consequently, less
coke consumption.
The PCI system design is capable of
injecting coal on a continuous and stable basis and
ensure accurate and uniform distribution

30
4

The coke savings from fluxed burden emanate from the following
causes :

better reducibility and enhanced indirect reduction (6-7 kg.C saved from
every 1 percent increase in indirect reduction);

use of higher blast temperatures because the thermal load is smaller

and the slag is pre-made; the primary slag melts at higher temperatures
and does so within a vertically narrow softening zone;

avoidance of carbon dioxide generated from limestone in the stack

which adversely affects indirect reduction;

transference of heat of calcination from the furnace to the

agglomerating plant.

30
5

 This is a unique design in which

large bell is replaced by a distributor
chute with 2 hoppers
A rotating chute is provided inside
the furnace top cone
Advantages:
Greater charge distribution
flexibility
more operational safety and
easy control over varying
charging particles
Less wearing parts: easy
maintenance
30
6

30
7

The advantages accruing from improved distribution

control can be summarised as follows

Increased productivity, decreased coke rate, improved

furnace life

Reduced refractory erosion

Improved wind acceptance and reduced hanging as well

as slips

Improved efficiency of gas utilisation and its indirect

reduction

30
8

Lower silicon content in hot metal and consistency in the

hot metal quality

Reduced tuyere losses and minimisation of scaffold

formation

Lower dust emission owing to uniform distribution of

fines.
All these advantages have improved the overall
efficiency, thereby making the process more competitive.

30
9

Coal-based Processes
Coal-based Rotary Kiln Processes like SL/RN
Codir, Accar, TDR, DRC
Coal-based Processes Using Rotary Hearth Furnaces
like FASTMET, COMET, INMETCO
Gas-based Processes
Gas-based Reduction in Stationary Retorts HYLProcess
Gas-based Shaft Furnace Processes: Midrex
Gas-based Direct Reduction in Fluidised Beds: Finmet

31
0

31
1

Non-metallurgical cheaper coals are readily available and

processes have been developed wherein carbon of such coals
is used to provide heat as well as the reducing gas.

Iron oxide in contact with carbon when heated results in

reduction of oxide to metallic iron. This reduction is more due to
the burning of carbon first followed by reduction by the carbon
monoxide gas generated from burning.

31
2

This is the reason why these processes are termed as 'coal-based'

processes. These are more popularly carried out in horizontal rotary
kilns. But vertical shaft furnaces have also been used as reactors
for this process in some rare cases.

This is also a counter current process to the extent that solid oxide
travels downwards and the gases formed on burning of coal ascend
with respect to the inclination of the kiln or the shaft reactor.

A long kiln, easily 70-80 m in length and few meters in diameter,

slightly inclined to the horizontal and rotating slowly around its own
axis is employed as a reactor in these processes.

31
3

The individual designs do differ in details. The charge is fed from

that end which is at a higher level. The charge travels under gravity
aided by the rotation motion, through several heating zones and the
reduced iron oxide product comes out of the other end of the kiln.

The throughput rate of solid iron oxide, its reducibility, its size,
gangue contents, the rotation of kiln and other related factors are
adjusted such that the oxide is reduced to the extent of 93-95% by
the time it travels the whole length of the kiln. This reduction is slow
and continuous and takes place over the length of the kiln. In this
process the iron oxide gets heated to a maximum of 1050C.

31
4

The gas-based shaft furnace processes, which have become by far the
most popular for the production of sponge iron, employ a vertical shaft
furnace in which, as in the case of a blast furnace, lump ore and pellets
are charged at the top using a charging system similar to a blast furnace.
Reformed natural gas after pre-heating is introduced in the lower portion
of the shaft. As the hot reducing gas flows upwards, reduction takes
place continuously. Hence, these processes are often referred to as
continuous countercurrent moving bed processes. In this category, the
Midrex process is dominant, followed by HYL III and HYL IV.

31
5

This process was developed by Midland Ross Corporation of

Cleveland, USA in 1967. The reducing gas is, as usual, generated by
reforming natural gas. In the Midrex reforming system, a proprietary
nickel catalyst is used. A single reformer is utilised instead of a
reformer/heater combination

The iron oxide feed to a Midrex shaft can be in the form of pellets or
lump ore However, generally speaking, the charge consists of around
60% pellets and 40% lump ore of a particular type. Pellets are the
preferred feedstock owing to their superior physicochemical
characteristics compared with lump ores.,

31
6

31
7

The spent reducing gas (or top gas) leaving the shaft furnace at a
temperature of 400-450C is cooled and cleaned in a gas scrubber
before approximately 60% of the gas is returned to the reformer; and
the rest used as a fuel. The process gas is compressed and preheated before entering the reformer at around 900C, where it is
mixed with make-up natural gas. The reformed gas made up mostly of
carbon monoxide and hydrogen exits from the reformer at about
850C. The ratio of H2 to CO is controlled at about 1.5-1.8, so that
the reducing quality of the gas is maintained for best operations. The
iron oxide burden is first heated to the process temperature before it is
metallised by the upward flowing, countercurrent reducing gas
injected at 800-870C through tuyeres located at the bottom of the
cylindrical section of the shaft.

31
8

The reduced material then passes through a transition zone before

reaching the lower conical section of the furnace. Low carbon DR1 (less
than 1.5% C) is directly cooled using a circulating stream of cooled
exhaust gas introduced in the conical section, before cold DRl is
discharged. When higher carbon DRl (up to 4.0% C) has to be produced,
natural gas is introduced along with cooling gas into the conical section.
In this zone, natural gas readily decomposes (cracks) in the presence of
highly reactive metallic DRI, thereby generating nascent carbon, which
gets absorbed in the product before it is discharged. In both cases, the
final product is DRl with 93-94% metallisation, with the desired
carbon content. A large number of Midrex plants are operating
successfully all over the world.

31
9

32
0

Until recently, blast furnaces were the only source of hot

metal on a bulk scale. However, with the world
witnessing a gradual shift from integrated steel plants
using the Blast Fumace-BOF combination to a
multiplicity of smaller mini-mills essentially based on
electric arc furnaces, alternative means of producing hot
metal using Smelting Reduction (SR) have come into
existence.
32
1

32
2

As the name implies, smelting reduction involves both reduction and

smelting, i.e. melting accompanied by chemical reaction(s).
The unit operations that take place in any smelting reduction process
are summarised in next figure.
In an ideal SR reactor, in the strictest sense, all the reduction
reactions should take place together in the liquid state in a single-step.
In actual practice, for effective process control, most SR processes
utilise two reactors and at least two, if not three, process steps that
include: the removal of oxygen from the oxide in the solid-state to
varying extents in stage one, followed by the removal of the remaining
oxygen via liquid phase reduction reactions in stage two. Sometimes,
the latter is completed in two steps rather than one, i.e. SR then
becomes a three-stage operation.
The initial reduction of iron oxide begins in the temperature range of
850C to 1050C. The basic reactions involved are as follows:

32
3

32
4

32
5

Up to this stage, SR is similar to solid-state DR. Beyond solid-state

reduction of haematite/ magnetite to wustite, smelting reduction,
essentially involves reduction of molten FeO by CO.

This gives rise to far higher transport rates owing to convection, and
a remarkable increase in the conversion rate because of
enlargement in the specific phase contact area. The latter is a direct
consequence of the. dispersed nature of the phases. These two
major advantages of SR accrue because of the formation of liquid
phases, which does not happen in DR.

32
6

From studies on the oxidation and reduction behaviour of pure molten

iron oxide by CO/ CO2 mixtures at l500C, it has been concluded that the
rate controlling step for both oxidation and reduction is the inter-diffusion
of iron and oxygen atoms within the melt. It has also been found that
mass transport plays an important role in the reduction kinetics. The
overall reaction rate is proportional to the square root of the gas flow rate.
Therefore, all out efforts are made to increase the amount of gas
that is available for reduction.

To generate sufficient amount of reducing gas, all SR processes

consume fairly large quantities of reductant (normally coal). Having
generated the large volume of gas that is required, its effective utilisation
becomes extremely important.

32
7

However, this gives rise to one of the inherent deficiencies of SR,

i.e. with most SR reactor configurations given the productivity
requirements, the entire gas cannot be fully utilised in the process,
and rich gas at a high temperature leaves the SR reactor.
It is for this reason that the use of the export gas in any SR process
has a marked influence on the cost of the hot metal made. In fact, in
many cases, unless the net export gas from the SR reactor is
gainfully utilised, ironmaking itself becomes totally uneconomical.
Without sufficient credit for the off-gas, the cost of hot metal made
by smelting reduction can be as much as 40-50% higher than that
of blast furnace hot metal, despite starting with less expensive
and inferior grade raw materials.
The reverse is the case if adequate credit can be obtained from the
utilisation of the off-gases. The most convenient way to utilise the
exit gas is to cogenerate electrical power. Alternatively, the exit gas
can be fed to a shaft furnace direct reduction unit, located adjacent
to the smelting reduction reactor.

32
8

This essentially eliminates the cokemaking step and uses directly

coal as the fuel as well as the reducing agent. This process
produces molten iron in a two-step reduction melting operation.
One reactor is melter-gasifier and the other is pre-reducer. In the
pre-reducer, iron oxide is reduced in counter-flow principle. The hot
sponge is discharged by screw conveyors into the melting reactor.
Coal is introduced in the melting-gassifying zone along with
oxygen gas at the rate of 500-600 Nm3/thm. The flow velocity is
chosen such that temperature in the range of 1500-1800C is main
tained.
32
9

The reducing gas containing nearly 85% CO is hot dedusted and cooled
to 800-900C before leading it into the pre-reducer. The process is
designed to operate at up to 5 bars.

Fluxes like limestone, dolomite, silica sand, etc. are added along with the
ore to finally make up the right quality of slag.

Around 500-700 Kg of fixed carbon is required to make a tonne of hot

metal. Along with 300 Nm3/thm oxygen gas.

The export gas generated in the melting zone is rich and may have
calorific value of around 7500 kJ/Nm3 and can be used in the plant
usefully

33
0

33
1

33
2

In the FINEX Process fine ore is preheated and reduced to DRI in a

train of four or three stage fluidized bed reactors.

The fine DRI is compacted and then charged in the form of Hot
Compacted Iron (HCI) into the melter gasifier. So, before charging
to the melter- gasifier unit of the FINEX unit, this material is
compacted in a hot briquetting press to give hot compacted iron
(HCI)

since the melter- gasifier can not use fine material (to ensure
permeability in the bed).

Non-coking coal is briquetted and is fed to the melter gasifier where

it is gasified with oxygen
33
3

By
Dr. Smarajit Sarkar
Associate Professor
Dept. of Metallurgical and Materials
Engg.
National institute of Technology,
Rourkela

33
4

Introduction

Changing Pattern of Steel Making

Modern steel making BOF / LD steel making

Silicon Reaction

Manganese reaction

Phosphorous Reaction

Carbon Reaction

Vacuum Degassing

33
5

Steelmaking is conversion of pig iron containing about

10 wt weight of carbon , silicon, manganese,
phosphorus, sulphur etc to steel with a controlled
amount of impurities to the extent of about 1 weight
percent.

With the exception of sulphur removal of all other

impurities is favored under oxidizing conditions.

In the case of oxidation of carbon the product, being a

gas, passes off into the atmosphere but rest of the oxide
products shall remain in contact with the iron melt in the
form of a slag phase.
33
6

SiO2, MnO and P2O5 formed by oxidation of Si, Mn and P,

respectively will join the slag phase.

The formation of these oxides can be facilitated by

decreasing their activities which is possible by providing
oxides of opposite chemical character serving as flux.

As SiO2 and P2O5 are acid oxides a basic flux is required

for formation and easy removal of the slag.

A strong basic slag is formed by addition of CaO and / or

MgO to absorb P2O5 and SiO2.

The removal of carbon will take place in the form of

gaseous products (CO).
33
7

33
8

33
9

34
0

During refining, controlled oxidation of the impurities in hot metal

(with the exception of sulphur) takes place once oxygen is blown at
supersonic speeds onto the liquid bath.

The interaction of the oxygen jet(s) with the bath produces crater(s)
on the surface, from the outer lip(s) of which, a large number of tiny
metal droplets get splashed.

These droplets reside for a short time in the slag above the bath.
Therefore, the existence of a metal-slag-gas emulsion within the
vessel, virtually during the entire blowing/refining period is an integral
part of BOF steelmaking.
34
1

This is the reason why the slag-metal reactions like

dephosphorisation and gas-metal reactions like decarburisation
proceed so rapidly in the BOF process

The droplets ultimately return to the metal bath. The extent of

emulsification varies at different stages of the blowing period, as
depicted schematically .

A minimum amount of slag, with the desired characteristics, is

necessary for ensuring that the emulsion is stable, i.e. the slag
should not be too viscous, or too 'watery'. Only in this way can the
kinetics of the removal of the impurities be enhanced.

34
2

For encouraging quick formation of the appropriate type of slag,

lime/dolomite/other fluxing agents with adequate reactivity are added right from
the beginning of the blow. The reactivity of the fluxing agents, primarily lime
(consumption 60-100 kg/tls), determines how quickly slag is formed (typically
within 4-5 minutes after the commencement of the blow).

The rate at which oxygen is blown through the lance, the number of openings
(holes) on the lance tip, the distance between the lance tip and the bath surface
(lance height), the characteristics of the oxygen jets as they impinge on the bath
surface, the volume, basicity and fluidity of the slag, the temperature conditions
in the bath and many other operational variables influence the refining.

34
3

34
4

34
5

There are two distinct zones of refining in a LD vessel viz. the

reactions in the emulsion and in the bulk phase. The
contribution of bulk refining, i.e. refining in impact zone and at
the bulk slag-metal interface, is dominant in the beginning
since emulsion is yet to form properly. It has also been
believed that substantial decarburisation of droplets can occur
because of its free exposure to an oxidising gas, particularly
in the beginning. As the emulsion builds up the emulsion
refining attains a dominant role. The bulk phase refining
dominates again towards the end when the emulsion
collapses.
34
6

Conditions for dephosphorisation are that the slag should be basic, thin and
oxidising and, that the temperature should be low.

Dephosphorisation, therefore, does not take place efficiently until such a slag is
formed. Such a slag is formed in LD process only after the initial 4-6 minutes of
blowing.

The rate of dephosphorisation picks up concurrently with the rate of

decarburisation.

For efficient decarburisation as well as dephosphorisation the slag should,

therefore, form as early as possible in the process. If a preformed slag is
present as in a double slag practice wherein the second, slag is retained in the
vessel in part or full, the decarburisation rate curve rises more steeply in the
beginning

34
7

Dephosphorisation is very rapid in the emulsion because of the

increased interfacial area and efficient mass transport.
Phosphorus should, therefore, be fully eliminated before the
emulsion collapses. If this is not achieved the heat will have to
be kept waiting for dephosphorisation to take place and, in the
bulk phase, it is extremely slow as compared to that in the
emulsion. In general dephosphorisation should be over by the
time carbon is down to 07-10%, i.e. well ahead of the collapse
of emulsion which begins at around 03%C.

34
8

The relative rates of dephosphorisation and decarburisation can be

controlled by adjusting the lance height or by adjusting the flow rate of
oxygen.

Raising the height of the lance or decreasing the oxygen pressure

decreases the gas-metal reactions in the emulsion (i.e. decarburisation)
and vice versa.

The dephosphorisation reaction is thus relatively increased by the above

change and vice versa. Towards the end when temperature is high the
danger of phosphorus reversion does exist but it can be prevented by
maintaining a high basicity of the slag.

34
9

The process of decarburization includes at least three stages:

supply of reagents - carbon and oxygen - to the reaction site;

the reaction [c] + [0] proper; and

evolution of reaction products - CO bubbles into the gaseous phase .

. The apparent activation energy of the reaction [C] + [0] = CO is

relatively small (according to various researchers, E = 80000-120000
J/mol), which suggests that the reaction occurs practically
instantaneously. The solubility of CO in molten steel is also negligible.
Accordingly, the process can be limited by either the first or the third
stage.

35
0

The nature of kinetic curves of carbon burning-off at its various

concentrations is different: on attaining a certain 'critical level of
concentration of carbon (0.15-0.35%), the rate of carbon oxidation is
observed to drop noticeably.

It has also been established in experiments that the critical carbon

concentration is determined by the intensity of supply of oxidant to the bath
(it increases with increasing intensity of oxygen supply and decreases
during bath boil or metal stirring).

35
1

Thus, at carbon concentrations above the critical value, the intensity of

decarburisation reaction is determined by the supply of the oxidant and at
those below the critical value, by carbon diffusion to the reaction place .

This means practically that, if the carbon content of the metal is sufficiently
high, the rate of carbon oxidation will be higher at a higher intensity of oxygen
supply. At low concentrations of carbon, however, a higher level of intensity of
oxygen supply will not produce the desired effect and the bath should be
agitated forcedly (in order to intensify the supply of carbon to the reaction
place) so as to increase the rate of carbon oxidation.

35
2

The rate of decarburization can also be limited by the third stage, the evolution of CO.
For a bubble of CO to form in metal, It must overcome the pressure of the column of
metal (pm), slag (psl), and of the atmosphere (pat) above the bubble and also the forces
of the cohesion with the liquid, 2/r i.e.
pCOev pm + psl + pat + 2/r

The value of 2/r becomes practically sensible at low values of bubble radius: at r > I
mm it can be neglected. Formation of bubbles in the bulk of liquid metal is practically
impossible.. They can only form on interfaces between. phases, such as slag - metal,
non-metallic inclusion - metal, gas bubble - metal or lining - metal. The most favorable
conditions for the nucleation of CO bubbles exist on boundaries between the metal
and refractory lining which has a rough surface and is poorly wettable by the metal

35
3

35
4

35
5

35
6

Slag evolution
During Blow

35
7

High silicon pig iron is required in the acid

steelmaking processes to make relatively acid slag
to ensure longer life of the refractory lining.

Oxidation of silicon also generates sufficient heat

required in case of the Bessemer process.

However basic steelmaking processes need low

silicon iron because the entire amount of acid silica
due to the oxidation of silicon has to be neutralized
by lime to produce slag with basicity (CaO / SiO 2
ratio) between 2 and 4 needed for effective
desulphurisation and dephosphorisation.
35
8

Due to the strong attraction between iron and silicon,

the Fe-Si system exhibits large negative deviation
from the Raoults low. The activity coefficient of silicon
in iron in presence of other elements is given by :

log fSi = 0.18%C + 0.11% Si + 0.058% Al

-0.058 %S + 0.025 % V + 0.014 % Cu
+ 0.005 % Ni
+ 0.002 % Mn 0.0023 % Co 0. 23 %O

Oxidation of silicon is an exothermic reaction and

provides some of the heat necessary for rise of
temperature of the bath during blowing.

35
9

Si O reaction is governed by G0 vs T equation:

[Si ]+ 2 [O] = (SiO2 ), Go = -14200 + 55.0 T cals.

The activity coefficient of oxygen decreases and that of silicon increases

with increasing silicon content in iron.

Silica is a very stable oxide, hence once silicon is oxidised to SiO2 the
danger of its reversion does not arise.

36
0

36
1

36
2

The extremely low activity of silica in basic steelmaking

slag poses no danger of preferential reduction of silica
like that of phosphorus removal.

In basic steelmaking process the silicon content of pig

iron should be kept as low as possible to decrease the
lime consumption with the prime objective of controlling
the required basicity for phosphorus removal at a
minimum slag volume.

In case of high silicon entering the basic steelmaking

furnace double slag practice has to be adopted.

Alternatively, external desiliconisation of the hot metal

has to be done outside the blast furnace before charging
it in a basic steelmaking furnace.

36
3

About 50 to 75% of the manganese in the

burden gets reduced along with the pig iron
resulting its manganese content between 0.5
to 2.5%.
During steelmaking major amount of
manganese is lost into the slag and very little
is utilized to meet the specifications.
Some manganese is required to control the
deleterious effects of sulphur and oxygen and
also for improvement of mechanical properties
of the steel.

36
4

Hence conditions for maximum recovery of

manganese can be derived by considering the
equilibria:
(Fe2+) + [Mn] = (Mn2+) +[Fe]
(FeO) (+[
[Fe]
a Mn]
) a = (MnO)
(
) f+ %
Fe
K

or K

Mn 2

Fe

( a Fe 2 ) a Mn

Mn 2

Fe

( 23)

( Fe 2 ) f Mn % Mn

( Mn 2 ) % Fe

( 24 )

( Fe 2 ) % Mn

At equilibrium the Mn slag-metal distribution

( relation
)
( isFe 2given
by
)
Mn 2
K
% Mn
% Fe

( 25)

36
5

From the equation it is apparent that the

conditions for the highest possible recovery of Mn
i.e. minimum slag-metal distribution ratio are
i) min (Fe2+), requiring a low FeO content in the
slag.
ii) min K requires a low SiO2 content and a high
temperature as evident from the relation showing
effect
log K various
3.1 SiO anions
2.5 POin the
2.4slag.
O 1 .5 F
4
4

3
4

36
6

From the figure it is

evident that for slags
containing about 20%
MnO, a maximum of 0.1%
Mn is found in metal.
The slag containing 50%
SiO2 (the rest being FeO
and MnO), with increasing
Mn content of the metal
the (MnO) content of the
slag increases whereas
the oxygen content of the
metal decreases and
silicon content increases.
36
7

Despite its very low boiling point significant

amount of P gets dissolved in pig iron due to
strong attraction for iron.
Making use of the interaction coefficients for
the effect of various elements on the activity
coefficient of phosphorus in iron, the activity
of P can be estimated by the expression:
logfP = 0.13%C + 0.13%O + 0.12%Si +
0.062%P + 0.024%Cu + 0.028%S +
0.006%Mn 0.0002%Ni 0.03%Cr

36
8

A very close stability of FeO, Fe2O3 and P2O5 is

evident from the iron and phosphorus lines in
the Ellingham diagram.
Hence practically all the phosphorus present in
the ore gets reduced along with iron in the blast
furnace and joins pig iron.
During steelmaking the activity of P2O5 in the
slag of basicity 2.4 is reduced drastically to 10 15-10-20.
Activity of P2O5 in steelmakig slag is very low
even if it contains 25% P2O5.
36
9

37
0

i.e. 2[P] + 5[O] + 3(O2-) = 2 (PO43-)

(12)

Thus for effective removal of phosphorus basic

steelmaking processes have to employ slags of
high basicity.
The distribution of phosphorus between slag and
metal can be dessribed as
2[P] + 5(FeO) + 3 (CaO) = (3 CaO.P2O5) + 5[Fe]
2
a PO
3
4

a i.e.
.a 2[P]
.a (3O 2 )
2
P

5
O

Applying Temkin rule 2+

5[O] + 3(O: ) = 2 (PO43-)

2
PO
3

(14 )

f P % P fO % O
2

(13)

3
O 2

The dephosphorising index, D P which is the ratio of phosphorus in slag to that in metal, is given as
DP

( PO 3 )
4

% P

1/ 2

%O5/ 2 (O

)3/ 2

(15)
37
1

From the figure it is clear that DP

increases with increase in the
(FeO) content upto 15% due to
the high oxidizing power.
Beyond this DP decreases due to
decrease in the lime proportion.

Dephosphorisation is more
effective at lower temperature
because DP increases with
decrease of temperature.

37
2

37
3

The soda ash is 100 times more effective compared to

lime on molar basis but it is avoided in practice due to its
severe corrosive action on furnace lining.
The magnesia content of a basic steelmaking slag
reaches equilibrium with the lining hence not under
control and MnO depends on charge and hence not much
adjustable.
The steel maker has the option of controlling lime, silica
and FeO.
For charges containing high % P more than one slags are
made to dephosphorise metal bath to the desired level.
In brief high basicity, low temperature, and FeO content
around 15% favour dephosphorisation of metal by basic
slags.

37
4

The optimum conditions for dephosphorisation

can be
derived( from
the equation defining the index:
)
PO
5/ 2
DP
K % O ( O )3 / 2
% P
3
4

1/ 2

1. Basic

slag gives a high value of O22. High lime content lime is the divalent oxide
making the largest contribution to K (log K' =
21N'Ca++ + 18 N'Mg++ + 13N'Mn++ + 12 N'Fe++
3. Ferrous oxide close to 15% .
4. Low temperature gives a high value of K.
37
5

In refining of steel oxidation of Si, Mn and P takes

place at the slag-metal interface.

The oxidation of carbon practically does not take place

at the slag-metal interface because of the difficulty of
nucleation of CO bubbles there.

C-O reaction takes place at the gas metal interface

since it eliminates the necessity of nucleating gas
bubbles.

During refining of steel oxygen has to dissolve first in

the bath before it reacts with the dissolved impurities.

In the absence of other slag forming constituents at

1600oC liquid iron can dissolve oxygen up to at 0.23
wt.%
37
6

In steel making the reaction between carbon

and dissolved oxygen is of utmost importance.

Generally pig iron contains about 4 wt%

carbon.

The solubility of carbon in steel is effected by

the presence of impurities and alloying
elements.

Presence of Nb, V, Cr, Mn and W increase

solubility of carbon in iron where as presence
of Co, Ni, Sn and Cu decrease it.

37
7

Thus solubility of carbon in steel can be calculated by combining the binary data from the following equation:

37
8

Oxidation of Carbon can be discussed

according to the reaction:
C + O = CO,
G0= -5350 9.48T cals.

pCO
pCO
K

a c aO [ fc%C ][ fo%O]
pCO
pCO
[% C ][% O ]

K fc fo
K

At any chosen pressure of CO, % C vs % O

indicates inverse hyperbolic relationship
37
9

During oxidation period oxygen is continuously transferred

from the slag to the bath, where it continuously reacts with
carbon to give CO.
The main resistance to the oxygen flow is the slagmetal and
the metalgas interfaces, whereas inside the steel bath the
transfer of dissolved oxygen is very fast.

38
0

The activity coefficient

of carbon in iron
increases with
increasing carbon
content and that of
oxygen decreases with
increasing carbon
content.
The net result is that the
product [% C] [% O] for
a given pCO decreases
slightly with increasing
carbon content as
shown in Figure

38
1

Since steel making is a dynamic process, the

concentration of carbon and oxygen in the bulk
metal phase is not in equilibrium with the
prevailing CO-pressure in the bubbles.

At the gas bubblemetal interface the reaction is

close to equilibrium.

The experimentally observed excess oxygen and

carbon in the bulk metal phase is thus helpful in
transfer of the reactants by diffusion to the gasmetal interface in the violently stirred metal bath.

38
2

As [% O] increases with (aFeO) in slag and [% O]

decreases with [% C] in the bath.

it follows that the iron oxide contents of the slag

increases with decreasing carbon in steel during
refining.

Hence there is a general trend in the variation of slag

composition with the carbon content of the metal.

For a given total iron oxide in a slag, a lower carbon in

the steel corresponds to a higher sum of (% SiO 2 + %
P2O5) in the slag.

38
3

Within the range of basic slags, for a given sum of %

CaO + % MgO + % MnO the carbon content of steel
does not vary much with the FeO content of the slag.

38
4

During steelmaking i.e. refining of pig iron where

impurities like carbon, silicon, manganese and
phosphorus are eliminated to the required level oxygen,
nitrogen and hydrogen get dissolved as harmful
impurities.

As solubility decreases with decrease of temperature

excess gases dissolved in steel are liberated during
solidification.

The evolution of the gas gives rise to the formation of

skin or pin holes, blow holes, pipes etc.

The unsoundness caused by these cavities affect the

mechanical properties of steel

38
5

Nitrogen pick up during steel making:

open atmosphere
raw material charged
during melting and/or refining

Effect of nitrogen in steel:

yield-point phenomena
AlN causes intergranular fracture
nitrogen stabilizes the austenitic structure

Factors affecting the nitrogen solubility in steel.

partial pressure of nitrogen in the blast
time of contact
length of after blow and
the bath temperature

38
6

[wt.%H] =

Since nitrogen dissolves atomically in liquid iron and steel

in very small proportion its solubility can be discussed in
terms of Sieverts and Henrys laws

There is slow rise in solubility in solid state with increasing

temperature but at the melting point it increases very
rapidly. It also rises in liquid steel but at a slow rate.

Presence of vanadium, niobium, tantalum, chromium,

manganese, molybdenum, and tungsten increases
nitrogen solubility in iron whereas it decreases in
presence of nickel, cobalt, silicon and carbon

38
7

38
8

Hydrogen pick up steel making:

wet solid and rusty charge materials
atmospheric humidity
wet refractory channels, runners and containers

Effect of hydrogen in steel

Decreases ductility
Appearance of hairline cracks seriously affect the
mechanical properties
Formation of blow holes and pin holes.

38
9

Water vapour coming in

contact with steel or slag leads
to the formation of hydrogen
which gets dissolved in steel
melt as per reaction:
H2O (g) = 2[H]1wt.% + [0]1 wt.%

At the melting point of iron

solubility in delta iron is
approximately 10 mL/ 100g.
Beyond this hydrogen will be
rejected during solidification to
produce unsound porous ingots
due to gas evolution.

39
0

Thus partial pressure of hydrogen, and

composition of steel and its temperature
control the hydrogen content of steel.
According to Sieverts law solubility of
hydrogen in pure iron is expressed as:

Presence of niobium, tantalum, titanium and

nickel increases the solubility of hydrogen in
iron whereas it decreases in pressure of
carbon, silicon, chromium and cobalt.
39
1

The objectives of vacuum degassing include removal

of hydrogen from steel to avoid long annealing
treatment, removal of oxygen as carbon monoxide
and production of steels with very low carbon content
(< 0.03%).

The principle is based on the usefulness of the

Sieverts law relationship.

The equation demonstrates that subjecting the

molten steel to vacuum will decrease the hydrogen,
nitrogen as well as the oxygen content of the steel
according to the following reasons:

39
2

2[H] = H2 (g)
2[N] = N2 (g)
[C] + [O] = CO (g)
The effectiveness of vacuum treatment
increases with increase in the surface area
of liquid steel exposed to vacuum.
For this purpose metal is allowed to flow in
the form of thin stream or even fall as
droplets to accelerate the degassing
process.
39
3

A number of methods available on commercial scale for

vacuum treatment of steel may categorized into three
groups :
1. Ladle Degassing
The teeming ladle filled with steel to one fourth of its
height is placed inside a vacuum chamber.
the melt is stirred either by bubbling argon or by
electromagnetic induction
Introduction of gas for stirring provides interface which
facilitates degassing.
In general pumping is carried out to attain the
ultimate vacuum of 1-10 mm Hg. which is supposed to
be adequate for degassing.

39
4

2.

Stream Degassing
In this case molten steel is allowed to flow
down under vacuum as a stream from the
furnace to ladle to another ladle or a mould.
A very high rate of degassing is achieved
due to large increase in surface area of
molten steel in the form of falling droplets.
Thus choice of proper vacuum pump and
vacuum chamber is important to achieve
the adequate level of degassing.

39
5

3.

Circuilation Degassing

R-H degassing process

The average rate of circulation
is
12 tons/min.
Twenty minutes are required
to
treat 100 tons of steel to bring
down 90% reduction of
hydrogen content.

39
6

D-H Vessel.

The
chamber
through 50-60
cycle time of 20
steel is exposed

is
moved
cm with a
sec. 10-15%
at a time.

7-10 cycles are required to

expose the entire steel
once.
Adequate
obtained
in 20-30
minutes.

degassing
cycles

in

is

15-20

39
7

Smarajit Sarkar
Department of Metallurgical and Materials Engineering
NIT Rourkela

39
8

39
9

Oxidation of carbon : Bottom blowing increases sharply the

intensity of bath stirring and increases the area of gas-metal
boundaries (10-20 times the values typical of top blowing) .

Since the hydrocarbons supplied into the bath together with

oxygen dissociate into H2, H2O and CO2 gas bubbles in the
bath have a lower partial pressure of carbon monoxide (Pco )

All these factors facilitate substantially the formation and

evolution of carbon monoxide, which leads to a higher rate of
decarburization in bottom blowing

40
0

 The

degree of oxidation of metal and slag

Removal

of phosphorous: Since the slag of

the bottom-blown converter process have a
low degree of oxidation almost during the
whole operation, the conditions existing
during these periods are unfavorable for
phosphorus removal

40
1

Problems arise when the layer of foaming slag created on the

surface of the molten metal exceeds the height of the vessel and
overflows, causing metal loss, process disruption and
environmental pollution. This phenomenon is commonly referred to
as slopping.

40
2

Better mixing and homogeneity in the bath offer the following

advantages:

Less slopping, since non-homogeneity causes formation of regions with

high supersaturation and consequent violent reactions and ejections.

Better mixing and mass transfer in the metal bath with closer approach
to equilibrium for [C]-[O]-CO reaction, and consequently, lower bath
oxygen content at the same carbon content

40
3

Better slag-metal mixing and mass transfer and

consequently, closer approach to slag- metal equilibrium,
leading to:
o lower FeO in slag and hence higher Fe yield
o transfer of more phosphorus from the metal to the slag (i.e.
better bath dephosphorisation)
o transfer of more Mn from the slag to the metal, and thus
better Mn recovery
o lower nitrogen and hydrogen contents of the bath.
More reliable temperature measurement and sampling of
metal and slag, and thus better process control
Faster dissolution of the scrap added into the metal bath

40
4

small amount of inert gas, about 3% of the volume of oxygen

blown from top, introduced from bottom, agitates the bath so

effectively that slopping is almost eliminated.
However

for obtaining near equilibrium state of the system

inside the vessel a substantial amount of gas has to be

introduced from the bottom.
If

20-30% of the total oxygen, if blown from bottom, can cause

adequate stirring for the system to achieve near equilibrium

conditions. The increase beyond 30% therefore contributes
negligible addition of benefits.

40
5

The more the oxygen fraction blown from bottom the

less is the post combustion of CO gas and
consequently less is the scrap consumption in the
charge under identical conditions of processing.

Blowing of inert gas from bottom has a chilling effect

on bath and hence should be minimum. On the
contrary the more is the gas blown the more is the
stirring effect and resultant better metallurgical results.
A optimum choice therefore has to be made
judiciously.
40
6

40
7

As compared to top blowing, the hybrid blowing

eliminates the temperature and concentration
gradients and effects improved blowing control, less
slopping and higher blowing rates. It also reduces over
oxidation and improves the yield. It leads the process
to near equilibrium with resultant effective
dephosphorisation and desulphurisation and ability to
make very low carbon steels.

40
8

 What

is blown from the bottom, inert gas or oxygen?

How much inert gas is blown from the bottom?
At what stage of the blow the inert gas is blown,
although the blow, at the end of the blow, after the
blow ends and so on?
What inert gas is blown, argon, nitrogen or their
combination?
How the inert gas is blown, permeable plug, tuyere,
etc.?
What oxidising media is blown from bottom, oxygen or
air?
If oxygen is blown from bottom as well then how much
of the total oxygen is blown from bottom ?
40
9

41
0

The processes have been developed to obtain the combined advantages of

both LD and OBM to the extent possible. Therefore the metallurgical
performance of a hybrid process has to be evaluated in relation to these two
extremes, namely the LD and the OBM. The parameters on which this can
be done are :

Iron content of the slag as a function of carbon content of bath

Oxidation levels in slag and metal

Manganese content of the bath at the turndown

Desulphurisation efficiency in terms of partition coefficient

Dephosphorisation efficiency in terms of partition coefficient

Hydrogen and nitrogen contents of the bath at turndown

Yield of liquid steel

41
1

41
2

The oxidizing conditions of a heat in a steelmaking plant, the

presence of oxidizing slag, and the interaction of the metal with the
surrounding atmosphere at tapping and teeming - all these factors
are responsible for the fact that the dissolved oxygen in steel has a
definite, often elevated, activity at the moment of steel tapping. The
procedure by which the activity of oxygen can be lowered to the
required limit is called deoxidation. Steel subjected to deoxidation is
termed 'deoxidized'. If deoxidized steel is 'quiet during solidification
in moulds, with almost no gases evolving from it, it is called 'killed
steel'.

41
3

If the metal is tapped and teemed without being deoxidized, the reaction
[O] + [C] = COg will take place between the dissolved oxygen and
carbon as the metal is cooled slowly in the mould. Bubbles of carbon
monoxide evolve from the solidifying metal, agitate the metal in the
mould vigorously, and the metal surface is seen to 'boil'. Such steel is
called 'wild'; when solidified, it will be termed 'rimming steel' .

In some cases, only partial deoxidation is carried out, i.e. oxygen is only
partially removed from the metal. The remaining dissolved oxygen
causes the metal to boil for a short time. This type of steel is termed
'semi-killed'.

41
4

Thus, practically all steels are deoxidized to some or other extent so

as to lower the activity of dissolved oxygen to the specified limit.

The activity of oxygen in the metal can be lowered by two methods: (I)
by lowering the oxygen concentration, or
(2) by combining oxygen into stable compounds.

There are the following main practical methods for deoxidation of steel:
(a) precipitation deoxidation, or deoxidation in the bulk;
(b) diffusion deoxidation;
(c) treatment with synthetic slags; and
(d) vacuum treatment.

41
5

The advantages of continuous casting (over ingot

casting) are:
It is directly possible to cast blooms, slabs and
billets, thus eliminating blooming, slabbing mills
completely, and billet mills to a large extent.
Better quality of the cast product.
Higher crude-to-finished steel yield (about 10 to
20% more than ingot casting).
Higher extent of automation and process control.

41
6

41
7

Solidification must be completed before the withdrawal

rolls.

The liquid core should be bowl-shaped as shown in the

Figure and not pointed at the bottom (as indicated by the
dotted lines), since the latter increases the tendency for
undesirable centerline (i.e. axial) macro-segregation and
porosity

The solidified shell of metal should be strong enough at

the exit region of the mould so that it does not crack or
breakout under pressure of the liquid.

41
8

The surface area-to-volume ratio per unit length of

continuously cast ingot is larger than that for ingot
casting. As a consequence, the linear rate of
solidification (dx/dt) is an order of magnitude
higher than that in ingot casting.

The dendrite arm spacing in continuously cast

products is smaller compared with that in ingot
casting.

41
9

42
0

Macro-segregation is less, and is restricted to the

centreline zone only.

Endogenous inclusions are smaller in size, since they

get less time to grow. For the same reason, the blow
holes are, on an average, smaller in size.

Inclusions get less time to float-up. Therefore, any

non-metallic particle coming into the melt at the later
stages tends to remain entrapped in the cast product.

42
1

In addition to more rapid freezing, continuous casting differs

from ingot casting in several ways. These are noted below.
Mathematically speaking, continuously cast ingot is infinitely
long. Hence, the heat flow is essentially in the transverse
direction, and there is no end-effect as is the case in ingot
casting (e.g. bottom cone of negative segregation, pipe at the
top, etc.).

The depth of the liquid metal pool is several metres long.

Hence, the ferrostatic pressure of the liquid is high during the
latter stages of solidification, resulting in significant difficulties of
blow-hole formation.

42
2

Since the ingot is withdrawn continuously from the mould, the frozen
layer of steel is subjected to stresses. This is aggravated by the
stresses arising out of thermal expansion/ contraction and phase
transformations.

Such stresses are the highest at the surface. Moreover, when the
ingot comes out of the mould, the thickness of the frozen steel shell is
not very appreciable. Furthermore, it is at around 1100-1200C, and is
therefore, weak. All these factors tend to cause cracks at the surface
of the ingot leading to rejections.

Use of a tundish between the ladle and the mould results in extra
temperature loss. Therefore, better refractory lining in the ladles,
tundish, etc. are required in order to minimise corrosion and erosion
by molten metal.
42
3

Smarajit Sarkar
Department of Metallurgical and Materials Engineering
NIT Rourkela

42
4

Primary steelmaking is aimed at fast melting

and rapid refining. It is capable of refining at
a macro level to arrive at broad steel
specifications, but is not designed to meet
the stringent demands on steel quality, and
consistency of composition and temperature
that is required for very sophisticated grades
of steel. In order to achieve such
requirements, liquid steel from primary
steelmaking units has to be further refined in
the ladle after tapping. This is known as
Secondary Steelmaking.
42
5

improvement

in quality
improvement in production rate
decrease in energy consumption
use of relatively cheaper grade or
alternative raw materials
use of alternate sources of energy
higher recovery of alloying elements.

42
6

 Lower

impurity contents .
Better cleanliness. (i.e. lower inclusion
contents)
Stringent quality control. (i.e. less variation
from heat-to-heat)
Microalloying to impart superior properties.
Better surface quality and homogeneity in
the cast product.

42
7

 The

term clean steel should mean a steel

free of inclusions. However, no steel can
be free from all inclusions.
Macro-inclusions are the primary harmful
ones. Hence, a clean steel means a
cleaner steel, i.e., one containing a much
lower level of harmful macro-inclusions.)

42
8

In practice, it is customary to divide

inclusions by size into macro inclusions and
micro inclusions. Macro inclusions ought to
be eliminated because of their harmful
effects. However, the presence of micro
inclusions can be tolerated, since they do
not necessarily have a harmful effect on the
properties of steel and can even be
beneficial. They can, for example, restrict
grain growth, increase yield strength and
hardness, and act as nuclei for the
precipitation of carbides, nitrides, etc.
42
9

 The

critical inclusion size is not fixed but

depends on many factors, including service
requirements.
Broadly speaking, it is in the range of 5 to 500
m (5 X 10-3 to 0.5 mm). It decreases with an
increase in yield stress. In high-strength steels,
its size will be very small.
Scientists advocated the use of fracture
mechanics concepts for theoretical estimation of
the critical size for a specific situation.
43
0

Precipitation due to reaction from molten steel or during

freezing because of reaction between dissolved oxygen
and the deoxidisers, with consequent formation of oxides
(also reaction with dissolved sulphur as well). These are
known as endogenous inclusions.
Mechanical and chemical erosion of the refractory lining
Entrapment of slag particles in steel
Oxygen pick up from the atmosphere, especially during
teeming, and consequent oxide formation.
Inclusions originating from contact with external sources
as listed in items 2 to 4 above, are called exogenous
inclusions.

43
1

With a lower wettability (higher value of Me

inc

),

an inclusion can be retained in contact with the

metal by lower forces, and therefore, can break
off more easily and float up in the metal. On the
contrary, inclusion which are wetted readily by the
metal, cannot break off from it as easily.

43
2

Carryover slag from the furnace into the ladle

should be minimised, since it contains high
percentage of FeO + MnO and makes efficient
deoxidation fairly difficult.

Deoxidation products should be chemically

stable. Otherwise, they would tend to
decompose and transfer oxygen back into liquid
steel. Si02 and Al203 are preferred to MnO.
Moreover the products should preferably be
liquid for faster growth by agglomeration and
hence faster removal by floatation. Complex
deoxidation gives this advantage.

43
3

o
o

Stirring of the melt in the ladle by argon flowing through

bottom tuyeres is a must for mixing and homogenisation,
faster growth, and floatation of the deoxidation products.
However, very high gas flow rates are not desirable from
the cleanliness point of view, since it has the following
adverse effects:
Too vigorous stirring of the metal can cause
disintegration of earlier formed inclusion conglomerates.
Re-entrainment of slag particles into molten steel.
Increased erosion of refractories and consequent
generation of exogenous inclusions.
More ejection of metal droplets into the atmosphere with
consequent oxide formation.

43
4

43
5

The varieties of secondary steelmaking

processes that have proved to be of
commercial value can broadly be categorised
as under:
Stirring treatments
Synthetic slag refining with stirring
Vacuum treatments
Decarburisation techniques
Injection metallurgy
Plunging techniques
Post-solidification treatments.
43
6

43
7

 Ladle

degassing processes (VD, VOD, VAD)

Stream

degassing processes

Circulation

degassing processes (DH and RH).

43
8

43
9

44
0

Molten steel is contained in the ladle. The two legs of the vacuum
chamber (known as Snorkels) are immersed into the melt. Argon is
injected into the up leg.

Rising and expanding argon bubbles provide pumping action and lift the
liquid into the vacuum chamber, where it disintegrates into fine droplets,
gets degassed and comes down through the down leg snorkel, causing
melt circulation.

The entire vacuum chamber is refractory lined. There is provision for

argon injection from the bottom, heating, alloy additions, sampling and
sighting as well as video display of the interior of the vacuum chamber.

44
1

44
2

Why RH-OB Process?

To meet increasing demand for cold-rolled steel sheets with improved
mechanical properties, and to cope with the change from batch-type to
continuous annealing, the production of ULC steel (C < 20 ppm) is
increasing.
A major problem in the conventional RH process is that the time
required to achieve such low carbon is so long that carbon content at
BOF tapping should be lowered. However, this is accompanied by
excessive oxidation of molten steel and loss of iron oxide in the slag.
It adversely affects surface the quality of sheet as well.

44
3

Hence, decarburization in RH degasser is to be speeded up.

This is achieved by some oxygen blowing (OB) during
degassing.

The RH-OB process, which uses an oxygen blowing facility

during degassing, was originally developed for
decarburization of stainless steel by Nippon Steel Corp.,
Japan, in 1972.

Subsequently, it was employed for the manufacture of ULC

steels.

The present thrust is to decrease carbon content from

something like 300 ppm to 10 or 20 ppm within 10 min.
Cont

44
4

44
5

Conventional AOD, no top blowing is involved. Only a

mixture of argon and oxygen is blown through the
immersed side tuyeres. However, the present AOD
converters are mostly fitted with concurrent facilities for
top blowing of either only oxygen, or oxygen plus inert
gas mixtures using a supersonic lance as in BOF
steelmaking.

44
6

Initially, when the carbon content of the melt is high, blowing

through the top lance is predominant though the gas mixture
introduced through the side tuyeres also contains a high
percentage of oxygen.

However, as decarburisation proceeds, oxygen blowing from

the top is reduced in stages and argon blowing increased. As
stated earlier, some stainless steel grades contain nitrogen as
a part of the specifications, in which case, nitrogen is
employed in place of argon in the final stages.

44
7

Simplified by Hiltey and Kaveney

44
8

44
9

45
0

45
1

This process produces molten iron in a two-step reduction melting

operation. One reactor is melter-gasifier and the other is prereducer. In the pre-reducer, iron oxide is reduced in counter-flow
principle. The hot sponge is discharged by screw conveyors into the
melting reactor.

Coal is introduced in the melting-gassifying zone along with

oxygen gas at the rate of 500-600 Nm3/thm. The flow velocity is
chosen such that temperature in the range of 1500-1800C is main
tained. The reducing gas containing nearly 85% CO is hot dedusted
and cooled to 800-900C before leading it into the pre-reducer

45
2

45
3

In the FINEX Process fine ore is preheated and reduced to DRI in a

train of four or three stage fluidized bed reactors.

The fine DRI is compacted and then charged in the form of Hot
Compacted Iron (HCI) into the melter gasifier. So, before charging
to the melter- gasifier unit of the FINEX unit, this material is
compacted in a hot briquetting press to give hot compacted iron
(HCI)

since the melter- gasifier can not use fine material (to ensure
permeability in the bed).

Non-coking coal is briquetted and is fed to the melter gasifier where

it is gasified with oxygen
45
4

As a standard guide the temperature rise attainable by

oxidation of 001 % of each of the element dissolved in
liquid iron at 1400C by oxygen at 25C is calculated
assuming that no heat is lost to the surroundings and
such data are shown below
.

45
5

Ahindra Ghosh and Amit Chatterjee: Ironmaking and Steelmaking Theory and Practice, PrenticeHall of India Private Limited, 2008
Anil K. Biswas: Principles of Blast Furnace Ironmaking, SBA Publication,1999
R.H.Tupkary and V.R.Tupkary: An Introduction to Modern Iron Making, Khanna Publishers.
R.H.Tupkary and V.R.Tupkary: An Introduction to Modern Steel Making, Khanna Publishers.
David H. Wakelin (ed.): The Making, Shaping and Treating of Steel (Ironmaking Volume), The
AISE Steel Foundation, 2004.
Richard J.Fruehan (ed.): The Making, Shaping and Treating of Steel (Steeelmaking Volume), The
AISE Steel Foundation, 2004.
A.Ghosh, Secondary Steel Making Principle & Applications, CRC Press 2001.
R.G.Ward: Physical Chemistry of iron & steel making, ELBS and Edward Arnold, 1962.
F.P.Edneral: Electrometallurgy of Steel and Ferro-Alloys, Vol.1 Mir Publishers,1979
B. Ozturk and R. J. Fruehan,: "Kinetics of the Reaction of SiO(g) with Carbon Saturated Iron":
Metall. Trans. B, Vol. 16B, 1985, p. 121.
B. Ozturk and R. J. Fruehan: "The Reaction of SiO(g) with Liquid Slags, Metall. Trans.B,
Volume 17B, 1986, p. 397.
B. Ozturk and R. J. Fruehan:.Transfer of Silicon in Blast Furnace": , Proceedings of the fifth
International Iron and Steel Congress, Washington D.C., 1986, p. 959.
P. F. Nogueira and R. J. Fruehan: Blast Furnace Softening and Melting Phenomena - Melting
Onset in Acid and Basic Pellets", , ISS-AIME lronmaking Conference, 2002, pp. 585.

45
6

Paulo Nogueira, Richard Fruehan: "Blast Furnace Burden Softening and Melting
Phenomena-Part I Pellet Bulk Interaction Observation", , Metallurgical and Materials
Transactions B, Volume 35B, 2004, pp. 829.
P.F. Nogueira, Richard J. Fruehan: 'Fundamental Studies on Blast Furnace Burden
Softening and Melting", Proceedings of 2nd International Meeting on lronmaking,
September 2004, Vitoria, Brazil.
Paulo F. Nogueira, Richard J. Fruehan, "Blast Furnace Softening and Melting
Phenomena - Part III: Melt Onset and Initial Microstructal Transformation in Pellets",
submitted to Materials and Metallurgical Transactions B.
Paulo F. Nogueira, Richard J. Fruehan :Blast Furnace Burden Softening and Melting
Phenomena-Part II Evolution of the Structure of the Pellets", Metallurgical and
Materials Transactions, Volume 36B, 2005, pp. 583
MA Jitang: Injecuion of flux into Blast Furnace via Tuyeres for optimizing slag
formation ISIJ International, Volume 39, No7 1999,pp697
Y.S.Lee, J.R.Kim, S.H.Yi and D.J.Min: Viscous behavior of CaO-SiO2-Al2O3-MgOFeO Slag, Proceedings of VIIInternational Conferenceon -Molten slags,fluxes and
salts, The South African Institute of Minig and Metallurgy, 2004,pp225

45
7

45
8

Smarajit Sarkar
Department of Metallurgical and Materials Engineering
NIT Rourkela

45
9