EP2350353B1

EP2350353B1 - Method and means for extracting heat from aluminium electrolysis cells

Info

Publication number: EP2350353B1
Application number: EP09823878.5A
Authority: EP
Inventors: Sigmund GJØRVEN; Yves Ladam; Bjørn Petter MOXNES; Petter NEKSÅ; ASBJøRN SOLHEIM
Original assignee: Norsk Hydro ASA
Current assignee: Norsk Hydro ASA
Priority date: 2008-10-31
Filing date: 2009-10-26
Publication date: 2016-06-29
Anticipated expiration: 2029-10-26
Also published as: WO2010050823A1; CA2741168A1; BRPI0919993A2; AR074082A1; AU2009310492A1; EA201100709A1; EA020514B1; CA2741168C; NO20110740A1; WO2010050823A8; CN102203325A; CN102203325B; NZ592384A; NO20084938L; AU2009310492B2; EP2350353A1; EP2350353A4; NO337977B1

Description

The present invention relates to a method and means for extracting heat from an electrolysis cell for production of aluminium. Specifically, it relates to the cooling of the anode/stubs/yoke assembly by heat conduction upwards along the anode stem, and the enhancement and control of this cooling effect.
The anode assembly in aluminium cells consists of the anode stem (rod), the anode yoke with stubs (studs), and the carbon anode block. The stem is attached at its upper end to the anode beam by means of a clamp, and its lower end is connected to the anode yoke. The stubs are integrated with the anode carbon block. The anode stem can be made of aluminium or copper, while the yoke is made of aluminium, copper or as normal made of steel. The stubs are made of steel. The electric and mechanic connection between the stem and the yoke is constituted by a bimetallic plate. One conventional way of fastening the stubs in holes in the carbon block is by means of cast iron.
Besides supplying the electrical current to the anode and providing the mechanical connection to the anode beam, thus fixing the anode in its correct position, the anode stem plays an important role in the energy balance of the cell. Approximately 50 percent of the electrical energy input to the cell is lost as heat. Up to 50 percent of the heat loss takes place at the top of the cell, and the major part of this again is through the anode.
Typically, for a 300 kA cell about 6-7 kW of heat is conducted through each anode carbon block from the electrolyte and upwards. Some of this passes through the anode cover material on top of the anode, but most of the heat (about 5 kW per anode) is conducted through the stubs and into the yoke. About 4 kW is then dissipated from the yoke and stubs by electromagnetic radiation and convective heat transfer, while the remaining 1 kW is conducted into the anode rod. Part of the latter heat is dissipated into the gas between the top crust and the superstructure, and part of it is dissipated outside the superstructure.
The energy balance in an aluminium cell is very delicate. It is uttermost important to keep the energy balance right, since the cell operation heavily relies on having a layer of frozen electrolyte at the inner walls of the cell to protect the lining. When increasing the amperage in existing potlines, numerous actions must be taken in order to adapt to the higher current. Well-known measures are the use of cathode carbon with high electric conductivity, accommodation of larger (longer) anodes, increasing the dimensions of the cell cavity by using thinner sidewalls, and decreasing the anode-cathode distance (ACD). However, there are upper limits for the anode dimensions, and a lower limit for the ACD that can be used without excessive loss of current efficiency and without risking magneto-hydrodynamic cell instability. From a certain point on, further increase of amperage is only possible by keeping the ACD constant and taking measures to increase the heat flow out of the cell.
Arguably, the easiest way to increase the heat losses is by increasing the number of stubs in each anode, or by increasing the diameter of the stubs. Besides increasing the heat loss, this has the inherent benefit of decreasing the electric resistance of the anode assembly. However, the increase in the heat loss through the stubs is less than proportional to the increase in the cross-sectional area, and the larger stub dimensions may give problems with anode cracking.
Increased heat losses from the stubs/yoke will also lead to increased temperature in the raw gas. There are, at least, three reasons why this is not desired; 1) Increased maintenance costs related to the filter bags in the dry scrubber if the temperature increases above their designed operating temperature, 2) It is important to keep the temperature of the superstructure below certain limits due to the numerous electromechanical installations in this area, and 3) There may be increased heat stress on the operators working in the vicinity of the cell. The extra heat losses must therefore be compensated by increased air suction into the cells. However, the air flow in the exhaust ducts and the gas scrubbing system is the far largest mass flow in an aluminium plant (e.g., 80 t air/t Al), and the cost of transporting the gas is approximately proportional to the cube of the volumetric flow. Moreover, increased suction rate may also require a scaling up of the equipment related to the dry scrubbing system.
One way of resolving the problem with increased raw gas temperature without increasing the suction rate, is to cool down the raw gas by spraying water mist into the raw gas ducts, as disclosed in WO 2004 064984 . One probable disadvantage related to this way of cooling the raw gas is increased corrosion in the raw gas ducts. Furthermore, the moisture content in the alumina fed to the cells will increase, which probably gives higher HF emissions to the environment. A better way of decreasing the temperature in front of the dry scrubber is probably to place one or more heat exchangers in the raw gas flow. The problems related to fouling in the dusty and contaminated raw gas appears to be solved; see the description in WO 2006 009459 .
It was recently disclosed that a decrease in the raw gas temperature, as well as a strongly increased heat flow through the anode, can be achieved by active cooling of the anode yokes ( WO 2006 088375 ). The potential of amperage increase, as well as the amount of heat taken out from the raw gas, appears to be extraordinarily high in this concept. Still, the modification of the anode yokes and the necessary installations at the cell's superstructure may require unacceptably high investments in some cases.
NO 318 164 B1 corresponds to WO 2004/018737 that discloses a method for control of inert electrodes in an electrolysis cell for aluminium production. The problem to be solved is to reduce dissolution of the anode material by transporting heat away from the active anode surface and to reduce deposit formation on the active surface of the cathode by preferably keeping the temperature of this surface higher than that of the electrolyte. By solving this problem, the electrolytic process based on inert electrodes can be enhanced.
US 4,737,247 relates to an inert electrically nonconductive material for spaces in inert anode-inert cathode assemblies. A spacer is provided having a hollow cavity through which a heat pipe coolant, such as potassium or sodium, could be passed to freeze a protective layer of bath around the spacer. A layer of frozen bath forming around the spacer protects the inert anode-inert cathode assembly form attack by the corrosive fluoride-containing molten salt bath of the cell.
WO 2006/007863 discloses an apparatus for the production of metals comprising an anodic system of dimensionally stable anodes with their lower portion immersed in a liquid electrolytic bath, wherein a solid electrically conductive layer is maintained on the external surface of the anodes by providing cooling on the internal body of the anodes by using manifolds for the intake and exit of the cooling fluid. Similarly, the cathodic system can be provided with a cooling system.
One main purpose of cooling the anode assembly as described in accordance with the present invention, is to be able to raise the amperage on the cell while maintaining the side and end ledge (frozen bath) in the bath phase without reducing the ACD, without increasing the dimension of the stub and yoke and thereby without increasing the temperature of the raw gas. Removing heat from the anode with an active cooling will also increase the efficiency of stub, yoke and stem as a heat sink for heat leaving the interpolar distance where most of the heat is generated. The reason for this is because the specific electrical and thermal conductivity of steel will increase and thereby leading to an increased heat loss through the stub and yoke and also because less internal heat will be generated in the material (steel). Calculation on a heat balance model with active cooling of the anodes has shown possibility for a 10 % increase in the amperage maintaining the interpolar distance and keeping the side ledge constant.
The basic idea in the present invention is to extract more heat from the interior of the cell, as well as reducing the heat dissipated into the raw gas, by increasing the amount of heat conducted from the cell along the anode stem. Enhancement of the heat removal from the cell can be achieved by improvement of the conduction along the stem or by installing a convective heat transfer circuit machined inside or fixed on the stem. The heat transfer fluid is circulated down to the yoke where it is heated up. It brings back this heat outside of the superstructure where the heat is released. Heat intake and release can be enhanced by phase transition of the refrigerant (boiling and condensation).
Further, the raw gas temperature can be reduced by applying thermal insulation at least partly at the anode stem inside the superstructure in that it limits the dissipation of heat into the raw gas.
In accordance to the invention, there can be removed heat in an amount that influences the overall thermal balance of the cell.
These and further advantages can be achieved with the invention in accordance to the accompanying claims.
In the following, the invention shall be described further by examples and figures where:

Fig. 1 discloses in general an anode assembly,
Fig. 2 a-b disclose two embodiments of cross sectional views of anode stems in accordance with the invention,
Fig. 3 discloses a diagram showing temperature gradients along an anode stem, calculated for four cases as discussed in the following text.

In Fig. 1 there is disclosed an anode assembly for an electrolysis cell that comprises an anode stem 1 which is connected to an anode beam 2 and an anode yoke 3 from which stubs 4 provide further electric contact to a carbon anode 5. The anode stem is cooled by increasing the surface area of the stem above the cell's superstructure 6, or by applying a cooling medium that circulates along the stem. The anode cooling is combined with the use of a thermal insulation material 7 at the anode stem below (inside) the superstructure.
In Fig. 2a and 2b there is shown two embodiments for arranging medium transport inside the anode stem 1. The Figures show possible technical solutions, which may also be used in combination with cooling of the anode yoke ( WO 2006 088375 ).
In Fig. 2a, the anode stem 1 contains a longitudinal pipe 22 for the cold fluid supplied or recycled at the top, and another longitudinal pipe 23 for the hot fluid coming from the bottom of the stem or from the yoke and the bottom of the stem. The latter pipe is thermally insulated 24 in order to avoid heating of the cold fluid or the anode stem itself. The pipes can be made two in parallel as in Fig. 2a or concentric as in Fig. 2b.
In Figure 2b the anode stem 1' contains a longitudinal pipe 22' for cold fluid supplied or recycled at the top, and another longitudinal pipe 23' for hot fluid coming from the bottom of the stem or from the yoke and the bottom of the stem. The pipes are arranged concentric with a layer of insulation 24' between them.
The preferred technical solution should as earlier stated be a fluid that evaporates at the lower part of the stem or within the anode yoke, and is condensed at the upper part of the stem. Since there is a relatively large surface of contact between the anode beam and the stem, the heat from the top of the stem can be extracted by cooling the anode beam. This eliminates the extra work needed during anode replacement, if the fluid supply to and from the stem or yoke must be connected and disconnected.
The anode stem should be supplied with a relief valve, in case increasing temperature should lead to an unacceptable pressure build-up.
Circulation of the cooling medium can be forced by a pump or a compressor. Circulation can also be simply triggered by buoyancy. This is the classical concept of thermosiphon. The heat transfer fluid is heated at the bottom (yoke). It expands and flows to the top (outside the electrolysis cell) where it is cooled. Its density increases and it falls back to the yoke. In this prospect CO2 based thermosiphon was found particularly promising. CO2 is an inert gas reducing safety issues, and heat exchange properties are very good. Calculations showed that 0.014kg/s of CO2 at 50bars could carry 3kW between the hot side (yoke) at 300°C and the top of the stem maintained at 100°C. If the heat transfer fluid is filled at a pressure larger than the critical pressure (70bars), the thermosiphon operates in transcritical mode. Very large density difference between the cold and hot sides, and then large flows can be achieved without phase transition which greatly reduces the risk of instabilities.
In order ensure a large heat extraction, the heat transfer fluid must be cooled above the superstructure. There are numerous ways of realising this cooling. The simplest way, but not the more effective, is to increase the surface area of heat transfer circuit above the superstructure with cooling fins. Those fins could for instance be sprayed by water or by a forced flow of air. The forced air flow can be provided by a fan, a lance delivering pressurized air, or by any other appropriate means.
A more advanced solution would be to couple the top the heat transfer circuit with an external cooling module. Heat exchange between the heat transfer fluid and refrigerant could be ensured by a proper heat exchanger. To increase cooling the vapour in the top of the hanger, the pipe that transport the warm gas upwards through the hanger is widened at the top of the hanger, i.e. to a small container. The container should be placed above the area where the current goes into the hanger from the anode beam.
However a solution that requires opening of the cooling circuit can be a tedious operation. Solid contact between fins of the heat transfer and the cooling circuit is another possibility, i.e. a cooling hood that is mounted on the top of each anode hanger will ensure a large surface area and good heat transmission to the cooling circuit.
An option that would solve all problems related to connection and disconnection during replacement of an anode would be to dissipate the heat into the anode beam by conduction across the electrical contact surface. This may require cooling of the anode beam, which would lead to added benefits such as decreased ohmic resistance and better mechanical properties of the anode beam (increased creep resistance).
Ideally the heat extracted should be utilized for power production. The cooling circuit would then preferably be of Rankine type with an expansion turbine driving a generator.
Heat extracted from several anode stems can be collected and led to an energy conversion unit conveniently arranged outside the pot room.
Recently thermionic materials have been developed. Such material installed on the fin of the heat transfer circuit would ensure cooling and convert the heat into electricity without complex connection.
As should already be clear from the descriptions and argumentation above, this way of extracting heat will add to the potential of amperage increase, as well as reducing the demand for higher rate of air suction following amperage increase. But it should also be mentioned that:

By reducing the temperature in the yoke and anode hanger the electrical conductivity through the hanger and yoke is increased, i.e. saving energy.
The invention will help stabilizing the temperature in the hanger and yoke at a lower level than today and make it possible to remove the bimetallic joint. If not removing it, it will live for a longer period.
With a more stable temperature in the hanger and yoke, measurement of the amperage through the individual hangers can be measured indirectly more accurate than today, by measuring the voltage drop over a specified part of the hanger.
Reduced temperature in the raw gas due to cooling of the anode assembly will lead to a lower pressure in the cell resulting in less Nm3 air needed to be sucked from the cell (reduced energy consumption on fans) to keep a certain under pressure in the cell.
Less Nm3 sucked from the cell means less dimensions (reduced investment) on the dry scrubber system. Lower temperature on the raw gas means less maintenance (reduced maintenance cost) on the filter bags in the dry scrubber.
With less heat given away from the anode assembly to the cell, less heat will be lead through the hoods and into the working zone, in other words it will be less heat stress on the operators. This is especially important in the summer time or in parts of the world with a high temperature in the pot room.
By regulating the cooling of the anode assembly it will be possible to change the net heat input into the cell. This can be used when the power in the pot line is reduced for a shorter or longer time by removing less heat from the hanger. In this way the number of cells that has to be shut down due to lack of enough power will be reduced. This is not possible to do, if a solution with increased stubs/ yoke/hanger dimension is chosen as a mean to increase the amperage.
The proposed technical solution can also be used by regulating the effect input to the cell under normal operation instead of moving the anode up and down (power pulsing). If the cell needs more heat, less heat is removed from all or some of the anode assemblies on the cell, and if the cell needs less heat more heat could be removed from the anode assemblies than normal. In this way the need for upwards and downwards movements of the anode to increase or reduce the heat input to the cell will be less and therefore it will be possible to keep a more constant interpolar distance (ACD). By keeping the ACD more constant the fluctuation in the bath level will be reduced, and also the process control will be improved since movements of the anode normally will disturb the resistance signal to the regulator deciding the alumina addition.
By cooling the yoke the need for long anode stubs (typical 30 cm) will be reduced, and thereby it will it be possible to reduce the specific energy consumption due to lower voltage drop in the stubs. A reduction of 10 cm should not be a problem. This will also increase the heat loss from the stubs.
Reduced length of the stubs will allow for higher anodes without increasing the height of the superstructure. (Reduced investment cost)
A colder anode yoke will reduce the maintenance cost of the bimetallic plate in the hanger due to lower temperature in the bimetallic plate, and also reduce the cowboy effect due to less thermal expansion of the yoke, and thereby less expansion force working on the stubs.
If the temperature on the stubs is reduced, the possibility of anode cracking due to a higher thermal expansion on the stubs than on the anode will be reduced.
A lower temperature on the yoke will also make it more easy to use other materials in the yoke than steel, by instance copper with a higher thermal conductivity and higher electrical conductivity than steel. Even an aluminium yoke could be considered.

In order to illustrate and emphasize the main ideas and features of the present invention, a simplified model of the anode stem and its surroundings was made.
The model takes into account the thermal conduction along the anode stem and the heat dissipated from the stem. The heat transferred from the stem to the surroundings was calculated using a single heat transfer coefficient intended to contain both the convectional heat transfer and the electromagnetic radiation. As already indicated, the model was not intended to be very accurate, but still, the results should be regarded as much better than order-of-magnitude-estimates. In the calculations, the boundary between the lower end of the anode stem and the bimetallic plate was assumed to be constant (280 °C).

Four cases were taken into consideration, as briefly explained below,

Case 1:	No thermal insulation on stem, no extra cooling (reference case, today's standard).
Case 2:	No thermal insulation on stem, stem cooled to 50 °C 1 m from the lower end.
Case 3:	Stem thermally insulated below (inside) the superstructure, and cooled to 50 °C 1 m from the lower end.
Case 4:	Stem thermally insulated below (inside) the superstructure, but no extra cooling.

The results of the calculation are given in Table I (heat flows) and in Figure 2 (temperature gradients along the stem).
When comparing Case 2 and Case 1 (the reference case), one observes that the cooling of the stem outside the superstructure brings about an increase in the amount of heat conducted into the anode stem. This effect would, of course, be even more pronounced if the anode stem was cooled to a lower temperature, or cooled closer to the yoke.
Case 3 is comparable to Case 2, except that the stem is thermally insulated below (inside) the superstructure. In this case, the amount of heat conducted into the stem becomes lower, but on the other hand, the heat dissipated into the raw gas is eliminated. Insulating the yoke is therefore an effective means of reducing the raw gas temperature. When comparing Case 3 and Case 4, however, it is clear that insulating the stem should only be done in combination with cooling, or else there will be a considerable decrease in the heat conducted into the stem.
There are numerous ways of realising cooling of the anode stem. The simplest way, but not the more effective, is to increase the surface area of the stem above the superstructure, i.e., supply the anode stem with cooling fins. Other ways of realizing cooling of the anode stem are described in previous embodiments. However, it should be understood that a realization of the invention is not limited to those embodiments. Table I. Heat flows (in W) into and out from the anode stem in four different cases as described in the text above.

Case 1 Case 2 Case 3 Case 4

Conducted from yoke to stem 1211 1404 1121 610

Dissipated outside superstructure 512 964 1121 610

Dissipated into raw gas 699 440 0 0

Claims

Method for extracting heat from an electrolysis cell for production of aluminium, the cell comprising a superstructure (6) with an interior part with one or more suspended carbon anode(s), each anode being suspended by an anode yoke (3) attached to one lower end of an anode stem (1) which is attached to an anode beam at its upper end, the anode beam (2) being arranged outside the superstructure (6), where heat is extracted via the anode stem (1), characterised in that the anode stem (1) is cooled so it extracts heat from the interior to the exterior of the superstructure, and wherein thermal insulation material (7) is applied at least at a part of the anode stem (1) that is inside the superstructure.
Method in accordance with claim 1,
characterised in that thermal insulation is applied at the whole part of the anode stem that is inside the superstructure.
Method in accordance with claim 1,
characterised in that the cooling takes place by increased heat loss from the stem (1) above the cell's superstructure (6) by an increased surface area of the stem (1).
Method in accordance with claim 1,
characterised in that,
heat is extracted via the anode stem (1) by cooling of the anode beam (2).
Method in accordance with claim 1,
characterised in that the cooling takes place by applying a cooling medium that circulates along the anode stem (1).
Method in accordance with claim 5,
characterised in that a separate closed cooling circuit is integrated in the elements of the stem, and possibly yoke and stubs, or individual circuits in each of them, transferring heat to a cooling medium that circulates along the anode stem (1), directly by the same cooling medium or indirectly.
Method in accordance with claim 5,
characterised in that it is applied a cooling medium that is a gas, a liquid, or a liquid that may evaporate and condense, in particular CO₂, i.e. to avoid high pressure connection to the cooling loop which simplifies anode changing procedures.
Method in accordance with claim 5,
characterised in that the cooling medium is circulated by natural convection or by forced convection using a pump or a compressor.
Method in accordance with claims 1-8,
characterised in that heat is transported from the lower part of the anode stem (1) and dissipated into the pot room by natural convection from cooling fins that preferably are water sprayed or exposed to a forced air flow.
Method in accordance with claims 1-8,
characterised in that heat is transported from the lower part of the anode stem (1) by means of a cooling agent and dissipated outside the pot room or into a heat exchanger, where heat can be recovered for power production.
Method in accordance with claim 1-8,
characterised in that heat is transported from the lower part of the anode stem (1) by means of a cooling agent and dissipated by thermionic material producing electricity.
Method in accordance with claim 5,
characterised in that the cooling medium is applied to the stem (1) via pipes (22, 23) attached to the outside of the stem, or in channels inside the stem.
Means for extracting heat from an electrolysis cell for production of aluminium, the cell comprising a superstructure (6) with an interior part with one or more suspended carbon anode(s) (5), the anode(s) being suspended by an anode yoke (3) attached to one lower end of an anode stem (1) which at its upper end is attached to an anode beam (2) arranged outside the superstructure (6),
characterised in that
the anode stem (1) inside the superstructure (6) is at least partly thermally insulated with thermal insulation material (7) and is further adapted for being cooled by a cooling medium that is circulated along the anode stem (1) via pipes (22, 23) attached to the outside of the stem, or in channels inside the stem, whereby heat is extracted from the inside of the superstructure to the outside of the superstructure by a closed cooling circuit.
Means in accordance with claim 13,
characterised in that
it is applied a cooling medium that is a gas, a liquid, or a liquid that may evaporate and condense, in particular CO₂, i.e. to avoid high pressure connection to the cooling loop which simplifies anode changing procedures.