REFRACTORY DESIGN EFFECT ON SUBMERGED ARC FURNACE LIFE AND
PRODUCTION PERFORMANCE – THE ZIMBABWE ALLOYS EXPERIENCE
J. Chirasha and N.R. Shoko
Zimbabwe Alloys International P.O. Box 530, Gweru-Zimbabwe
E-mail:
[email protected] and
[email protected]ABSTRACT
Refractory design remains the most fundamental factor in furnace construction. Refractory design details and
proven technology consideration is of paramount importance to come up with an optimum furnace life and
to maintain productivity in the life span of the furnace lining.
The failure to produce a technically or operationally durable refractory design lining could give rise to pre-
mature deterioration of the refractory lining and furnace shell.
The premature failure of the furnace lining will lead to extensive heat losses, furnace bath shift, change in
taphole orientation and alteration of the Pitch Circle Diameter. A combination of these were experienced at
Zimbabwe Alloys’ new high carbon ferrochrome furnace.
The Zimbabwe Alloys experience has shown an excellent furnace performance in the first two years of op-
eration although signs of an anomaly showed within the first six months of the new furnace. Production
slowed down due to induced problems from refractory failure causing furnace distortions impacting on the
availability of the furnace and smelting capacity.
This paper presents the lessons learnt, measures taken to deal with the premature failure and constraints
within which the situation was managed.
1. INTRODUCTION
Zimbabwe Alloys has been producing ferrochrome for the past 53 years and has produced 5 different ferro-
chrome and other alloys amounting to more than three million tonnes.
The furnace under discussion was originally commissioned as a High Carbon Ferrochrome production unit
in 1974.
High Carbon Ferrochrome production was discontinued in 1983 due to market constraints. The furnace was
then converted to producing Ferrosilicon Chrome.
Ferrosilicon chrome production continued until the year 2000. The market changes saw the 31MVA Fur-
nace converted from Ferrosilicon Chrome to High Carbon Ferrochrome in 2000 as the latter product had bet-
ter margins of return.
The furnace was operated without problems for the first six months of operation. Movements at the back
of the furnace between the shell and the base plate initiated the distortions that later posed challenges related
to furnace availability, production, efficiency and safety.
This paper describes the effect of refractory design on furnace operation, efficiency and details lining design
shortcomings as observed by the furnace lining behaviour during the furnace production life span.
Refractory Design Effect ... 729
2. LINING DESIGN
The HCFeCr refractory lining design is fairly straight forward because theoretical modeling for these linings
are readily available.
Technical suitability of the lining has been proven in most operations but there is need to pay attention to
detail as the building of the furnace is taking place. The detail lining design omissions can give rise to serious
operational and efficiency problems impacting on economic production.
Zimbabwe Alloys designed and lined its HCFeCr Furnace as shown in Figure 1.0. It was later found out
that expansion for the entire lining of the furnace was not provided for. Bricks were lined directly against the
shell. More over the brick to brick expansion was not adequate allowed for from the base of the furnace to
the furnace side walls.
10500mm
4993mm
Cold
ram
paste
Cold ram paste Cold ram paste
Carbon Block
KEY
85% Alumina Tongue & Groove 85% Alumina Brick 40% Alumina Firebrick
40% Alumina Brick 60% Alumina Carbon block
Precast Taphole block 60% Alumina Brick Hi-Strength castable
90% Alumina 90% Alumina
Figure 1: Refractory Lining Section
The base of the furnace was levelled off using a castable material 1300 in order to facilitate the 60% Alu-
mina brick laying which started at the bottom.
A heavier duty 85% Alumina brick followed the 60% Alumina and this also was used predominantly around
the taphole area. A Tongue and Grove 85% Alumina brick was put on top of ordinary 85% Alumina brick.
The detailed lining design is shown in Figure 1.
Production started in July 2000 after the furnace was converted and lined for HCFeCr production.
3. CAMPAIGN STATISTICS
3.1. Production Attainment And Efficiencies
Production improved from the time the Furnace was commissioned in August 2000. The 35000 tonnes mark
was attained in the first full year of operation and the upward trend continued in 2002 with 45 000 tonnes of
production.
In 2003 production started deteriorating and bottomed at 32000 tonnes in 2005, a 30% reduction from 2002
production although the furnace running time was constant.
Figure 2 below shows the production trend from the year of commissioning to December 2005.
730 INFACON XI
45000
40000
35000
Tonnes
30000
25000
20000
15000
2000 2001 2002 2003 2004 2005
Ye ar
Figure 2: HCFeCr Production
The specific power consumption (KWh/t) which is a direct measure of furnace efficiency utilization of pow-
er improved from 4500 kwh/t to the lowest figure of 3400 kwh/t between 2002 and 2003. A value of power
utilization of this magnitude could qualify to be one of the best in the convertional HCFeCr process.
Figure 3 below shows the change in specific power consumption in the period of five years of furnace op-
eration.
5000
4500
KWh/t
4000
3500
3000
2000 2001 2002 2003 2004 2005
Ye ar
Figure 3: HCFeCr Specific Power Consumption
Chromium Unit Recovery which is the business of ferrochrome production also followed a favourable trend
in tandem with both production and power utilization. A recovery of 89% was achieved in 2002 as the best
performance of this furnace during its stable period.
Figure 4 below shows the recoveries recorded in the period under review. A visible deterioration of recov-
eries from 2003 to 2005 showing poor furnace performance.
90
80
%
70
60
2000 2001 2002 2003 2004 2005
Ye ar
Figure 4: Chromium Recovery
Refractory Design Effect ... 731
3.2. Production Milestones
A few major land marks were achieved when the Furnace was at peak performance.
• A daily production of 210.8 tonnes per day was attained in 2002 breaking another one of 202.4 tonnes
per day in 2001.
This was a mega achievement on a furnace rated 31MVA whose maximum theoretical production is
143 tonnes per day.
• A weekly record production of 171.1 tonnes per day was attained with an all time monthly record of
146.7 tonnes per day, better than the designed production capacity.
4. FURNACE PERFORMANCE AND PHYSICAL APPREARANCE CHANGES
Figure 2 to 4 indicated a significant unfavourable shift on production, kwh/t and recovery from 2003 to 2005.
This was a major signal flagging on the furnace performance to the effect that something was still not original
on the furnace.
4.1. Shell Distortions
After the first 18 months of operation, the furnace base started lifting to about 300-400mm from the horizontal
support beams.
The front section of the shell as an effect of the force on the base movement displaced about 600mm out-
wards together with tap launder. A view of the furnace from the top showing the new shell position as com-
pared to the original position is shown in Figure 5.
Original Refractory
position
Original shell
diameter
Shell position after movement
Figure 5: Furnace plan showing relative movement
The taphole lifted 400mm up and 200mm sideways to the right. Weak zones were created and the furnace
shell started cracking vertically from the base upwards on either side of the taphole.
4.1.1. Immediate action
It was suggested to cut-off the bolts joining the furnace flange and the base plates to release the tension. After
this, the base plate got to its original position but the furnace flange remained up leaving a 400mm gap ex-
posing the refractory. A 400mm wide plate was rolled and welded to close the gap to protect the refractory.
732 INFACON XI
Grouting of the Furnace was carried out each time when flames could be seen on the shell or hot spots. Fur-
ther movement occurred intermittently increasing the distortion on the furnace.
4.1.2. Furnace distortion effects
Frequent furnace breakouts were experienced and furnace shell cracked. Welding of cracks and grouting con-
tinued as a short-term measure.
Production deteriorated at the peak of these problems and difficult tapping conditions prevailed. The move-
ment of the furnace could have affected the PCD and furnace bath causing huge heat losses affecting the water
reticulation pipework resulting in water leaks. There had to be major modification on the tapping equipment
due to the movement of the taphole and the whole front section in order to circumvent alloy losses.
The availability of the furnace also reduced due to breakouts and frequent water leaks, these affected both
power utilization efficiency (Kwh/t) and chromium recovery.
4.1.3. Furnace distortion source
The production performance was poor due to many disruptions to the extent that it became unacceptable. An
investigation was carried on the furnace, particularly on refractory design to actual furnace lining.
Refractory type and quality was verified correct. There was no provision for expansion gap. Bricks were
lined or put against the shell on the sides and therefore during expansion they pushed out the shell in all di-
rections but more severely around the taphole area as this was the weakest area of resistance as a result of
activity related to the tapping operation. Expansion also between the bricks themselves was not provided
which created even a bigger force on the shell.
5. THE DECISION TO RELINE/REBUILD
Poor furnace performance and frequent disruption of production became the main consideration to reline or
rebuild the furnace. A decision to reline could not be simply reached because of the expensive nature of such
a job and the furnace had only operated less than 5 years compared to the 8-10 years life expectancy.
5.1. Patch The Existing Lining, Replace The Distorted Section Or Reline?
Relining a furnace is a disruptive activity on cash flow and other activities as a result very detailed consider-
ation was given to all possible alternatives.
The alternatives identified were:
• Repair the furnace as and when a lining area failure occurred.
• Effect a partial repair on either side of taphole area.
• A full reline of the furnace.
The option of patching the furnace as and when a lining failure occurred was considered to be a temporary
solution and not very conducive to the continued reliability of the furnace as a result of the possibility of un-
timely disruption of the operation. This was in any event, the modus operandi whilst the planning of a mean-
ingful repair was being pursued. Effecting a partial repair either side of the furnace taphole area was a very
feasible option.
A full reline of the furnace was the best idea because it minimized the risk on further operation but the time
and cost of the project was prohibitive since the Company was operating a single furnace.
6. FURNACE PARTIAL REPAIR
On further assessment after a partial reline was agreed upon, the available resource could only allow a 43O
either side of the taphole instead of the 60O but was a good enough repair.
Refractory Design Effect ... 733
6.1. Dig Out
The shell area around the taphole was removed to cover 43O on either side of the shell to facilitate the fur-
nace digout and replace the refractory. Digging was done into the feed zone on this part of the section.
Figures 6 to 8 below shows the process of furnace digging:
Figure 6: Furnace digging in progress
Figure 7: Left side Cavity after brick movement due Figure 8: Right side Cavity after brick movement
to distortion due to distortion
6.2. Metallurgical/Structural Observations
The right side of the taphole had suffered severe refractory movement as was seen by distorted brickwork.
On the left side, the movement left a big cavity between the shell and some refractory bricks.
Figure 9 below shows the analysis of samples taken at different and related points during furnace digout. A
mixture of fused, semi-fused and unfused raw materials were observed together with metal and slag.
The expansion of bricks and the subsequent distortion left weak zones in the refractory lining, these were
observed as the alloy and slag were found between layers of distorted refractory bricks which areas caused
breakout.
The hearth remained intact in spite of all the movement experienced by the furnace shell and particularly
around the taphole.
6.3. Refractory Lining
Refractory lining started after wide consultations with refractory consultants and experienced Furnace lining
companies. The following recommendations were to be adhered to as the lining of the furnace was being
done:
734 INFACON XI
Samples Analyses
Cr2O3 FeO MgO CaO Al2O3
A 51.6 29.1 14.2 1.26 9.1
C 35.6 29.0 15.7 1.01 16.8
E 44.8 27.6 12.0 1.12 15.1
G 12.4 6.5 34.7 2.12 20.8
H 15.6 26.0 16.4 2.76 15.0
Cr Si P
B B 57.1 4.86 0.043
D D 65.7 3.31 0.041
E
F F 48.4 3.97 0.041
A
C
H
G
Figure 9: Analysis of Material from Dig-out Area
(i) A 75mm gap was left between the new refractory and new shell.
(ii) An expansion gap of 2mm between the bricks was left.
(iii) New taphole block installation to the original position.
(iv) A ceram GT90 carbon based material filled the 75mm expansion gap.
Figure 10 to 12 shows the structure of the furnace as the work progressed to partially rebuild the furnace.
Figure 10: Initial stage of Refractory
Installation
Figure 11: Final stage of Refractory
Installation
Refractory Design Effect ... 735
Figure 12: Complete Refractory and Shell
Installation
Figure 13 below shows the new shell in position relative to the old one.
New Furnace
Shell Section
(Installed)
1
2Original
expansion Tap Hole BuckStay
Radial Expansion
of Furnace Shell
Figure 13: State of Furnace after repairs
The furnace base cooling was also modified in the process and furnace side walls in front of the taphole had
cooling mechanism installed. The repair lasted 14 days.
7. PERFORMANCE AFTER THE REPAIRS
There was a significant shift on furnace output, kwh/t and recovery after the furnace repairs.
The Graphs, Figure 14 – 16 shows the comparison of production, chromium recovery and specific power
consumption six months prior to the partial repairs and six months post the repairs.
Production improved by 5622 tonnes for a period of six months (36.6%) after furnace repairs. The chro-
mium recovery significantly shifted to close to 90% after the furnace partial rebuild.
736 INFACON XI
Figure 14: Production
Figure 15: Chromium Recovery
Figure 16: Specific Power Consumption
8. LESSONS LEARNT
The following points summarise the major learning points that emanated from the problems on the 31 MVA
Zimbabwe Alloys HCFeCr Furnace.
• Refractory lining fundamentals cannot be overlooked and the details must be thoroughly checked.
• The furnace performance and life can be affected by such small details during the refractory furnace
design to the effect of business closure if such a premature failure cannot be correctly addressed.
Refractory Design Effect ... 737
• The correct expansion between the shell and the bricks themselves must be calculated accurately.
Expansion gaps must be installed all the times during furnace lining in relation to the type of the
refractory.
• There could have been a merit to quickly arrest the problem as it started, however the pressure on pro-
duction and cashflow constraints distracted the events. In future it is important to weigh out the ben-
efits today as opposed to rectifying the problem and avoid high production losses which translates
into improved revenue.
• Grouting the furnace on areas of hot spots may aggravate shell distortion due to increased expansion
after filling up some gaps in the refractory.
9. CONCLUSION
Refractory lining is by far one of the most important features in furnace design to ensure good performance
over the life span of the furnace whilst a good return on investment is achieved.
