.
The metallurgy of tin smelting In a submerged-
arc furnace
by H. A. UYS*, Ing. Metallurgy (Pretoria) (Visitor)
SYNOPSIS
The control of tin smelting demands a knowledge of slag characteristics and of the equilibrium reaction that
governs the process. These factors are explained by reference to the tin-smelting operation at Iscor, Vanderbijlpark.
In the primary cycle, it is important to obtain the highest possible FeO content in the slag. In the secondary cycle,
a primary slag high in FeO gives a hardhead (iron-tin alloy) high in iron, which is essential if a high ratio of FeO to
SnO is to be obtained in the secondary slag. This determines the economy of the smelting operation: the iron that
is introduced into the system from the ore concentrate must go out in the secondary slag. Additions of lime are
necessary in the secondary cycle but must be limited if the volume of slag is to be kept Iow.
SAMEVATTING
Die beheer van tinsmelting vereis 'n kennis van slakeienskappe en van die ewewigsreaksie wat die proses beheer.
Hierdie faktore word verduidelik met betrekking tot die tinsmeltbewerking by Yskor, Vanderbijlpark.
Dit is belangrik om in die primere siklus die hoogste moontlike FeO-inhoud in die slak te kry. In die sekondere
siklus gee 'n primere slak met 'n hoe FeO-inhoud 'n yster-tinlegering met 'n hoe ysterinhoud wat noodsaaklik is om
'n hoe verhouding van FeO tot SnO in die sekondere slak te kry. Dit bepaal die ekonomie van die smeltbewerking:
die yster wat uit die ertskonsentraat by die stelsel gevoeg word, moet in die sekondere slak uitgaan. Die byvoeging
van kalk is in die sekondere siklus nodig maar moet be perk word as die volume van die slak laag gehou moet word.
Introduction FeO and SnO act as fluxes in the primary slag and
lend a basic character to it. This requires a certain
At the Iscor (Vanderbijlpark) works, tin is smelted
minimum amount of iron in the hardhead. The relation-
from a cassiterite concentrate in a submerged-arc
ship between the SnO and the FeO in the primary slag
furnace with a power input of 350 kV A, the electrodes
is shown in Fig. 3, which indicates the ideal operating
being amplidyne controlled. The sidewall of the furnace
range with 30 to 40 per cent FeO.
is lined with fireclay bricks and the hearth with carbon
blocks. The waste gas, which is sucked off round the
perimeter of the furnace, is collected in a bag filter. The The Secondary Cycle
layout of the plant is shown in Fig. 1.
The slag arising from the primary cycle is smelted in
the secondary cycle with coal char as the reducing
The Process
agent and burnt lime as the flux. The FeO and SnO,
The tin-smelting operation is divided into two cycles which acted as fluxes in the primary slag, are now
as shown in the flow diagram of Fig. 2. reduced to fairly low levels.
After tapping, which is done every 2 hours, the slag is
The Primary Cycle decanted and scooped off, the liquid hardhead being
granulated in a strong stream of water. This hardhead
Tin is smelted from the cassiterite concentrate by the
is used again during the primary cycle of the following
reducing action of coal char, and iron is smelted from
campaign.
the hardhead (iron-tin alloy) produced in the secondary
Despite the lime that was added, which must be
cycle of the previous campaign. The reduction is strictly
limited to keep the volume of slag low, the slag acquires
controlled to produce the best slag for smelting in the
an acid character. A considerable portion of the tin
secondary cycle. In the primary slag, the activity of
that is lost in the secondary slag is in the form of trapped
SnO is higher than that of FeO, but the ratio of FeO to
metal droplets, and the viscosity of the slag is important
SnO can be increased by an increase of the reducing
in limiting losses. In practice, the last slag that is scooped
agent. However, over-reduction leads to an increase in
off is returned to the furnace so that such losses are
the iron content of the tin. The control criterion is the
minimized.
formation of a reasonable amount (5 to 10 per cent) of
The secondary slag, which has a tin content of 2 to 3
hardhead in the ladle on cooling.
per cent, is sold for its (Nb, Ta)20s content, which de-
After tapping, which is done every 1,5 hours, the ladle
pends on the source of the concentrate and the percent-
is deslagged and the tin is allowed to cool in the ladle
age concentration; these determine the volume of the
until most of the dissolved iron has crystallized out as
slag. Fig. 4 shows the slag as a function of the tin in the
hardhead, which is then scooped off. Further refining is
concentrate.
done in a heated kettle at a temperature just above the
When the concentrate is leaner, it is important to
melting-point of tin: air is bubbled through until the
increase the ratio of iron to tin so that the most
iron content has been reduced to 0,1 per cent.
economical ratio of FeO to SnO can be attained in the
*South African Iron & Steel Industrial Corp. Ltd (Iscor), secondary slag. Fig. 5 shows the relationship between the
Vanderbijlpark, Transvaal. FeO and the SnO in the secondary slag from a typical
JOURNAL OF THE SOUTH AFRICAN INSTITUTE QF MINING ANt>MH AI..LI,IRGY JANUARY.1977 121
STORE
TRANS-
FORMER
FURNACE
REFINING
KETTLE BAGFILTER
D
MOULDS
COAL CHAR BUNKER
COOLING AREA FOR
c==>
CONCENTRATE BUNKER
0 LADLESTIN
SCALE]
CONCENTRATE BUNKER t
HARDHEAD
CRUSH~R
PRIMARY SLAG I
SECONDARY
SLAG
DUMP
Fig. I-Layout of the tin-smelting plant at Iscor, Vanderbljlpark
122 JANUARY 1977 JOURNAL OF THE SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY
CASSITERITE
COAL CHAR PRIMARY CYCLE
HARDHEAD
COAL CHAR ~
SECONDARY CYCLE
LIME
CASTING OF ELECTRODES
FOR ELECTROLYTIC
TINNING LINE
SALES OF SLAG
FOR ITS
(Nb,Ta)20S
VALUE
MIXING OF HARDHEAD FO
USE IN THE FOLLOWING
CAMPAIGN
Fig. 2-Flow diagram for the tlnsmelter at Iscor, Vanderbijlpark
350
45
z
40 0300
I-
8nO% '-...
~
35 ~'250
b8
tI
...
30 fI)
200
30
55 60 65
Fig. 3- The relationship between SnO and FeO In the 8n in Concehtrate %
prl mary slag
Fig. 4- The secondary slag as a function of the tin In the
concentrate
JOURNAL OF THE SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY JANUARY 1977 123
20
. 15
reO %
10
Fig. 5- The relationship between FeO and SnO in the
secondary slag
CaO 40 60 70 80 90 FeO
Fig. 6-Composition of secondary slags in the ternary system CaO-FeO-SIO
124 JANUARY 1977 JOURNAL OF THE SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY
campaign. The importance of controlling the ratio of By linear regression analysis, the following relation-
FeO to SnO in the primary cycle becomes clear when the ship was obtained from actual production data, giving
aim is to achieve a high ratio of iron to tin in the hard- a correlation coefficient of 0,68:
head. K=25,7(CaOjSiO2)-4,34.
In the ternary phase diagram of the system The ratio of FeO to SnO in the secondary slag is thus
CaO-FeO-SiO2 (Fig. 6), typical compositions of the dependent on both the ratio of iron to tin in the hard-
secondary slag are indicated around the 1300°C liquidus head and the ratio of CaO to SiO2 in the slag.
line. The ratio of CaO to SiO2 plays an important role
in determining the activity of the FeO and SnO in the
Conclusion
slag.
The secondary cycle can be represented by the equili- For efficient tin smelting, it is necessary to maintain
brium reaction between the Fe and the Sn in the metal a high ratio of FeO to SnO in the secondary slag and
and the FeO and SnO in the slag: so ensure that iron is not accumulated in the system.
Fe+SnO ~ FeO+Sn The best solution is to achieve a high ratio of FeO to
FeOjSnO = K(FejSn). SnO in the primary slag. This will ensure a high ratio
The value of the equilibrium constant K is dependent of iron to tin in the hardhead of the secondary cycle,
on the ratio of CaO to SiO2 in the slag, as shown in which in turn will give a high ratio of FeO to SnO in the
Fig. 7. secondary slag.
20
15
K
10
,1
Fig. 7-The relationship between the reaction constant K in the equation FeOjSnO=K
(FejSn) and the ratio of CaO to SiC in the secondary slag
,
')
JOURNAL OF THE SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY JANUARY 1977 125
