INFLUENCE OF THE CATHODE SURFACE GEOMETRY

ON THE METAL PAD CURRENT DENSITY

Marc Dupuis¹ and Valdis Bojarevics²

¹GéniSim Inc., 3111 Alger St., Jonquière, Québec, Canada G7S 2M9

marc.dupuis@genisim.com

² University of Greenwich, School of Computing and Mathematics,

30 Park Row, London, SE10 9LS, UK

V.Bojarevics@gre.ac.uk

Keywords: Modeling, MHD, cell stability, irregular cathode surface, current density

Abstract

For the retrofit study presented in

[1] using the irregular top

surface cathode design presented in Figure 1, the cell voltage was

In the recent years, the Chinese aluminum industry has started to

reduced from 4.17 V to 3.85 V which is a reduction of 320 mV.

extensively use irregular top surface cathode blocks in its new cell

Out of that 320 mV reduction, 274 mV came from a reduction of

designs. The increase popularity of these new type of cell designs

the cell ACD (see Table 7 of [1]), clearly indicating that the new

in China is explained by the fact that they can be operated at a

cell design increased the cell stability as compared to the previous

much lower cell voltage.

one.

The cell voltage can be reduced because irregular top cathode

Yet, the cell stability analysis based on MHD-Valdis code

surface designs seem to increase the MHD cell stability which

presented in [4] predicted that adding such transversal ridges,

allows the cell to be operated at a reduced ACD. No satisfactory

while keeping the same metal level, hence, reducing the metal

explanation as to why the usage of irregular top cathode surface

volume, will decrease the cell stability. At best, if the metal

promotes MHD cell stability has been presented up to now.

volume is kept constant, the presence of those transversal ridges is

predicted to have a negligible impact on the cell stability.

The present work concentrates on the influence of the cathode

surface geometry on the metal pad current density as potential

The discrepancy between the cell stability analysis and the

observations was not addressed in [4]. It is important to notice that

cause of the change in the MHD cell stability behavior.

for standard flat top surface cathode design the MHD-Valdis code

was observed to be quite reliable in [5], so clearly, more research

Introduction

work was required.

A typical cell retrofit story involving the replacement of a

Study of the Impact of Cathode Surface Geometry on the

standard flat top surface cathode design by an irregular top surface

Cathode Surface Current Density

cathode design has been presented in [1]. Figure

1 below, a

reproduction of Figure 5 presented in [1], is showing an example

The local variation of the thickness of carbon above the collector

of geometry of the irregular cathode block surface. This is not the

bar(s) has, among other parameters, an impact on the current

only design used, since references

[2,

3] present alternative

density field on the cathode surface and hence the current density

irregular top surface cathode designs.

field in the metal pad. This effect was recognized as a key to the

prediction of the acceleration of the erosion rate of the cathode

reported in [6].

Yet this effect was not considered when the option to define the

geometry of the top cathode was added to the MHD-Valdis code

in order to produce the cell stability analysis presented in [7]. This

was the case simply because the deformation of the top cathode

surface like the one presented in Figure 2 (Figure 8 of [7]), was

caused by the global deformation of the cell. Obviously, as can be

seen in Figure 3 (Figure 9 of [8]), the global deformation of the

cell due to the cathode panel swelling do affect the geometry of

the cathode surface but is not affecting the thickness of the carbon

above the collector bar(s), and hence is not affecting the current

density field on the top cathode surface.

Since the usage of an irregular top surface cathode design do

affect the local variation of the thickness of carbon above the

collector bar(s), this variation must be considered in the

calculation of the cathode surface current density. Since no study

Figure 1. Example of irregular top surface cathode design

has been presented yet on that specific subject, this section is

presenting the result of such a study using a full cell side slice

thermo-electric 300 kA cell model.

Since the cathode carbon material is far more resistive than the

metal, the current has no incentive to enter into those ridges in

order to reach the collector bar. This is indeed what the full cell

side slice thermo-electric model solution is indicating. See Figure

5 for the cathode top surface current density solution.

Figure 2. Metal pad bottom profile input for an impact of

cell deformation cell stability study

Figure 5. Current density in the cathode block in A/m²

The resulting current density in the metal pad is presented in

Total displacement (m)

Figure

6, this time using a vector representation in order to

distinguish between the horizontal and vertical components of the

Figure 3. Cathode panel displacement results from thermo-

current density.

mechanical potshell and lining modeling

Study of the Impact of Longitudinal Ridges

Figure

4 presents the first model geometry case with four

longitudinal ridges. This is similar to the geometry presented in

Figure 5 of [2].

Figure 6. Current density in the metal pad in A/m²

As it is the horizontal component of the current density that is

promoting cell instability, it is pertinent to compare the horizontal

current density in the middle of the metal pad with and without

those four longitudinal ridges (keeping the same metal level). As

can be seen in Figure 7, the four longitudinal ridges are adding

local gradient of current density as the current has to go around

those ridges in order to enter the cathode block in the lower flat

sections between them.

The next step would be to analyse the impact of that change on

Figure 4. Full cell side slice thermo-electric model geometry

the cell stability by using a cell stability analysis code like MHD-

with four longitudinal ridges

Valdis, but unfortunately the required version of the code was still

under development when this study was carried out.

Figure 7. Comparison of the current density in the metal pad,

with and without ridges, in A/m²

Study of the Impact of Transversal Ridges

Figure 9. Current density in the cathode block in A/m²

The second case studied is the case of the addition of transversal

ridges like the ones presented in Figure 4 of [2]. Figure 8 presents

the model temperature solution with the transversal ridge. That

geometry is not that different from the geometry presented in

Figure 1.

Figure 10. Current density in the metal pad in A/m²

Study of the Impact of Cathode Surface Geometry on the Cell

Stability

Figure 8. Full cell side slice thermo-electric model thermal

solution with a transversal ridge

500 kA Flat Cathode Surface Base Case Model

The base case of that cell stability comparison study is the 500 kA

Figure 9 shows that, as for the previous case, the cathode top

cell design presented in Figure

1 of

[9]. The

“classical”

surface current density is quite affected by the presence of the

asymmetric busbar layout is presented in Figure 11.

ridge.

The current density on the top surface of the cathode and at the

Figure 10 shows the resulting current density in the metal pad.

middle of the metal pad is presented in Figure 12. Since the

This time the ridge introduces a horizontal current density

component in the third dimension (the X direction in the model).

busbar network is perfectly balanced and the ledge toe position

has been optimized, there is essentially no horizontal current in

Again, the next step is to analyse the impact of that change on the

the longitudinal direction (JX) in the solution.

cell stability by using a cell stability analysis code like MHD-

Valdis, and again required version of the code was still under

Since the magnitude of the vertical component of the magnetic

development at the time this study was carried out.

field (BZ) is key to the cell stability, that solution is presented in

But for this very specific case, the flexibility of the available code

Figure 13. Finally the evolution of the interface position during

is permitting the user to build and hence analyse this case. That

the transient analysis is presented in Figure 14, from which that

work is presented in the next section.

cell design is predicted to be stable.

500 kA with Transversal Ridges Case Model

As in the previous study [4], the geometry of the top cathode

surface must be entered in MHD-Valdis’ BOTTOM input file.

This ensures that the code will account for that geometry in the

calculation of the metal pad current density and the subsequent

CFD solution of the metal flow.

But in the available version at the time of this study, this doesn’t

ensure that the geometry of the top cathode surface is affecting the

current density on that top cathode surface. Fortunately, in that

specific case, it is possible to ensure that by taking advantage of

the code user input flexibility.

The procedure to follow to achieve that is:

Replace each double bars block in the model by 3

single bar block

Change the flex to network busbar connections

Figure 11. Geometry of the 500 kA base case model showing

accordingly in MHD-Valdis’ BUSNET input file

the current intensity solution in each conductor in A

Manually disconnect all the flexes of the middle block,

(so block 2,5,8 etc) in MHD-Valdis’ BARSIN input file

Run MHD-Valdis with the option to use input from

BARSIN activated

Figure 15 presents the cell geometry obtained by following this

procedure. Figure

16 presents the current density solution

obtained following this cell geometry setup. This solution is only

an approximation of the correct solution as no current at all can

enter in the ridges.

Figure

17 presents the resulting magnetic field that is also

affected. Finally, Figure 18 presents the resulting transient cell

Figure 12. Current density solution on the top surface of the

stability results using the same metal pad depth and the same

cathode in A/m²

ledge toe position.

As for the cell stability study presented in

[4], the cell with

transverse ridges is predicted to be less stable than the base case

cell with flat bottom when keeping the same metal depth, hence,

decreasing the metal volume.

Figure 13. Vertical component of the magnetic field solution

in the middle of the metal pad in T

Figure 15a. Geometry of the 500 kA with transversal ridges

case model (BARSIN and BUSNET files)

Figure 14. Evolution of the interface position (m)

transversal ridges on the cell stability and that effect is negative or

in the best case, where the metal volume is conserved [4], neutral,

something else must be responsible for the observed gain of cell

stability.

In Figure 19 (Figure 4 of [1]), it can be seen that before the

retrofit, the ledge toe is extending a lot on the flat cathode surface.

This is no longer the case in Figure 1 after the retrofit. It is well

known that ledge toe extension under the anode shadow is bad for

the cell stability. In order to illustrate that, the base case flat

cathode surface model will be rerun this time to more accurately

represent the conditions of operation of cells in China before the

retrofit.

Figure 15b. Geometry of the 500 kA with transversal ridges

case model (BOTTOM file)

Figure 16. Current density solution on the top surface of the

cathode in A/m²

Figure 19. Typical ledge toe extension in cells prior to retrofitted

cathode with ridges in China

Compared to the base case model, this case has 5 cm less metal

and about 20 cm more ledge toe extension. The resulting current

density solution is presented in Figure 20.

Figure 17. Vertical component of the magnetic field solution

in the middle of the metal pad in T

Figure 20. Current density solution on the top surface of the

cathode in A/m²

The excessive ledge toe extension introduces a lot of extra

Figure 18. Evolution of the interface position (m)

horizontal current particularly increasing the JX in the end of the

cell where none were present in the base case with optimum ledge

500 kA Base Case Model with Less Metal and More Ledge

toe position. Figure 21 presents the corresponding magnetic field

This leaves intact the discrepancy between the cell stability

solution. Finally, Figure 22 presents the obtained transient cell

analysis results and the observations in China. If this time, it is

stability results.

assumed that the model represents in totality the effect of adding

The cell stability analysis that was performed for a cell with

transversal ridges on its cathode surface taking into account the

two ways those ridges affect the metal pad current density. The

conclusion of the study is that those ridges decrease the cell

stability if the metal height is kept the same (less metal volume).

A new version of the program (see the accompanying paper in this

volume) not available when the present study was carried out is

accounting for the effect and is bringing very similar general

conclusions.

Figure 21. Vertical component of the magnetic field solution

in the middle of the metal pad in T

Since the new results confirm the results of the previous study [4],

the discrepancy between the cell stability analysis and the

observations still needed to be explained.

The last part of the paper addresses this by suggesting that it is the

improvement of the ledge toe position that improved the observed

cell stability not the impact of the ridges on the metal pad current

density or the metal pad flow pattern.

Figure 22. Evolution of the interface position (m)

References

In a short time after the beginning of the simulation, the bath-

J. Zhou et al., “Depth Analysis and Potential Exploitation of

metal deforms and touches the anode, stopping the simulation so

Energy-Saving and Consumption-Reduction of Aluminum

that configuration is predicted to be unstable.

Reduction Pot,” TMS Light Metals, 2012, 601-606.

N. Feng et al., “Research and Application of Energy Saving

Thus clearly, despite the fact that on their own, the introduction of

Technology for Aluminum Reduction in China,” TMS Light

ridges on the surface of the cathode has a negative or at best

Metals 2012, 563-568.

neutral impact on the cell stability, it appears that the

N. Feng et al., “Energy Reduction Technology for Aluminum

improvement of the ledge toe position between Figure 19 and 1 is

Electrolysis: Choice of the Cell Voltage,” TMS Light Metals

the main reason responsible for the gain of cell stability observed

2013, 549-552.

in the industrial cells in China.

V. Bojarevics, “MHD of Aluminium Cells with the Effect of

Channels and Cathode Perturbation Elements,” TMS Light

Conclusions

Metals 2013, 609-614.

S. Ruan et al., “Production Application Study on Magneto-

In the first part of the paper, it was demonstrated that ridges on

Hydro-Dynamic Stability of a Large Prebaked Anode

cathode surface affect the top cathode surface current density.

Aluminim Reduction Cell,” TMS Light Metals 2013, 603-

This influences the metal pad current density in two ways, the first

607.

one by locally changing the depth of metal and the second way by

M. Dupuis,

“Development of a

3D Transient Thermo-

affecting the top cathode surface current density.

Electric Cathode Panel Erosion Model of an Aluminum

Reduction Cell,” COM Light Metals 2000, 169-178.

Depth of the metal variation is taken into account in the available

M. Dupuis, V. Bojarevics and D. Richard, “Impact of the

version of MHD-Valdis at the time this study was carried out by

Vertical Potshell Deformation on the MHD Cell Stability

providing the geometry of that top cathode surface in the

Behavior of a 500 kA Aluminum Electrolysis Cell,” TMS

BOTTOM input file.

Light Metals 2008, 409-412.

M. Dupuis,

“Mathematical Modeling of Aluminum

Changed to the cathode surface current density is not

Reduction Cell Potshell Deformation,” TMS Light Metals

automatically taken into account in the available version of MHD-

2010, 417-422.

Valdis but for the specific case of transversal ridges, it has been

M. Dupuis and V. Bojarevis, “Retrofit of a 500 kA cell

taken into account by performing the appropriate adjustments to

design into a 600 kA cell design,” ALUMINIUM, 87 (1/2),

the MHD-Valdis input files.

2011, 52-55.