See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/282013566
Improved Environmental and Energy Recovery Performance with New
Furnace Hood Design at Elkem Thamshavn
Conference Paper · June 2002
CITATIONS READS
2 304
11 authors, including:
Gerhard Schüßler Halvard Tveit
Medizinische Universität Innsbruck Norwegian University of Science and Technology
283 PUBLICATIONS 2,603 CITATIONS 35 PUBLICATIONS 354 CITATIONS
SEE PROFILE SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Promiljø View project
Psychosomatic training program View project
All content following this page was uploaded by Halvard Tveit on 22 September 2015.
The user has requested enhancement of the downloaded file.
Silicon for the Chemical Industry VI
Loen, Norway, June 17-21, 2002
Ed.: H.A. Øye, H.M. Rong, L. Nygaard, G. Schüssler, J.Kr. Tuset
Trondheim, Norway, 2002
Improved Environmental and Energy Recovery
Performance with New Furnace Hood Design at
Elkem Thamshavn
Almås, Kurt1), Delbeck, Håkon K.2), Halland, T. 1), Rong, Harry M.2) and
Tveit, Halvard1)
1) Elkem Thamshavn, P.O.Box 10, N-7301 Orkanger
2) Elkem ASA, Silicon division, P.O.Box 5211 Majorstua, N-0303 Oslo
Abstract
One of the few silicon furnaces in the world with energy recovery is located at Elkem
Thamshavn in Orkanger. In October 2000 the furnace was equipped with a new high
pressure steam producing hood that increased the energy recovery significantly. The
paper discusses experience with the new hood after more than one year of successful
operation.
Introduction
Silicon is produced by carbothermic reduction of silicondioxide in open or semi-closed
electric arc furnaces. The silicondioxide is added to the furnace as quartz stones with a
size ranging from 10 to150 mm. Sand cannot be used in the furnace since it will clog
the furnace charge completely. The carbon is added to the furnace as coal, charcoal or
coke. It is also typical to add wood chips to the furnace. Wood chips are a source of
carbon, but the main reason to add wood chips is to get enough porosity in the furnace
mix to ensure a good distribution of the gas formed at the electrode tip. Charcoal will
also improve the porosity of the mix. The particle sizes of all raw materials are
optimised to give the highest possible silicon recovery. The silicon process is highly
endothermic and the temperature where the reaction takes place has to be above 1811C
and much higher to obtain maximum silicon recovery. Electricity is added to the furnace
via three large carbon electrodes to get high enough temperature. An arc will form on
the electrode tips, and in this region the temperature is high enough to make the reaction
take place. The carbon electrodes will be consumed in the process due to the reactions
taking place at the electrode tip. The gas components produced in the furnace (CO and
SiO) will react with oxygen from the air on top of the furnace to CO 2 and SiO2. The
process of producing silicon is very complicated and to get more information we refer to
the text book by Schei et al. [1].
Significantly more energy (electricity and reduction materials) has to be added to the
furnace than we need to perform the chemical reduction of quartz. Figure 1 presents the
energy distribution in a conventional silicon furnace. The diagram shows that only
about 31% of the energy input to the furnace is used for the reduction of quartz. The
remainder is heat loss to the surroundings. It is not difficult to understand that the heat
loss represents both a significant economic and an environmental impact. Although the
energy conservation is relatively high in silicon production, there is room for
improvement. In the future higher energy prices are expected to ensure that there is
more efficient use of energy. The question is whether we can continue to produce
silicon wasting all this energy when we are lacking energy in many parts of the
world?
Sensible heat Chemical effect
in product Off-gas
in product
120 %
7% 74 %
Reduction
material
135 %
Furnace Stack
(P1)
235 %
Electric
130 %
power Hood and duct
(water-cooled)
100 %
Electric loss
to offgas
Loss through Charging equipment
(water-cooled) 30 %
refractories Electric loss Low-temperature
Electrode holders
(water-cooled) cooling water
Figure 1: Energy flow in conventional silicon production
Energy recovery in silicon operation
Energy recovery from silicon and ferrosilicon furnaces is difficult. Normal problems are
the high temperatures, clogging of the tubes reducing the heat transfer between the off-
gas and the heat transfer media (steam or water) and the high investment cost for the
recovery equipment. However, it is possible to overcome these problems and today
several plants are recovering energy to a greater or lesser extent. Two different energy
recovery concepts are mainly used today.
Shot-cleaned high pressure water tube boilers for electrical power production and/or
steam/hot water:
Elkem Thamshavn, Norway (silicon and ferrosilicon)
Elkem Bjølvefossen, Norway (ferrosilicon)
Lilleby Metall, Norway (ferrosilicon)
Vargøen, Sweden (ferrosilicon and ferrochrome)
Self-cleaning gas tube boilers for low pressure steam / hot water:
Elkem Bremanger, Norway (ferrosilicon)
Rana Metall, Norway (ferrosilicon)
Finnfjord Smelteverk, Norway, (ferrosilicon)
The different methods of energy recovery have been previously discussed by
Delbeck [2].
2
The most effective form of energy recovery is to produce steam/hot water that is
used in district heating or heating for other industrial activity close to the silicon plant.
With such a system it is possible to recover as much as about 42% of the total energy
put into the furnace (see Figure 2, about 100% of the electrical energy input to the
furnace). However, most of the silicon (and ferrosilicon) plants are found in minor
communities with considerable distances to potential users of steam or hot water. In
these plants the only effective way of energy recovery is to produce electrical power.
The produced electrical power is either used in the smelting process or put into the grid.
In plants that do not have the possibility to produce electrical energy, the maximum
energy recovery is limited by the internal need for steam or hot water. Figure 2 shows
the energy flow with maximum energy recovered as electrical energy. Due to the high
loss in the turbine only about 13% of the energy can be recovered as electrical power.
The Vargøen plant in Sweden is connected to the district heating and is also
supplying steam to Holmen Bruk, a paper mill close to the plant. Vargøen was the first
plant installing steam producing furnace hoods and chimneys in addition to the boiler
for the off-gas. Vargøen is probably the ferroalloy plant with the most effective energy
recovery in the world. The Lilleby Metall plant in Trondheim should also be mentioned
in this respect because the plant is located close to the centre of Trondheim and
connected to the district heating of the city. The energy recovery is very high, but not at
the same level as Vargøen since the furnace hood and chimney are the conventional
types.
Sensible Chemical
heat in effect in
product product
7% 74 % Loss in turbine
Offgas
and generator
25-30% 75%
Reduction
material
135%
Furnace
Stack
235 % (P1)
Boiler Turbine
Electric
power 130 %
25-30%
100 %
Electric loss Electric
to off-gas power
production
Loss through Charging Low-temperature
refractories equipment cooling water
Electric loss
(water-cooled)
Electrode holders
(water-cooled)
Figure 2: Energy flow in silicon production with maximal energy recovered as electrical
power production
Maximum energy recovery means that the furnaces have to be operated completely
differently from conventional open furnaces. To optimise energy recovery we have to
reduce the amount of false air entering the furnace to give the highest possible
temperature in the off-gas. On the other hand we need to add enough air to the furnace
to ensure that the CO and SiO gas is completely oxidised. We also need to limit the
temperature peaks that can destroy the equipment. Further we have to make sure that the
temperature is in the range where effective dry filtration can be performed. It is also
3
important that the off-gas suction is high enough to avoid that dust enters the furnace
building when the stoking machine is used to distribute the raw materials.
Conventional silicon furnaces are normally open furnaces with a rather low off-gas
temperature. These furnaces are not suited for energy recovery. To recover energy the
furnaces have to be semi-closed and the equipment inside the furnace should stand
temperatures up to 1000C and peaks with even higher temperatures.
Energy recovery at Elkem Thamshavn
Furnace 1 at Elkem Thamshavn is the only silicon furnace with energy recovery in the
world (after the Lilleby plant was converted back to ferrosilicon). Orkanger, a small
town about 2 km from the plant does not have any system for district heating and
therefore the only way to recover energy is to produce electrical power. The shot-
cleaned high pressure boiler and the turbine were installed at Elkem Thamshavn as early
as in 1981. The boiler is still in operation recovering about 7% (16 % of the electrical
input) of the energy in the off-gas and is expected to recover energy for another 10-20
years. The turbine had large excess capacity and in October 2000 furnace 1 was started
up with a new steam producing smoke hood.
The smoke hood was based on the concept developed by Vargøen, but improved to
satisfy the requirements for modern silicon operations. Important criteria for the smoke
hood were:
Increased load on the furnace
Improved operation time and silicon recovery
Increased energy recovery
Improved air quality inside the plant
Improved combustion and reduced emission to air
Improved quality on the silica
Improved lifetime of the hood
During the project we realised that it would be a challenge to construct a new smoke
hood fulfilling all the above criteria. The data available for the old smoke hood were
very limited. We did not know how to improve the hood design. We also had to put the
new hood into an old building with its limitations. Our first step was therefore to create
a 3D-model of the old smoke hood to understand where the problems where. A
comprehensive validation of the different problems with the existing hood was carried
out. The model was developed by NTNU/SINTEF in Trondheim and made with the
simulation program named FLUENT[3]. The model of the old hood predicted all the
problems we had experienced. The old smoke hood did not collect the off-gas properly
especially when one of the doors was opened to stoke the furnace. The combustion was
not complete causing carbon contamination in the silica. Clogging of the off-gas outlet
was a severe problem causing reduced operating time for cleaning purposes. The
pressure loss in the off-gas system was high and the fans had to be operated at high load
to control temperature and the off-gas collection. The high pressure loss resulted in very
high temperature peaks when the furnace charge collapsed during the stoking process.
The temperature peaks opened the chimney and resulted in increased off-gas emissions
to the surroundings. All these problems and the fact that the old hood was close to
technical breakdown were the basic motivation for this project.
4
The new smoke hood design was developed with very active participation from the
operators at Elkem Thamshavn. Together we tested the new ideas and adapted the
model to simulate the effect of the changes. The design of the old smoke hood and the
final design of the new smoke hood are shown in Figure 3. The model predicted that the
new design should fulfil the criteria listed above.
Figure 3: Design of the old (left) and new (right) smoke hood at Elkem Thamshavn
The inner surface of the new smoke hood is built up of steel tubes. The area of the
steel tubes in contact with the off-gas is about 300 m2. Inside the tubes we have water
with a pressure of about 50 bar and a temperature of about 265C. The water is
circulated inside the tubes and if we lose the water circulation a tube rupture occur
immediately with steam / water blow out as a result. A lot of water entering the furnace
can cause explosions. The steel tubes have to be cooled to survive inside the furnace and
if we lose the cooling effect the steel tube will burn and disappear after a short time.
A picture of the inside of the furnace is shown in Figure 4. This shows the tubes in
the wall, one of the electrodes, the burning charge around the electrode and the charging
tube for the raw materials. Figure 5 shows the outside of the furnace during stoking with
one door open. It is seen that unlike conventional furnaces, furnace 1 at Elkem
Thamshavn is semi-closed to reduce the loss of energy.
The furnace with the new hood was started in October 2000. We had done whatever
we could to ensure that the hood worked perfectly. However, we all know that in such
large new projects there is a risk that some factors could have been overseen or not
considered critical and that these factors could cause problems later on. During the
ramping up of the furnace we had two incidences of tube rupture caused by insufficient
water circulation. The designed water circulation could not match the high fluctuating
heat load and thermal strain. As a result steam and water entered the furnace. Figure 6
shows the hole in the tube located in the roof of the hood. A thorough examination was
carried out to detect the reason for the blow out. It was concluded that in that part of the
hood the water circulation was too low. To correct the problem we welded the tube,
improved the circulation and started up the furnace again After the two minor problems,
the hood has been working according to our expectations. We will now discuss each of
the criteria set up in the beginning of the project and evaluate the results.
5
Figure 4: Picture of the inside of the new smoke hood
Figure 5: Picture of the outside of the new smoke hood
6
Figure 6: The devils eye!
Energy recovery a way to reduce cost in silicon production
Increased load: It was not possible to run a higher load than about 18 MW with the
old hood. The water cooled tubes in the roof of the old hood had started to break down
after 18 years of operation. After installing the new steam producing hood the load has
been increased with more than 20%. The hood is no longer the limitation for increased
load and we should not exclude higher load in the future.
Improved operation time and silicon recovery: In the old hood one of the
biggest problems was the clogging of the chimney / hood outlets with silica. The fans
had to be operated at higher load to prevent the off-gas entering the furnace building.
Finally the furnace had to be taken out of operation for cleaning of the hood outlets.
With the new design the clogging almost disappeared, and the operating time improved
significantly.
The silicon recovery is the percentage of silicon tapped from the furnace related to the
amount of silicon added as raw materials to the furnace. The remaining silicon is lost as
silica fume. With the new hood design we improved the silicon recovery by about 10%.
The improvement is a result of several factors where the possibility to increase the
furnace load to an optimal set-point seems to be most important. But also improved
operating time, better distribution of raw materials and improved control of the gas flow
in the furnace seem to contribute to the improvement.
Increased energy recovery: With the new steam producing hood the recovery
of energy has increased to about 10% of the total energy input to the furnace. This is
equal to about 22% of the electrical energy input to the furnace. It is difficult to give an
7
exact number of the energy recovery at furnace 1 since the off-gas from both furnaces at
Elkem Thamshavn enters the same main boiler and the same turbine converts all the
steam to electricity. However, the total energy recovery after the start up of the new
hood was even higher than what we set as an optimistic case in the investment
calculation.
As a case we have collected the data from Q2 2001[5]. In this period the furnace had
an energy consumption of 10.8 MWh/MT silicon. The recovered energy from furnace 1
was 2.4 MWh/MT giving a energy consumption of 8.4 MWh/MT silicon. The amount
of auxiliary energy (fans, pumps, light, etc.) used at the plant is about 1.0 MWh/MT
silicon. If we include all electrical energy we will need about 9.4 MWh to produce one
tonne of silicon. To our knowledge no other silicon producers have ever reported that
low energy consumption to produce silicon.
Improved air quality in the furnace hall: The new hood design reduced the
emission of off-gas to the furnace hall significantly. We are measuring the off-gas
amount each operator is exposed to, and planned measurements in the near future are
expected to confirm the improved conditions in the furnace building. Improved hood
design has also reduced the emission of “diffuse off-gas” to the surroundings leaving
via the furnace building. However, Elkem Thamshavn still has to reduce the emissions
from the tapping area.
Improved combustion and reduced emission to air: With the new hood
we have been able to increase the temperature in the off-gas. We have also obtained a
uniform temperature profile in the whole furnace area. Higher and more stable
temperature gives better combustion in the furnace. Improved combustion has reduced
the amount of soot, PAH, NOx, etc.
Improved quality on the silica: One of the most critical factors in the silica is
the carbon content[4]. Improved combustion has reduced the carbon content
significantly.
Improved lifetime of the hood: It will still take many years before we can
confirm that the lifetime of the new hood is better than the old design. However,
normally the lifetime of a hood is partly determined by corrosion of the tubes from the
inside. To avoid the corrosion we are using pure water (deionised) in a closed system.
The main boiler has been using this type of water since it was started up more than 20
years ago. We are therefore rather optimistic and hope that the lifetime also meets our
expectations.
Cost reduction: Increased load, improved operation time, increased silicon
recovery, increased energy recovery and improved silica quality have all contributed to
reduced production cost at Elkem Thamshavn. Energy recovery is one of the reasons for
Elkem Thamshavn producing at full capacity in a period with a very difficult market.
The energy recovery system at Elkem Thamshavn shows that it is possible to combine
an environmentally conscious production with good economy.
Challenges for the future
Energy recovery systems are very expensive. Silicon production has never given a
satisfactory return to the owners and during the last 30 years many producers have been
in severe financial problems. Therefore the required amount of cash has not been
present or the willingness to invest in energy recovery systems has been very low. The
governments in each country with silicon production should look into possibilities to
8
help the silicon producers with financial solutions. Without such help it is unlikely that
we will be able to utilise the potential of improved energy recovery.
The strong competition from China, with silicon produced with high energy
consumption, has threatened the western silicon production. However, in order to
improve the competitiveness of silicones and aluminium alloys we have to put more
effort in obtaining a more environmentally friendly production of silicon. The silicon
produced with energy recovery systems attached to the furnace (and thereby the
silicones, silicon wafer and aluminium alloy) has a less negative impact on nature.
Conclusion
A significant amount of energy could be recovered from the silicon furnaces around the
world. Elkem has developed such technology and shown that it is possible to obtain
both reduced cost and improved environmental effects by installing energy recovery
systems. Energy recovery units are very expensive and the pay back time for these kinds
of investments is very long. Due to the difficult economic situation for the silicon
industry during the last 30 years it is not likely that many energy recovery units will be
built unless governments are willing to assist in the financing of energy recovery
projects.
Acknowledgement
We greatly appreciate financial support from the Research council of Norway through
the KLIMATEK programme and to Elkem ASA for allowing the publication of these
results.
References
1. A. Schei, J. Kr. Tuset and H. Tveit, Production of High Silicon Alloys, Tapir Forlag,
Trondheim 1998, Norway.
2. H. K. Delbeck, Energy Recovery From Hot Waste Gas in the Ferroalloy and Silicon
Industry, Proceedings from Silicon for the Chemical Industry V, Tromsø, Norway
May 29-June 2, 2000, editors: H.A. Øye, H.M. Rong, L. Nygaard, G. Schüssler,
J.Kr. Tuset, Trondheim Norway 2000.
3. H. Tveit, H. Laux, S.T. Johansen, V. Lund, CFD Study of New and Existing Design
of the Thamshavn Furnace 1 Off-gas System, August 1999.
4. H. K. Delbeck Energy and Silica Powder Recovery from Hot Off Gas in the
Ferroalloy and Silicon Metal Industry, UD/NHO Environmental Seminars in Beijing
and Weihai, China, June, 2001
5. G. Halvorsen and G. Schüssler, Sustainable Silicon Production, 1st International
Silicon Days, Munich, September, 2001
View publication stats
