Buy

Yazawa International Symposium
METALLURGICAL AND MA TERlALS PROCESSING: PRINCIPLES AND TECllNOLOGlES
VoIumeU: HIGH·TEMPERATURE METALS PRODUCTION
Edited by F. Kongoli, K.ltagaki. C. Yamauchi and H.Y. Sohn TMS (The Minerals, Metals & Mattrials Society), 1003
APPLICATION OF THE POROUS PLUG SYSTEM IN THE ANODE
FURNACE AT ONSAN SMELTER
Sang-Su Lee
l
, Back-Sang Kim\ Sei-Rhim Choi
l
ILG-Nikko Copper, Inc.;
Daejung-Ri 70, Ulju-Gun, Onsan-Eup; Uisan City, 689-892, Korea
Abstract
This paper shows the application of the porous plug system in the Mitsubishi continuous copper
smelting and converting process. This system has already been adopted and utilized in the steel
industry to enhance productivity. Recently, some copper smelters have adopted this system but
the results have not been satisfactory. The porous plug system was installed at the anode
furnace in December 2001 to reduce the oxidation time and decrease the oil consumption. A
good effect was that the nitrogen stirring of the melt through porous plugs improved the melt
heat transfer. As the result of the enhancing heat transfer the oil consumption and the build.up
inside the furnace were actually decreased. But until now the oxidation time has not been
sufficiently reduced. In the future, further research and tests are needed in order to increase the
productivity and save costs.
447
Introduction
The porous plug was used for the first time in iron manufacture industry. In early 1970s, the
continuous steel casters appeared in the steel industry and the porous plug was devised as a
necessity to this process. In the bottom/top of the ladle that moved the molten steel, the porous
plugs and lances were installed and an inert gas (nitrogen, argon) was injected to the molten
steel, which makes the molten steel bubble. The purpose of the inert gas purging is to obtain a
chemical homogeneity, reduce the thermal difference and increase the steel purity. All of these
actions contributed to the continuous steel casters.
In late 1970s, the ARBED steel company (Luxembourg) installed the porous plug to the bottom
of BOFs (Basic Oxygen Furnace) and succeeded in reducing the steelmaking time in the
converter. In 1984, Inco made an experiment on the application of the porous plug purging
system in the converting oxidation step and found out the relation between the oxygen
efficiency and the purging status during the oxidation step. Consequently they installed two
full-scale porous plug and oxygen lances. In 1990, in order to increase the agitation effects,
Reynolds Metal Company (USA, Alabama) installed the porous plug at six of its Aluminum
smelters.
Simply mentioned, the porous plug can be applicable to all kind of metal refining processes.
Consequently LG-NIKKO Inc. installed the porous plug system in the Anode furnace of MI
process in Nov. 2001. During the ten-month operation LG-NIKKO was able to verify the
effects of the porous plug purging system.
Outokumpu Process vs. Mitsubishi Process
At the moment, the most widely used smelting process is the Outokumpu Flash Process. Some
of its features are the following. The flash furnace is operated with highly 02-enriched air blast
and a minimum of fossil fuel. Most of the energy for heating and melting comes from Fe and S
oxidation. An adjustment of the 02/concentrate and flux/concentrate ratios at the input affects
the matte and slag compositions. An adjustment of the fossil fuel combustion rate and the
N2/02 ratio at the input blast can control the product temperature. The converter type that most
of the Outokumpu process smelters have adopted is the PS (Peirce-Smith converter). LG-
NIKKO Onsan plant also selected the PS converter. The flash furnace matte (Cu=62.S%) and
the electric furnace matte (Cu=68.S%) are poured into the PS converter by ladle. After copper
blowing, the sulfur contents decreases to 0.02-0.03%. This is explained with the batch
character of the PS-converter. As given below in Figure I CmS starts to oxidize at the point (a).
(a) CmS + x02.. .. CU2S1-x+ XS02
Then, the sulfur in the white metal (CuzS) is lowered to 19.6% at point (b). Continuous blowing
leads to point (c). From Point (b) to point (c), the blister copper (I % Sulfur) and the molten
white metal are mixed. However, due to their different structure the two liquid phases are
immiscible and the denser blister copper sinks to the bottom. At the vicinity of point (c) greater
care should be taken in order to prevent over oxidation as CmO. When copper starts to oxidize,
448
the flame color of the tuyeres changes to green and the operators have to prepare to stop the
copper blowing. In the Onsan plant, the decision of Cu bowing endpoint depends on the
temperature of off gas. When the off gas temperature goes down, operators visually check the
molten blister copper and decide to stop copper blowing.
Molten 1400
'blister' copper
"eu"
1000
Motten 'bUster' COf!)pEtr' ..... moHen 'white metal'
Molten
"'white metal'

4
Cu
8 12
Mass % S
20
Cu.p
Fig I. Cu-S Equilibrium Phase Diagram Showing Copper Making Reaction Path
(a,b,c,d, 1200 C) (Sharma, Chang, 1980)
Mitsubishi process consists of three furnaces (Smelting, Slag cleaning and Converting Furnace)
connected to each other by launders. Therefore, the operation of one furnace affects the next
one continuously. Dried feed (Cu 31 is blown to the Smelting Furnace along with the 02
- enriched air (50-55 vol% 02). The lance consists of two concentric pipes inserted through the
furnace roof. In the CL-furnace the matte and slag are separated due to their difference in the
specific gravity. The matte of about 68.5% Cu flows into the C-furnace. In the C-furnace, the
02-enriched air (33-36%) and limestone is blown through lances.
One of the features of the Mitsubishi process is that the sulfur content in the blister is as high as
0.65-0.7%, while blister sulfur content in the PS converter is about 0.02-0.03%. This is
explained with the fact that the PS converter is batch type operation while the C-furnace is a
continuous one. In Onsan, the slag/copper blowing is not divided in the C-Furnace operation.
The C-slag and blister flow out continuously. The sulfur content of the blister can be gradually
reduced by further blowing in the C- Furnace but this way the Cu content in tbe C-slag also
increases gradually. Consequently the ideal sulfur content in the blister is 0.65-0.7%.
449
Table I. Blister copper composition in the C-Fumace and PS converter
i
.%
Blister Copper
C-Fumace PS converter
Cu 98.24 98.60
I
S 0.840 0.025
Pb 0.292 0.244
Zn 0.005 0.003
I
Fe 0.002 0.039
In conclusion, Mitsubishi process has longer oxidation time than Outokumpu process. If there
are some methods that can reduce the oxidation time, the productivity of Mitsubishi process can
be certainly increased. One of those probable methods is the porous plug purging system in the
anode furnace.
Enlargement of Mitsubishi Process, Onsan plant #2
From April, 10 to May, 10, 2002, Onsan Mitsubishi plant was shut down for a 2-year periodic
relining. At the same period, Onsan Mitsubishi plant was enlarged and the annual anode
production increased from 210,000 to 260,000 ton. There are three important items of
enlargement process. First, the feeding system was changed. Installing a new S-feeding tank
and the change of the S-lance system were major modifications. So much more dried feed can
be put to the S-Furnace.
Before After
Fig 2. Changing S-furnace Lance system during the 2002 enlargement.
Second, the new anode furnace (#3 anode furnace) was installed in November 2001. With the
possible positive effects of the porous plug the increased blisters production
could be treated by only two anode furnaces consequently the porous plug system was installed
at #3 anode furnace.
Third, the balloon dust was increased. That is an off gas line after WHB (Waste Heat Boiler) to
ESP.
After the enlargement, the changed operation factors are shown in the table below.
450
Table 2. Mitsubishi Process operation factors in the Onsan plant #2.
Necessity of the Porous Plug Purging System
As previously mentioned LG-NIKKO Inc. installed the porous plug system for 2 major reasons.
First, as one offeatures of Mit sub is hi process, the sulfur content of the blister copper (0.84%) is
much higher than PS converter (0.025%). Consequently the anode furnace oxidation time of the
Mitsubishi process is longer. The Mitsubishi process in Onsan needs just about 4-4.5 hrslbatch
oxidation time (about 400Tons blister is treated per batch).
Second, because of the enlargement of Mitsubishi process in the Onsan plant, the annual anode
production increased from 210,000 ton to 260,000 ton and therefore an additional anode
furnace was installed. However, if oxidation time could be reduced, only two anode furnaces
could have been enough even for the enlarged capacity. However, during the last 6 months the
anticipated effects of the porous plug on the oxidation time has not been visibly shown.
Hypotheses on the Effects of the Porous Plug in the Anode Furnace
First, the oxidation time would be reduced. Because of the porous plug purging effects the
blister copper in the anode furnace would be more homogeneous compared to that one without
the purging effect.
Second, because of the thermal uniformity the energy consumption would be reduced. In the
Onsan Plant Mitsubishi process the Anode Furnace uses low sulfur B-C oil.
Third, the bottom and end plate side build up would be reduced. Such a build-up occurrence has
some relation with the thermal uniformity. If the buildup of interior anode furnace is reduced,
more blister copper could be poured into the anode furnace.
During the past six months only two anode furnaces has been in operation. The Anode furnace
#3 was installed in Nov. 2001. At that time the Anode Furnace #2 was repairing. When the
shutdown of the Mitsubishi process ended the relining work of Anode furnace #2 had finished,
but the anode furnace #1 repair work started again, as illustrated in Figure 3 below.
451
(
'AF#3
Just in AF#3, Porous
" plug was installoo:/
AF# 1 rUdAIOg.
·AF#2 start tor.palr
AF#3 I. stalled & rUAolog
' ..... :c •..........
AF#2 AF#l
,AptI.2002
May;2002·
Mitsubisfli .start to r.P9.1r
Process
Relining I:'X.-."'·ril-,OO---------'..,
Fig 3. The anode furnace running situation in the Onsan Mitsubishi process.
In order to find out the effects of the porous plug the data were compared between furnaces in
the following way: AF#3 vs AF#2 and AF#3 vs AF# I. While the energy consumption as well
as the formation of the build up was reduced during above period, the oxidation time was not
visibly reduced.
Comparison of the Energy Consumption
In order to find the extra effects of the porous plug a comparison of the consumption of B-C oil
was suitable since the total state of the anode furnace thermal homogeneity was in a functional
relation with the main burner B-C consumption. These effects are shown in Table 3 below.
Also, under normal operation, the consumption of B-C Oil is shown.
600-620 Literlhr
The graph in Figures 4 and 5 below is the B-C consumption at the main burner of the anode
furnace #1, 2, 3. The Oil consumption per batch was divided according to the casting amounts.
As shown in graph the effect of the porous plug in the energy consumption is visible. However
it was not proved that the effects on the energy consumption are from the porous plug or a new
main burner type. While the main burners installed in AF#I,2 were hand-operated the one
installed at AF#3 is fully automatic. This has to be proven in the future.
I Tuyere consump. depend on reduction time, oxygen contents of after oxidation blister copper.
452
17
16
. ~ ___ , __ ~ __ . iBAF#l usage L
IDA.F#3 usage I
15
I:i 14
m
"8 13
c
~
~ 12 1-
m
u
'rl 11
.,
10
batch
Fig 4. Comparison B-C Oil consumption AF #1 (w/o P.P) VS AF#3 (wlP.P)
[Liter/Anode Ton, Batch], (February, 2002)
12
11
- 1-
c-- ---
- - - ,--
N
.,.
'"
.,
"
N
.,.
'"
.,
"
N
.,.
\0
.,
"
~ ~ ~
.,
"
N ..
'"
.,
"
N ..
~
ro
"
" " " "
,-;
'i 'i
,-; ,-;
N
';' ';' ';'
N M
'( ';'
.. ..
.,. .,.
"' "' "'
u;
'"
~
,
~
,
~
,
~ ~
, , ,
.... ...
,.. ,..
...
,.. ,.. ,..
r- r- r- .... r-- .... .... r- r- .... .... .... .... r- .... .... ....
batch
Fig 5. Comparison B-C Oil consumption AF #2 (w/o P.P) VS AF#3 (wlP.P)
[Liter/Anode Ton, Batch], (July, 2002)
Reduced Buildup Inside the Anode Furnace.
N
'"
~
There were no build up inside ofthe AF #3 as shown in Fig. 6. In AF # I and 2 the bottom build
up was considerable in each end plate. Especially below the main burner hole a lot of accretion
was formed. These accretions reduce anode furnace capacity.
453
(a) ( b)
Fig 6. End Plate of AF #3.
(a) uptake side (b) Main burner side
Effect of the Porous Plug on the Reducing Oxidation
Fig 7 is a typical off gas volume change. When the oxidation starts the volume sharply
increases and afterwards the down stream 'slope is not steep. X-axis is the oxidation time. Due
to the chemical and thermal homogeneity of the blister copper created by the porous plug the
oxidation time can be additionally shortened. After that the slope becomes steeper. In the
Mitsubishi process, the AF oxidation time changes according to the matte grade, C-fumace
blowing and sulfur content in the blister copper. The Cu content in the blister copper is grasped
by the Cu contents in the C-slag. Around 13.5 % of Cu content in the C-slag is appropriate
although it is very difficult to maintain this value. Consequently the Cu and S contents of the
blister copper fluctuate and due to this feature of the Mitsubishi process the anode furnace
operation time goes up and down also.
vol ofAF uptake
[Nm 3jlllj
45,000
40,000
35,000
30,000
25,000
20,000
15,000
10,000
5,000
a
1 3 5 7
7-06 batch AF#3
9 11 13 15 17 19 21 23 25 27
10 m htue
Fig 7. Off gas volume change of the anode furnace uptake.
454
400
~
AF#3 AF#1
AVG.266 AVG.2S3
350
min/400ton min/400ton
... - .... - ~ = j
•
"" 300
g
~
I
g
250 -
::::
I
~
I
I!J
.il
200
e
150
I
!
! 1 :
100
) j'J
.1,
M ...
'" '"
U1 M ...
'" '"
u;
.. N
~
..
~
::;
'"
00 ..
'"
a M M
'"
M M .. ..
'" '"
M
'"
.. ..
'"
,
r:.
, , , , , , , , , , , ,
'"
,
'"
, ,
'"
N N
'"
N
'"
N
'"
N N
'"
N N N
'"
N
batch
Fig 8. Comparison of the Oxidation Time (w/P.P AF #3, wlo P.P. AF #1, February, 2002)
450 ...---------1
400 I ~ n , ~ ~ ~ ___ ~
g 300 -
~
~ 250
~
'g 200
150
)00
M
a
<-
'"
a
...
'"
M
'" a ... M
<- ~ I"-
M ... M
:!1
Z
N M
;..
... I"-
AF#3
AVG.276
min/400ton
M
~
'"
M ...
X
..
'" '"
,
.... ... ... ...
M
'"
... c
,
... ....
batch
AF#3
AVG.303
min/400ton
'"
Z
..
'" '"
0 M ...
'"
;.. ;..
... ...
::: '" '"
N
'"
0 ..
'"
N
M M
X
'i '" '"
~
'f
,
~ .... .... ... ... ....
<-
Fig 9. Comparison of the Oxidation Time (wlP.P. AF #3, wlo P.P. AF#2, July, 2002)
The graphs in Figures 8 and 9 above shows that the reduction of the oxidation step is not clearly
identified. On the contrary, the average oxidation time of AF #1 was shorter than AF #3.
Operators of the plant agreed that the reduction of the consumption and the thermal equality
eould be clearly observed. However, they were not sure whether the reduction of the oxidation
time was successful. In the Mitsubishi process, the oxidation endpoint is visually determined by
the operator. Of course the tendency of the volume of the uptake off gas and the SOx
concentration in the off gas could be referred to but the final decision depends on the operator's
experience. The average operator's experience is about 12 -18 years. In the case of an
emergency situation their experience is very helpful. But for normal situation, because of their
conservative inclination, they could not adjust easily themselves to the new equipment like the
porous plug system. It was presumed that the real oxidation might be finished before the
operator's end point. Therefore, more data gathering and studying might be needed.
455
Ladle Sample from Anode Furnace
Many samples were analyzed since information of the change of sulfur and oxygen contents
was needed.
On 9
th
February of 2002, 3 samples of each step were taken for analysis. Analyses were
performed by LG-NIKKO analytical departmeflts in Boehler Austria and in NA-Hamburg
Germany. The data are given below,
Table 4. Contents of sulfur and oxygen during oxidation
% I 18:00 20:40 21:20 21 :40 22:20 22:40
I
23:20
LG-NIKKO
02 I 0.077 0.082 0,152 0.19 0.208 0.311
I
0.351
S 0.127 0.055
0,07
0.063 0.038 0.038 0.025
NA Hamburg I 02 0.33 0.14 0.25 0.28 0.3 0.35
I
0.49
Germany S 0.12 0.11 0.05 0.04 0,04 0.04
I
0.02
Boehler
L 02
0.17 0.119 0.225 0.23 0.305 0.37 0.46
Austria
IS 0.21 0.04 0.02 0.01 0.02 0.02
I
0.01
0.6
0.5
'*'
0.4
III
.w
>::
0.3
Q)
.w
>::
0
0.2 .
U
0.1
o .
Fig 10. The tendency of sulfur and oxygen content during oxidation
Fig. 10 gives the tendency of sulfur and oxygen content during oxidation. Viewing the tendency
of sulfur diminution at 21:40 the oxidation could be stopped. But the real oxidation was stopped
at 23:30. The operators at Onsan smelter believed that there were problems of bad shape of the
cast anode due to the sulfur in blister copper. Consequently they oxidized the blister copper
more but it turned out to be excessive.
456
The Plan for the Output of the Porous Plug
The main problem that the oxidation time couldn't be reduced appeared to be the
misunderstanding of our operators as above, It appears that a more objective method for the
determination of the oxidation endpoint is required. Therefore, it appears to be necessary to
have a system to measure sulfur and oxygen contents on the spot. Visual checking is not
enough.
It is very difficult to gather data from the Mitsubishi process because of its continuous
character. In order to increase the efficiency of porous plug purging system invariable contents
of Cu and S in the blister copper are indispensable and, consequently, the C furnace operation
should be stable.
Conclusions
When the melting capacity of Mitsubishi process of Dnsan was enlarged in May 2002, an anode
furnace with porous plug system was introduced. It was assumed that the reduction of the
oxidation time might lead to the operation of only 2 anode furnaces with the introduction of
porous plug system, The porous plug system reduced the energy consumption and the accretion
but its effect on the oxidation time could not be clearly seen. It can be said that in order to
increase the porous plug purging effects a change in the way of determining the endpoint of the
oxidation time is required.
References
1. AX. Biswas, W.G. Davenport, "Extractive Metallurgy of Copper" Third edition 1994. Pl94
-202.
2. AJ. Rigby, Michael D. Lanyi "Porous plug in molten copper production and refining"
(Paper presented at Copper99-Cobre99 International Conference, \999)
457