SIDERAR Planta Canning

SIDERAR 1999 Experience in Pot Rebuilding

The line produced 578000 Metric tons from April 1991- February 1999.
The first year of operation the strip was entered into the bath at more than 630°C, so the inductors
were working at low power most of the time. Besides we melted pure Al, Zn an Al-Si to produce
our premixed jumbo ingots during the galvanizing runs or during lines stops due to the
unavailability of jumbo ingots in the local market.
Another bad practice was to bubble the inductor channels without turning the power off so as a
result, the nitrogen pipes were worn off and disintegrated into the channels.
So during Galvalume® runs the strip temperature was too high and the inductors worked too little.
Also the temperature variation was considerable when we introduced the cold ingots. All this
 SIDERAR Planta Canning

factors helped to increase the power factor drastically during the first 10 months as showed in the
graphic A.
1

0.9 February 1992

0.8

0.7

0.6

0.5

0.4

0.3 April 1991

0.2

0.1

0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53

From December 1991 to February 1992 we were close to lose the four inductors, when we started
to develop the cleaning method which will be explained in a separate paper.
In 1996 we had to rebuild the top of the front wall due “arise” of the bricks. There we found out that
removing the bricks was extremely difficult and that they were impossible to cut due to its
hardness. In our post repair report, we recommended to plan the rebuilding of the whole furnace
better than repair the remaining walls.

In 1997 we decide to call Junker an Lüngen to have a technical assistance to discuss the unusual
long period inductor life, to plan inductor replacement an to have a qualified opinion regarding the
furnace condition and the steps to be followed.
The technical visit took place in march 1998 when we found a collapsed bushing and during the
next 3 weeks the other 7 bushings collapsed too, in different degrees
Due to this we planned two scenarios, programmed rebuilding, emergency rebuilding.

The programmed rebuild was planned to take place overlapped with major line stop in March 1999.
In case of an unexpected collapse we agreed with Lüngen to have an emergency fabrication
 SIDERAR Planta Canning

program to receive perishable material by air mail ASAP, in order to minimize unavailability of
Galvalume® production.
Since March we stopped inductor cleaning and we limited the maximum power, so full hard runs
were restricted to a minimum.

Rebuilding process
We arrived to February 1999 with the inductors running but under critical condition. Also we knew
that the pot there was plenty of bottom dross, we didn’t change to 70% Zn since January 1998.
To empty the pot we did as follows:
a) bail out with moulds as much as possible
b) pump out to emergency molds
c) keep the inductors running until last moment
d) Turn the inductors of and empty the inductor throats and channels whit spoons

During the whole process we heated the bath with natural gas burners
Soon after we started to empty the pot, we found that there was more dross than expected. When
we arrive to the throats level we saw that they were full of dross.
We put the pumps to empty the last part but only 5 tons of were pumped out to reach the bottom
level due to the huge quantity of dross, as can be seen in photo sequence.
At this point the inductors started to have current peaks and it was impossible keep them running,
so we switched off one by one preserving the cooling air for inductor coils.
We started the disconnection of the electrical cables an we dismounted the air ducts. To avoid
heating of the electrical coils compressed air hoses are connected permanently to the inductor
boxes to ensure fresh air.
Into the inductor channels there was not liquid metal, we confirm it when we drilled a 19 mm hole
into the inductor horizontal channel plug.
All the operations to empty the pot is showed in the next photo sequence:
SIDERAR Planta Canning
SIDERAR Planta Canning
 SIDERAR Planta Canning

We took out three inductors without problems. The connection flange of Inductor #2 had a lot
congealed alloy as a consequence of some spilling during abnormal conditions, such as overflow
during non-ox furnace start up. This frozen metal was very hard to be removed.
 SIDERAR Planta Canning

We
kept the inductors cooled through compressed air and later with fans, until the inner temperature
reached 70 °C. This careful treatment we gave to the inductors allowed us to recover the eight
coils without any damage which was confirmed by Junker in a later detailed inspection, so we had
no need to buy additional coil to complete our inductors spare set.
After the inductors where dismounted, we started to cool the complete pot, first with air fans an
then with water. We needed to have low temperatures on the walls since we planned to use
explosives to minimize the demolition time. So during cooling process we drilled the holes to put
the explosives.
Before we started using the explosives, due to the walls temperature, we tried with pneumatic
hammers with poor results due to the hardness of the bricks.
All the demolition process was carried out while the line was operating with galvanizing pot. For
this reason we were very cautious with the amount of explosives, and the charges were increased
progressively in small amounts.
To put the charges we drilled between the bricks and refractory concrete.
We put explosives all around the pot, then we took all the loose pieces out and use the pneumatic
hammers were was no way to use explosives. Afterward we used explosives again. We repeated
this process three times until we reached the bottom of the pot. At this point only remained dross
and the bottom of the furnace as can be seen in photographs bellow.
The big stones of dross were very hard to be broken so we used hydraulic powers to break them
in pieces. Notice that the corner peaks (mountains) off dross were together with the bottom which
was about 70 to 100 mm thick. We broke the corners and we took the complete base in one piece
of approximately 4 tons (2,5x2,5x1m) after trying to broke it in vain.
The whole process of demolition took 4 complete days after cooling the pot.
The demolition process is resumed in the next photo sequence:
SIDERAR Planta Canning
SIDERAR Planta Canning
SIDERAR Planta Canning
 SIDERAR Planta Canning

Metal Penetration

Then we rebuild the furnace in 10 days. At the same time other concurrent jobs took place around
the pot area so a good coordination was essential.

At this point we left the pot empty heated only by electrical heaters in order to allow the installation
of the new Kholer equipment and to be able to measure everything.

As can be seen in different photos the liquid metal penetrated most of the joints, so the theory
which says that frozen metal is the best barrier to stop metal leakage was effective in our case.
For failure prediction regarding metal leaks we used thermographic analyzes and we don’t find
heat points.

The metal case of the furnace had little corrosion, we sand blasted it and painted with
anticorrosion paint
 SIDERAR Planta Canning

Several inductor bushings collapsed in such a way that produced electrical shortcuts and melted it.
Fortunately the inductor coils were not damaged at all.
 SIDERAR Planta Canning

The inductors failed on its weakest point which is the minimum distance line between the channels
and the coil bushings. There the channels grow and the refractory broke. When the stress was
high enough to cut the riventings of the bushings they collapse. This happens before any liquid
metal reaches the coils.

Conclusions:

• The complete job was very successful thanks to the detailed program, which included every
single task.
• The main problem was the dross accumulation, so we review our cleaning procedure to avoid
this problem in the future
• The inductors showed an excellent performance despite our initial bad practices and the lack
of premelt pot
• We strongly recommend the use of explosives for fast and safe demolition
• We are less concerned about inductor change since we test no problems during inductor
dismounting.
• All the inductors failed whit the same pattern, bushing collapse in the minimum distanced point
between the bushing and the channels

To facilitate the bottom dross cleaning method we designed, together with EFCO a fast removing
system for the lower portion of Galvalume snout, as explained in the pot cleaning paper.