See discussions, stats, and author profiles for this publication at: [Link]

net/publication/354778471

OPTIMISATION OF BLAST FURNACE TROUGH LINING REFRACTORY

PERFORMANCE

Conference Paper · May 2004

CITATION READS

1 711

4 authors, including:

Indra Chakraborty
Calderys
74 PUBLICATIONS 446 CITATIONS

SEE PROFILE

All content following this page was uploaded by Indra Chakraborty on 23 September 2021.

The user has requested enhancement of the downloaded file.

Tehran International Conference on Refractories, 4-6 May 2004

OPTIMISATION OF BLAST FURNACE TROUGH LINING REFRACTORY PERFORMANCE

I. N. Chakraborty, S. Bose, A. Mukherjee and D. K. Singh

Refractories Division, The Associated Cement Companies, India

Abstract: Performance of Blast Furnace (BF) Trough Refractories, both Ramming Mass and
Castable, depend not only on the material quality but has great dependence on the installation
practices as well as operating conditions of the furnace. In this paper our experience with trough
castable as well as ramming mass formulations, developed through our R & D effort, has been
discussed in length. We share the role of raw material base, including their chemistry, the
concentration of different constituents and their gradation on actual performance of the material
during use. Based on our experience of managing 16 BF troughs, the impact of trough design and
installation as well as operational practices on material performance has also been shared.

1. INTRODUCTION
Blast Furnace (BF) Troughs are designed to convey molten iron (HM) and slag, during tapping of the
BF, to transfer ladle. They also facilitate iron and slag separation while flowing. The effects of this
phenomenon on the refractory lining of runners are reaction with iron, slag and air, especially at the
iron – slag and slag – air interfaces. Abrasion of the refractory lining is also encountered, especially at
the impact area of the main trough and the tilting runner. Runner system must now meet the
expections of BF operators, i. e. long and predictable campaign lives, ease of installation, mechanised
dismantling and installation processes, safe working conditions; rapid drying capability; campaign life
without intermittent repairs and above all reduced operating costs. Sand with small amounts of clay
and coke, bonded by tar, did suffice in the past as trough lining material, but not any longer for high
capacity BFs. Todays Blast Furnace Trough (BFT) lining materials are Alumina based formulations
with Carbon and SiC. These products are available in both ramming mass and castable variety.
We at present maintain 16 BFTs in India, out of which 11 are lined with castables and the rest with
ramming masses. Both these formulations were developed through our in – house R & D effort. By
virtue of our over 10 years experience in this field, we have modified product bond system as well as
upgraded our refractory installation capabilities. In this process now we constantly achieve straight
BFT life of over 1,20,000 and 80,000 MT of HM through put with castable and ramming mass,
respectively. During this period of 10 years, we also have developed a pool of skilled human resources
and created requisite hardware facilities for BFT refractory installation and management.
The performance of BFT refractory depends on numerous parameters. In fact non – refractory related
aspects, viz. BF operational parameters, BFT design etc., play a dominant role in determining BFT life
(Fig. 1). In this paper we would share our experience of product development, product optimization
and briefly focus on role of lining design, installation practice and our BFT performance related
features.

2. BFT REFRACTORY CONSTITUENTS

Modern BFT formulations, both castable and ramming mass, are based on Al2O3 – SiC – C system.
The role of each of these constituents is discussed in this section.
Alumina source in BFT formulation, generally, is calcined bauxite and fused alumina aggregate. These
are the major constituent and form the bulk. The higher the alumina content, higher is the
refractoriness of the mix.
The presence of carbon increases the thermal conductivity and renders better thermal shock resistance.
Carbon is of utmost importance for corrosion resistance, since it prevents slag penetration and is not

250
I. N. Chakraborty, S. Bose, A. Mukherjee and D. K. Singh

wetted by slag. Carbon can be added in various forms such as pitch, resin, coke, carbon black or
graphite.
Silicon carbide also has high thermal conductivity and resists well against erosion as well as slag
attack and protects the carbon from oxidation. SiC also improves high temperature mechanical
properties of BFT formulation. The silicon carbides with a wide particle size distribution perform the
best in BFT formulation. It appears that the sizing of the silicon carbide is more critical than its purity.
In this context it is prudent to mention that the physical characteristics of BFT castable and ramming
mass formulations based on identical raw materials are different, especially their porosity and pore
size distribution (Fig. 2). These features reveal that the slag and metal penetration of ramming mass is
expected to be higher compared to that of castable.

3. CAUSES OF REFRACTORY WEAR IN TROUGH

In general it is strongly believed that BFT refractory wear is primarily attributable to the erosion
caused by hot slag as well as metal flow, chemical reaction with BF slag and related phenomena. A
recent work [1], however, reveals that erosion plays a very minor role in determining the BFT wear.
The analysis of the stress generated during the fluid flow reveals that the reason for the high wear at
the impact area, i. e. highest stressed area, is more complicated than normally thought. Normally it is
expected that the maximum wear would appear at the of re-circulation whirls at the impact point,
where the stresses are maximum, which is only a few Pascals. The maximum wear area, however,
does not correspond to this highest stress area as per calculation. Additionally, total power lost due to
friction in the stream over the whole runner is only a few kilowatts, which is too small and is not
sufficient to explain wear by only mechanical friction. It is, thus, concluded that mechanical factors
contribute very little in the wear of the BFT refractory lining.
During tapping the stream carries air bubbles and oxygen in the runner. At the impact point, these
bubbles are released by the molten mass and they reach the refractory walls downstream of the impact
point where oxidation reaction proceeds. It has, thus, been concluded mechanical erosion is linked
here with the corrosion and acts as a washing effect of the reacting layers.
In addition to the direct oxidation of carbonaceous constituents of the formulation following reactions
also are known to occur [2].
• Oxidation of SiC by reaction with CO gas generated by reaction between SiO2 of in the slag
and C in the pig iron.
(SiO2 + 2C → Si + 2CO, and subsequently
SiC + 2 CO → SiO2 + 3C)
• Oxidation of SiC by formation of FeO as per
SiO2 + 2 Fe → 2 FeO + Si, and subsequently
SiC + 2FeO → SiO2 + 2 Fe + C)
FeO of the aggregates as well as that of the overall formulation, thus, are important parameter, since
they determine the refractory wear of the BFT lining. The cracking and peeling-off in BFT material
occur normally by progressive sintering or with the increase of internal stress because of heating and
cooling cycles. The lining fails due to differential thermal expansion of the penetrated metal and
refractory during temperature change. Thermal shock on the lining can be reduced by:
• Reduced tap closing to opening time
• Using cover over trough to minimize the heat loss
A higher wear area is observed at the hot metal / slag interface, namely at the interface during the
pooling, where strong corrosion reactions occur between the slag, the hot metal and the refractory.
The refractory wear, thus, is linked with the reaction between the slag / refractory on one side and the
air / refractory - particularly oxidation on the other side. An increased oxidation resistance of the
carbon component can , thus, address aging symptoms of BFT material.
Based on microstructural analyses of used BFT material Fig. 3 schematically represents the wear
mechanism of BFT lining material when in contact with BF slag. The refractory wear appears to

251
Tehran International Conference on Refractories, 4-6 May 2004

proceed via oxidation of carbonaceous material, formation of low melting compounds and subsequent
dissolution of new phases as well as aggregates in slag. Improvements of direct, i. e. by atmospheric
oxygen, and indirect, i. e. by impurities like FeO, oxidation resistance of the product is a major route
for better BFT performance.

FACTORS AFFECTING TROUGH LINING LIFE

Lining Thickness Runner

shape Tapping sequence
14% 25%
HM Thrioughput Taphole
Design
Trough Lining
28%

33% HM Bath Depth

Fig. 1. Contribution of different factors in BF Trough refractory wear

Fig. 2. Comparative porosity and pore size distribution of BF Trough Ramming Mass and Castable

4. DEPENDENCE OF AGGREGATE WEAR ON ITS TYPE

During the development of the formulation numerous interactions studies between BF Slag, typical BF
slag compositions of Indian BFs are reported in Table 1, and BFT lining materials had been
conducted. Figs. 4 and 5 illustrate the effect of type, i. e. fused or sintered, of aluminous aggregates in
BFT formulations. It is fairly apparent that the aluminous aggregate quality plays a dominant role in
deciding the performance of these products.
Microscopic analyses, both Optical and Electron, reveal that the wear mechanism of these aggregates.
The phases formed at the interface and by interaction between alumina and penetrated liquid would
primarily be determined by the diffusion coefficient of the species present in BF slag, which are CaO,
SiO2, Al2O3, MgO and FeO. The more-mobile species tend to move ahead of the less-mobile one,

252
I. N. Chakraborty, S. Bose, A. Mukherjee and D. K. Singh

creating a local electostatic field, which may increase or decrease the diffusivity of the other ions,
depending on their charges and direction of migration. When one ion is much more mobile than any of
the oppositely charged ion, its effect on the diffusion coefficients of the oppositely charged ions is
relatively small, as in the case of the oxygen ion, which is much more mobile than the cations. It is
known that diffusion coefficient DFe > DM (M= Ca, Mg, Mn) > DAl >> Dsi. The concentration
distribution of these species in the aggregate - slag boundary layer can be explained in terms of the
mobility of the cations, which can be evaluated by the corresponding diffusion coefficients in the slag.
The results for both aggregates, in general, reveal that layer directly adjacent to the Al2O3 is CA6 and
the layer adjacent to this CA6 is hercynitic spinel comprising mainly aluminum, iron, manganese, and
magnesium cations. It is not difficult to understand the of spinel, since iron, calcium, manganese, and
magnesium ions move to the interface ahead of silicon ions and once they arrive at the interface, react
with Al2O3 to form the corresponding products. Although calcium ions diffused more slowly than iron
ions, but as its concentration is higher in the bulk slag, greater calcium iron ion concentration occur at
the interface, which explains CA6 formation in the boundary layer.

Fig. 3. Erosion Mechanism of BF Trough Al2O3 – SiC – C refractory, when in contact with BF Slag

Table 1. Blast Furnace Slag characteristics of SAIL

Parameters (% wt)
SiO2 BSL, Bokaro DSP, Durgapur RSP, Rourkela BSP, Bhilai
Al2O3 31.5 -3 2.0 32.5 - 33.5 31.0 - 33.0 33.0 - 34.5
CaO 22 - 24 23 - 24 22 - 23 17 - 19
MgO 32-33 31 - 33 30 - 32 32 - 33
FeO 10 - 11 8-9 8-9 10 - 11
MnO 0.4 - 0.6 0.5 - 0.7 0.55 - 0.7 0.55 - 0.6
S 0.7 - 0.8 0.8 - 0.9 0.4 - 0.7 0.7 - 0.8
CaO/SiO2 0.60 - 0.65 0.55 - 0.65 0.55 - 0.65 0.6 - 0.65
HM Temperature, deg C 0.96 - 1.0 0.95 - 1.0 0.91 - 0.97 0.99 - 1.0

253
Tehran International Conference on Refractories, 4-6 May 2004

The CA6 and spinel layers in the fused grain appear more continuous than those in sintered grains. A
thicker CA6 layer, ∼ 60 µm, forms in the fused aggregate compared to only ∼ 40 µm in the sintered
aggregate. CA6 layer formed in the fused grains has different morphology from the adjacent to the
sintered grain. At the normal slag temperature (1450 0C) CA6 layer that formed in the sintered
aggregate is thin and discontinuous. This morphology could be attributed to the presence of grain
boundaries and porosity in the sintered sample.
Additionally, Aluminum species from the Al2O3 also can diffuse readily through the grain boundaries
and the CA6 layer and dissolve into the slag; i. e. indirect dissolution of sintered aggregates can occur.
No slag penetration is observed in fused grains, apart from that at cracks or pores. In the sintered
aggregate slag penetration ∼ 300 µm into the grain was evident and are corroded and broken up by the
penetrated slag.

Fig. 4. Effect of base aggregate purity on the resistance to BF slag attack. In this case the base aggregate is
relatively impure

Fig. 5. Effect of base aggregate purity on the resistance to BF slag attack. In this case the base aggregate is
relatively pure

254
I. N. Chakraborty, S. Bose, A. Mukherjee and D. K. Singh

Continuous diffusion of FexO and MnO to the interface results in complex spinel formation. The
spinel layers, in both cases, are not dense and tended to disperse broadly toward the bulk slag.
According to MgO content of the slag, the morphology of the spinel changed from a few isolated
particles, to a discontinuously attached layer, to a continuously attached layer at 5%, 10% and 15 –
25% of MgO content, respectively. A lower-melting spinel increases its dissolution into the slag, while
the presence of grain boundaries and pores in the sintered grain hinder development of the
concentration profile in the boundary layer.

5. ROLE OF AGGREGATE GRADATION AND BINDER

Figs. 6–9 illustrate the results of interaction study between BFT castable formulations and BF slag.
Figs. 6 and 7 draw comparison between 2 hydraulically bonded (Bond A) castables with same
concentration of SiC but with different gradation. It is evident that product with wider SiC gradation
yield better slag resistance. Hydraulically bonded formulations, owing to presence of CaO in the
matrix, did not yield desired performance in large capacity BFs. Change in the aggregate quality or
change in the gradation failed to achieve the desired results.
Work with CaO free bond system revealed that with similar aggregate the slag attack resistance can be
improved significantly (Fig. 8). It is fairly evident that this formulation has significantly better BF slag
attack resistance. The limitation of this newly developed system, however, has been mixing of a non –
aqueous liquid (Bond B) at site [3]. Subsequent R & D work has been identification of CaO free
binder (Bond C), which could be supplied in single pack. The laboratory work reveals that this
formulation is better resistant to slag attack (Fig. 9) compare well to the earlier two - pack system.
Improvement in slag attack resistance for the latter two formulations is attributable to the lack of
formation low melting Mellilite, a solid solution of Ghelenite and Akermanite, in the matrix.

Fig. 6. Slag interaction study of BF trough castable, with wider particle size distribution of SiC, based on Bond
A and BF Slag

255
Tehran International Conference on Refractories, 4-6 May 2004

Fig. 7. Slag interaction study of BF Trough castable, with narrow particle size distribution of SiC, based on bond
A and BF Slag

Fig. 8. Slag interaction study of BF Trough castable based on bond B and BF Slag

Fig. 9. Slag interaction study of BF Trough castable based on bond C and BF Slag

256
I. N. Chakraborty, S. Bose, A. Mukherjee and D. K. Singh

6. ROLE OF DESIGN AND OPERATION ON BFT WEAR

The design of the trough is an extremely important parameter. If BFT is too short slag iron separation
would be poor, high wear would be expected in the skimmer block area and slag temperature would be
high, which is detrimental for secondary trough. On the other hand if it is too long, as a consequence
BFT expansion and corrosion would be high. BFT length, thus, must be optimum. Furthermore, if the
trough is too narrow, the flow would be turbulent and the lining temperature would be high, both lead
to increased corrosion of refractory lining. Conversely, if it is too wide, the thermal losses are unduly
high and larger quantity of refractory is required for the initial lining. The following additional
parameters related to trough design and other operational parameters influence refractory lining wear
in BFT:
• Slope of the trough, i. e. pooling or non – pooling trough, which determine the level of molten
mass between the taps. The runner with a slope of 1 to 1.5%, i. e. pooling type, is constantly
full of iron and thus, shows little wear at BFT bottom. The runner slope also controls the
velocity of the iron / slag stream
• Usage of main trough cover
• Whether the height of the slag / iron / oxygen interface is varied by altering slag runner heel.
• Slag composition – primarily its MgO content since it determines the slag viscosity. Slag
richer in MgO is less viscous and therefore more easily absorbed into the Castable than those
lower in MgO.
• The hot metal with low viscosity penetrates the cracks and pores of trough refractory lining
and causes spalling.
• A large taphole angle causes a short and deep turbulence and consequently increases wear at
the sides of the trough in the region of the impact areas.
• The width of the trough has an important effect on the advanced wear at the sides. With
increased trough width the specific consumption decreases in similar conditions.
As expected the reaction zone is located nearer to the hot side. The cooling of the main runner causes
dissipation of the heat from the refractory lining and has as important an effect on the wear
mechanisms by controlling the chemical attack. It is observed that the wear of the non -cooled
sidewall section is significantly higher than the cooled section. Cooling, thus, reduces the oxidation
and de-carburisation of certain components in BFT lining material. The refractory materials used for
lining the safety and back-up linings should, thus, have a high thermal conductivity. Following are the
additional features of good safety and back – up linings:
• High resistance to oxidation
• Good resistance to corrosion and impregnation
• High residual mechanical strength, which resist mechanical abuse at wrecking
• High resistance to thermal shock and high stability
• The back layer be strong and resistant to oxidation.
• Conversely the working lignin should be easily cleaned out before relining and should bear
strains.
When there is no way to change the design of the cast house or the process, the only options left for
better BFT performance are:
• Improve corrosion and erosion resistance of the product
• Monitor type and frequency of repairs
• Take utmost care during installation of the refractory

7. ROLE OF INSTALLATION AND PERFORMANCE

Goal during the installation of BFT refractory, castable or ramming mass, is to ensure that the installed
mass has the lowest possible porosity. All precautions and actions should be taken to achieve this goal.
Porosity control probably is the most important criterion, as has been seen earlier section. Permeability

257
Tehran International Conference on Refractories, 4-6 May 2004

of the installed mass, other than chemistry, determines BFT material performance during use. For
castables it must be ensured that the high intensity mixer of adequate capacities are available for the
avoidance of lamination. Subsequent to installation heating of the installed mass should be carried out
so that all the volatiles are removed, especially for castables, without causing any structural damage.
With correctly installed material, our product is customized based on the slag chemistry and other
operating conditions, the specific consumption of the castable and ramming mass are in the range of
0.27 – 0.39 and 0.35 – 0.45 kgs / MT of HM, respectively. Typical wear pattern of BFT maintained by
us for 1 mT BF is illustrated in Fig. 10.

Fig. 10. Wear Rate Pattern of different section of BF Trough when in use

8. CONCLUSIONS
Based on our experience in BFT management in India, we conclude
• BFT lining castable yield better performance compared to its ramming mass counterpart
owing to better pore structure.
• Fused aluminous aggregates tend to perform better in these formulations owing to resistance
to dissolution in BF slag by virtue of CA6 – protective layer formation.
• The gradation of aggregates plays a definite role in BFT wear pattern. Broader distribution
appears to yield better results.
• Minor constituents in the matrix should be so adjusted that no low melting compounds form
during use. This has been identified as the major cause of BFT lining wear.
• Appropriately designed BFT lining material should have high resistance to oxidation as well
as impregnation / corrosion and thermal shock.
• Safety and back up lining play an important role in the performance of BFT working lining.
• Improvements in material and installation quality are the only ways to improve BFT material
performance since trough design alteration in majority of the cases is not possible and
refractory suppliers have little control over BF operation.

258
 I. N. Chakraborty, S. Bose, A. Mukherjee and D. K. Singh

REFERENCES

1. M. Gondolf and J. P. Radal, “Numerical Modelling of the wear in Blast Furnace Main Trough”,
Proceedings Unitecr 2001 Mexico.
2. J. Niu et al, “ Effect of graphite grains on properties of castable for Blast Furnace Trough”, J.
Tech Assoc. of Refractories Japan, 22 (3) 219 – 222 (2002).
3. Subrata Banerjee, “Versatility of gel bonded castable / pumpable Refractories”, Refractory
Application and News, pp 1 – 3, Jan – Feb (2001).

259

View publication stats