www.aluminiumtoday.com

THERMAL ANALYSIS

Energy saving from a reduction pot

This paper puts forward the benefits of energy-saving in an aluminium reduction pot by minimising

heat loss through the use of a new thermal insulation lining, an insulated and sealed hood, composition

and thickness of the anode covering layer and installing an ‘irregular’ cathode with bevelled edges.

By Zhou Jianfei*, Marc Dupuis**, Yan Feiya*, Huang Jun* & Yi Xiaobing*

The key technologies applicable to primary

Heat output

maintained by reducing the heat

aluminium production are low-

1. Heat dissipation at pot top

dissipated from the pot and balancing the

temperature reduction, increasing the

- internal

cause:

cell

operating

heat input by reducing the voltage across

current,

on-line

measurement of

temperature (Topr)

the pot.

superheat,

Three-variable

control

- External cause: material and thickness

technology, anode slotting, bevelled

of anode covering materials, flue gas

Lowering voltage

cathode blocks, thermal insulation of pot

velocity and degree of sealing of pot

About half of the total power input of a

hoods, drainable cathodes and the

hood.

pot is used for electrolysis and half is lost

application of new lining designs and

2. Heat dissipation at pot side

through heat dissipation. Thus the focus

materials. These are being studied and

- internal cause: Tsuper heat

on achieving a lowering of pot voltage is

developed in China with the aim of raising

- External cause: bath level, metal level,

to reduce the heat lost from the pot. The

aluminium reduction technology to an

pot lining design.

optimum thickness of the side profile

advanced level.

3. Heat dissipation at pot bottom

ledge formed in the bath is a core factor

The consumption of energy and raw

- internal

cause:

cell

operating

in lowering the voltage.

materials for aluminium reduction has

temperature (Topr)

Too thick or too thin a ledge will have a

been high recently especially in regards to

- External cause: material and thickness

negative effect on the pot stability during

power consumption. Aluminium reduction

of cathode lining

production, affecting the voltage required

costs must be lowered.

and the current efficiency

(CE) and

The most efficient method to achieve

The objective is to minimise the voltage

possibly resulting in leakage from the pot.

this is to lower the consumption of the DC

drop across the cell by a combination of

Therefore, the distribution of heat

current used for electrolysis by increasing

adjusting the heat input, mainly by

dissipation and the formation of an

the current efficiency (CE) and reducing

adjusting the voltage drop across the cell

optimum thickness of the side ledge are

pot voltage.

between the anode and cathode (ACD).

key to pot design and has led to a new

The pot energy balance has been

The heat generated during electrolysis

thermal insulating lining being developed.

summarized by Warren Haupin^[1] as:

arises from the Joule heat produced by the

current passing through the bath between

Heat dissipation: Pot top

Heat input

the anode and the cathode.

As the heat lost from the top of the pot

Current (variable)

Stable production is maintained by

accounts for about half of the total heat

Voltage

ensuring a dynamic balance of heat lost by

loss the focus on lowering heat loss is in

2.1

Anode (constant)

the pot and heat input by the Joule effect

this area. For a conventional pot, the heat

2.2

Cathode (constant)

during operation.

lost from the pot top equates to 1.0V to

2.3

ACD (Anode Cathode Distance)

If the Joule heat generated is insufficient

1.2V (the figure is expressed in volts since

2.3.1

Bubble voltage drop (variable)

to balance the heat output, the pot

Topr = heat loss in kW / Cell current in kA)

2.3.2

Bath voltage drop - ACD,

gradually cools, and the process can stop

and the pot voltage is required to be

bath ratio (variable)

and the pot freeze.

above

4.16V. However, based on the

2.3.3

Back-EMF (constant)

The energy balance of a pot can be

statistics of the pots operating with

Area

Heat

Area

Heat

Dissipation Area

Dissipation

(%)

Dissipation Area

Dissipation

(%)

(kW)

(V)

(kW)

(V)

Anode area

Pot hood

Pot side

112.8

0.348

18.4

Anode area

Pot hood

Pot side

97.1

0.299

18.1

cover plate

Pot rim plate

15.5

0.047

2.5

Pot rim plate

21.0

0.064

3.5

Pot end

cover plate

15.0

0.046

2.4

cover plate

18.3

0.050

2.7

Sub-total

143.2

0.439

23.4

Sub-total

134.4

0.412

22.2

Pot

Pot top

43.4

0.133

7.1

Pot

Pot top

67.1

0.206

11.1

superstructure

Anode

guide bar

7.4

0.023

1.2

guide bar

6.9

0.021

1.1

Fume

181.5

0.557

29.7

Fume

111.5

0.342

18.5

Sub total

232.3

0.713

38.0

Sub total

185.5

0.569

30.7

Total

375.5

1.152

61.4

Total

319.8

0.981

52.8

Table 1 Distribution of heat loss from the anode area for a Type A conventional pot

Table 2 Distribution of heat loss from the anode area for a Type B pot

Aluminium International Today

March/April 2012

42 THERMAL ANALYSIS

www.aluminiumtoday.com

Conventional Irregular

measures to improve the insulating

A pot operating at a lower voltage must

Cathode

properties at the middle and at the ends

maintain the thermal balance at a lower

Lining thickness (cm)

of the pot by increasing the thickness of

superheat. Thus the thickness of the pot

ACD (cm)

5.4

4.5

the covering layer to ensure uniformity of

lining must be matched to ensure

Superheat (°C)

the total thermal balance of the pot so as

sufficient heat dissipation and a uniform

Pot voltage (V)

4.17

3.85

to achieve a uniform and regular side

yet not excessive ledge thickness.

ledge within the pot.

Simulation software to show the energy

Table 3 Comparison of physical parameters of different

The thickness of the cover materials has

balance of an aluminium reduction pot

350kA pots

the greatest impact on heat dissipation at

was developed by GAMI and combines

the top of the pot. Taking a 350kA pot as

the thermal field simulation software

various currents, heat dissipation at the

an example, the covering layer was made

developed by Prof Marc Dupuis.

top is below 0.98V for pots with a voltage

up of equal parts of crushed bath and

The software aims at modelling various

lower than 3.9V. A lowering of the heat

alumina.

size pots as well as advanced reduction

dissipated from the top of the pot thus

technologies and process parameters and

has the greatest effect on lowering overall

Application of new pot hood

employs models for heat loss at the

heat loss and contributes the greater part

A new energy-saving sealed pot hood has

cathode, pot sides and pot ends as well as

to a lowering of pot voltage.

been designed. This is lined with a high

an anode model.

There is an important relationship

temperature fire resistant layer and with

These are combined through a three-in-

between heat dissipated from the top of

thermally insulating composites on both

one unifying method^[3]. The models are

the cell and the pot voltage. Using data

the inner and outer surfaces of the hood.

based on data collected over a number of

from a 320kA pot in a Chinese smelter,

This insulating layer prevents the heat

years to verify optimum heat dissipation

this paper illustrates this relationship for

from the top of the pot escaping through

and correct ledge thickness.

two types of pot.

the sides of the hood and so reduces heat

The first type (Type A), is of conventional

loss from the pot. It was observed that the

Comparison of designs

design with a thinner cover material on the

external surface temperature of the pot

The results calculated for two different pot

anode and bath, a voltage of 4.16 to

hood was lowered by more than 10°C and

designs both operating at 350kA, one with

4.18V and heat dissipation of 1.153V in

the voltage reduced by 5 to 15mV using

a conventional lining and a flat bottom

the anode area. The total heat loss from

the new pot hood compared to a

cathode installed using conventional paste

the hood and superstructure of the pot in

conventional hood.

to seal the steel conduction bar into the

the anode area are summarised in Table 1.

cathode block. The second heat dissipation

The second type of pot (Type B), is a

Heat loss at pot sides and bottom

calculation was for a pot with a so called

redesigned version with a thicker layer of

The heat dissipated from the sides and

‘irregular’ cathode (a term adopted in China

covering material, a pot voltage of 3.8 to

bottom of the pot accounts for half of the

which first produced such cathodes in

3.85V and heat dissipation of 0.98V in the

total heat output. The degree of superheat

which the long edges of each carbon block

anode area, as summarised in Table 2.

of the electrolyte is directly related to heat

forming the base of the pot are bevelled -

dissipation, the principle of which is to

see Fig 2) and connected to the steel bus

Pot covering

maintain stable production at minimum

bars by sealing with molten cast iron poured

The composition and thickness of the pot

voltage by adjusting the superheat to

between the block and recesses bus bar.

covering layer regulates the thermal

ensure an energy balance. The design of

The physical parameters of the

balance which can be controlled by

the lining of the pot is indirectly related to

conventional pot with flat cathode are

adjusting the proportion of alumina in the

dissipation of heat. The purpose is to form

compared with those of the pot fitted with

mixture of alumina and crushed bath

a uniform but not excessive thickness of

the irregular cathode in Table 3.

electrolyte which makes up the covering

the ledge under a low superheat while

A comparison of heat dissipated at the

material as well as its thickness.

guaranteeing stable and effective

various parts of each pot is presented in

For large pots, it is necessary to take

production.

Table 4.

A comparison of the temperature

distribution

shows the

highest

Conventional

(mV)

New thermal

(mV)

temperature to be at the side steel plate of

lining structure

a conventional lined pot reaching 301°C

Anode voltage drop

346

Anode voltage drop

347

while that for the pot fitted with the

Clamp voltage drop

irregular cathode was 230°C (Fig 2).

Guide rod voltage drop

Comparing the profile ledge formed in

the two types of pots, the thickness of the

Explosive welding voltage drop

profile ledge at the sides of the

Anode stub welding drop

conventional pot was 10.9cm and that for

Voltage drop of iron/carbon joint

105

Voltage drop of iron/carbon joint

104

the redesigned pot 12.4cm and the ledge

Carbon block voltage drop

150

Carbon block voltage drop

151

thickness at the end of the pots was

Bath layer voltage drop

1502

Bath layer voltage drop

1228

16.7cm and 17.6cm respectively.

Bubble layer voltage drop

170

Bubble layer voltage drop

170

The ledge toe along the sides of the

Cathode voltage drop

284

Cathode voltage drop

229

pots showed a much greater difference at

Cathode steel bar voltage drop

109

Cathode steel bar voltage drop

106

18cm for the conventional and 30cm for

Cathode joint voltage drop

106

Cathode joint voltage drop

redesign and the ledge toe at the pot ends

Cathode carbon block voltage drop

was 13.8cm and 27cm respectively.

Counteraction electric potential

1672

Counteraction electric potential

1672

From a comparison of the distribution of

Voltage drop for busbar around pot

200

Voltage drop for busbar around pot

200

heat dissipation, the greatest heat loss for

Pot working voltage

4.174 (V)

Pot working voltage

3.846 (V)

the conventional pot is in the cathode

area, while that for the redesigned pot is

Table 4 Comparison of voltage distribution

in the anode area. The distribution of heat

March/April 2012

Aluminium International Today

44 THERMAL ANALYSIS

www.aluminiumtoday.com

Conventional

New design

ANSYS 11.OSP1

Pot region

Heat loss

% Total

Heat loss

% Total

78.201

Plot No1

187.516

296.83

heat loss

Heat loss

406.145

515.459

624.773

Anode total

0.971

47.02

0.928

54.16

734.088

843.402

Cathode total

1.094

52.98

0.785

45.84

952.717

of which side

0.663

32.11

0.388

22.65

Rim plate

0.103

4.99

0.068

3.97

Bottom

0.163

7.90

0.143

8.33

Steel bar head

0.165

7.98

0.187

10.89

Total

2.065

1.713

Energy utilisation ratio

45.21%

50.25%

Cell temp 3D

Table 5 Comparison of heat dissipation distribution

Fig 1 Calculated temperature distribution for a

loss in the anode and cathode areas in a

Conclusions

conventional pot with flat cathode (pot end)

conventional lined pot are 47.02%, and

The methods to reduce energy

52.98% respectively and for the new

consumption through pot voltage

ANSYS 11.OSP1

76.616

design 54.16% and 45.84% respectively.

reduction with respect to heat dissipation

Plot No1

186.021

295.426

Thus the proportions of heat loss in the

are:

404.832

514.237

623.642

two types of pot are reversed (Table 5).

733.048

842.453

In the cathode area, the heat dissipation

- The thickness and composition of the

951.858

from the pot sides accounted for the

anode covering material;

greatest heat loss amounting to 32.11%

- The use of a new-design of thermal

of the total pot heat loss in the case of the

insulation including new types of material.

conventional pot and falling to 22.65% of

- Compared to a conventional pot, the

the total for the redesigned pot, but still

reduction in cell voltage is around 200-

remaining the major contributor to heat

450mV.

loss in the cathode area.

- The energy consumption per tonne

Cell temp 3D

aluminium is reduced by around 640-

Economic benefit

1440kWh/t at 93% current efficiency.

With the prerequisite that the efficiency of

- The annual reduction in energy

Fig 2 Calculated temperature distribution in a redesigned

the redesigned lining is 1-2% lower than

consumption of the pot line is typically

pot with irregular cathode (pot end)

that of the conventional lining, the

between 32x107 to 72x107kWh/y for a

China Guangxi Branch?’Comprehensive test report for

redesign can still save 900kWh/t of electric

pot line of capacity 500kt/y.

physics field of 320kA prebaked anode pot’, 2007.7.

energy consumption (Table 6).

- Savings in operation cost of a 500kt/y

[6] Aluminium Corp of China, Zhengzhou Research

smelter are in the range of RMB160M

Institute,

‘Comprehensive test and energy-saving

Comparison of heat loss

($25.3M) to 360M ($57.17M ) per annum

potential research report for 320kA pot’, 2008.6.

Table 7 shows that the distribution of

based on a power price of RMB0.5

[7] Aluminium Corp of China, Zhengzhou Research

heat loss between the cathode and the

($0.07)/kWh.

Institute,

‘Comprehensive test report for related

anode for the new thermal insulation pot

parameters measurement for 400kA pot’, 2008.3.

fitted with the irregular cathode is the

References

[8] Aluminium Corp of China, Zhengzhou Research

reverse of that seen in a conventional lined

[1] Warren Haupin, Halvor Kvande , ‘Thermodynamics

Institute,

‘Comprehensive test and energy-saving

pot with +54% of total heat lost from the

of Electrochemical Reduction of Alumina’, TMS Light

potential research report for 350 kA pot’, 2010.10.

anode in the new design compared with

Metals, 2000, 379-384.

[9] Gui Yang Aluminium and Magnesium Design and

43% in a conventional pot.

[2] Jayson Tessier, Carl Duchesne, Claude Gauthier,

Research Institute and Aluminium Corp of China,

The largest difference is for the heat lost

Gilles Dufour, ‘Image Analysis for Estimation of Anode

Guizhou Branch,

‘Comprehensive test report for

from the sides of the pots which decreases

Cover Material Composition’, TMS Light Metals,

physics field of 240 kA prebaked anode pot’, 2008.7.

from 35% for a conventional pot to 25%

2008, 293-298.

for the new design.

[3] Tian Yingpu,

‘Pot anode cathode distance

Contact

In the past two years the application of

composing model and process energy consumption’,

*Chalieco Gami, 2 Jinzhu Road, Jingyang,

the new insulation lining material has

Light Metals, 2011.3.

Guiyang, Guizhou, China 550081

become a focus for the aluminium

[4] Energy source science and engineering college of

Email jf_zhou@qq.com

industry. Ceramic fibres and compounds

Central South University and Aluminium Corp of

Web www.gami.com.cn

of silica, magnesium-aluminium type

China Guangxi Branch ’Comprehensive test report for

**GéniSim Inc, 3111 Alger St, Jonquière, QC,

thermal insulation materials are widely

physics field of 160 kA prebaked anode pot’, 2007.7.

Canada, G7S 2M9

used to insulate the inside and outside of

[5] Energy source science and engineering college of

Email marc.dupuis@genisim.com

the steel pot structure.

Central South University and Aluminium Corp of

Web www.genisim.com

Parameter

Conventional

New design

Conventional lining

New thermal

(%)

insulation lining %

Current efficiency

94%

93%

Anode area

>54

Daily aluminium production

2650kg

2622kg

Cathode area

<46

Direct current consumption

13231kWh/t

12323kWh/t

Lateral part of pot

<25

Pot rim plate

Table 6 (above)

Table 7 (right) Distribution of heat loss in a conventional

Bottom of pot

Comparison of economic

lined pot and the new thermal insulation lined pot with

Collector bar head

benefit

irregular cathode

March/April 2012

Aluminium International Today