Note: Descriptions are shown in the official language in which they were submitted.
:
07Z~4Z `~`
This invention relates to a method of tapping aluminum from
a cell Eor recovery of aluminum by electrolysis of aluminum oxide
dissolved in a fluoride melt. ;- `
For the recovery of aluminum by electrolysis of aluminum i- `;
oxide (A1203, alumina) the latter is dissolved in a fluoride mel~,
which consists in the greatest part of cryolite Na3AlF6. This
melt is contained in a cell, the innerwalls of which consist of
amorphous carbon. Anodes of amorphous carbon dip from above into
the melt. The aluminum separated at the cathode collects in liquid
state on the bo~tom of the cell beneath the fluoride melt. Oxygen
is released at the anodes by the electrolytic decomposition of the `
aluminum oxide, and combines with the carbon of the anodes to C0
and C02. The electrolysis takes place in a temperature range of
about 940 to 975 C.
The method according to my invention allows the metal height
to be tapped to be accurately determined, having regard to the
' varying thickness of the lateral ledge, the burning away of the
i anodes, the varying interpolar distance and the desired metal lev- `
el, and the cell thereupon to be tapped to the desired metal level. `
~` 20 The method~according to my invention for tapping aluminum
1:
from a cell for recovery of aluminum by electrolysis of aluminum
oxide dissolved in a fluoride melt comprises the following oper-
ational steps carried out in succession:
a) At regular time intervals the instantaneous ohmic cell
resistance is calculated, the instantaneous values over a certain
period of time are smoothed and the difference AR between this
:
smoothed cell resistance and the base resistance established for
each cell is calculated;
b) 30 to 60 minutes after a normal cell servicing which is ;
whenever the difference ~R exceeds a limiting value given for each
cell, the anode beam is raised or lowered, in order to match the
existing ohmic resistance with the ohmic base resistance;
~ `" , .
':
1~)7ZO~Z `
c) The difference hB of the vertical levels of the - -~
anode beam i5 calculated from two values, of which the first is
taken 30 to 60 minutes after the normal cell service following
the previous tapping operation, and the second is taken 30 to 60
minutes after the last normal cell service before the next tap- -
ping operation referred to at (e~ below, -~-
d) The metal height H (mm) to be tapped is calculated --
. ,
according to the equation
H = J Q t f ~ ~B
m
I0 in which J signifies the mean direct current in kiloamps, t the ~;
time in hours which has passed since the previous tapping opera-
. tion, and f a proportionality factor
: . . . .
~ ( mm
f ( ); :
(KA . h)
e) Tapping is carried out to reduce the metal level
.~
by the height given by the equation under (d).
The regular time intervals mentioned under (a) can lie ~ ~;
between 2 seconds and 5 minutes. In practice time intervals of
; 20 10 seconds to 1 minute have proved to be advantageous.
, ,
The periods likewise mentioned under (a) can lle be- `;
tween 10 minutes and 1 hour. In practice advantageously periods
of 10 minutes are chosen. `
The method is applicable both to a single cell and
also to several cells connected in series.
In the drawings which illustrate embodiments of the ,
invention Figure 1 shows a schematic vertical section in the
longitudinal direction through part of an electrolysis cell.
Figure 2 is a diagram of the times at which the level of the anode
,
beam is taken. -~
.
The principle of an aluminum electrolysis cell with
prebaked anodes is apparent from Figure 1, which shows a
~ ~)7Z042 ~ `
schematic vertical section in the longitudinal direction
through part of ~n electrolysis cell. The steel shell 12,
which is lined with a thermal insulatlon 13 of heat-resisting,
heat-insulating material, e.g. chamotte, and with carbon 11,
contains the fluoride melt
":
,'~ ' '
,. ~ ,~ .
~': ;. :
!~
~ . .
,, .
' ,
' ~:
'
- 2a ~
-
7 Z 0~ 2
10 (the electrolyte). The aluminum 14 separated at the cathode lies on the ;~
carbon bottom 15 of the ce31. The surf~ce 16 of the liquid al~m~num constit-
utes the cathode. In the carbon lining 11 there are lnserted lron cathode
bars 17 (in this case transverse to the longitudinal direction of the cell),
which conduct the electr~cal direct current from the carbon lining 11 of the
cell 1 teral~ outl~ards. Anodes 18 of amorphous carbon dip ~rom above into
,~e
the flo~rid~ ~elt 10, and supply the direct current to the electrolyte. They
are firmly connected via conductor rods 19 and by clamps 20 wlth the anode
beam ~1. The anode beam can consist of one or more conducting bars.
The current flows from the cathode bars 17 of one cell to the anode
be~m 21 of the following cell througp conventional bu~ b~rs, not shown From
the anode beam 21 it flows through the conductor rods 19, the anodes 18~ the
electrolyte 10, the liquid aluminum l4, and the carbon lining 11 to the cath-
ode bars 17. The electrolyte 10 is covered with a crust 22 of solidified
m~lt and a layer of aluminum oxide 23 lying above it. Cavities 25 occur m
operation between electrolyte 10 and the solidifed crust 22 Against tne side
waLL9 o~ the carbon l m m g l~ there likewise forms a crust of solid electro-
lyte in the form of the lateral ledge 24. The thickness o~ the ledge 24 de-
termines the hori~ontal extent of the bath of ~`luid aluminum 14 and electro-
lyte 10. With rising temperature, the thickness of the ledge 24 generally
decreàses, with falling te~perature generally increases.
The average distance d from the lower sides 26 of the anodes to
the upper surface 16 of the liquid aluminum, which is also known as the inter-
polar distance, can be adjusted by lifting or lowering of the anode beam 21
with the help of the lifting mechanismus 27~ which are mounted on pillars 28.
~h1s operates on all the anodes. Each anode can however be adjusted by rais-
ing or lowering singly, if ~le respective clamp 20 lS opened, the conductor
rod 19 i9 shifted relatively to the anode beam 21 and finally the clamp 20
i9 again closed. Because of the attack by the oxygen released during elect-
rolysis, the anodes are consumed continuously on their lower face by about
- '
~ 2 ~ ~ 2
l.5 to 2 cms per day (anode burning) according to the type of cell, and 9im-
ultaneous1y the level of the liquid ~lum~num rises by about the sam~ amount
becau~e of the sep~ration of alumintun at the c~thode.
~ nen an anode i9 used up, it must be exchanged for a new ona. The
Cell i9 90 operated m practice that, some days after stàrting up~ the anodes
of the cell no longer have the same degree of consumption ~nd therefore after -
~use for several weeks they must be excha~ged ~eparately. For this reason one
finds anodes of different starting age operating together, as appears from
Figure l. ~-`
The horizontal surface, which contains the totality of the lower
faces o~ the anodes of a cel1, i9 known as the anode table. ; ;~
The princlple of an aluminum blectrolysls cell witn selfbaking an-
odes (Soederberg anodes) i9 the same as that of an aluminum electrolysis cell
with pre~ba~ed anodes.
Instead of pre-baked-anodes, anodes are used which, during the elec-
trolytic operation, are continually baked from a green electrode paste in a
ste~1 jacket by the heat of the cell, The direct current i9 supplied by
lateral steel rod~ or from above by vertical steel rods. These anodes are
renewed a~ required by pouring green electrode paste into the steel jacket.
By breaking ~k the upper electrolyte crust 22 (the crusted bath
surface) the aluminum oxide 23 which is above it is brought into the elect-
rolyte lO. This operation is known as ~ervicing of the cell. In the course
of the electro1ysis the electrolyte becomes depleted in aluminum oxide. At
a 10wer concentration of for examp1e l to 2,S% of aluminum oxide :in the elec-
trolyte, there arise~ the anode effect, which resu1ts in a sudden increase
in voltage from the normal 4 to 4.5 volts for example to 20 volts and above.
Then at the latest the crust must be broken in and the A1203 concentration
be raised by addition of new aluminum oxide.
In normal operation the celL is usually serviced periodically, even -
if no arlode effect occurs. This ceIl servicing will be referred to in what
-4-
,, ' ~
~ Oq ZO 4Z
follows as "nonmal cell servicing~0 It occurs for example every two to six
hoursO In addition, as stated ~bove, upon every anode e~fect the crust of
the bath must be broken in and the A1203 concentration raised by ~ddition of
fresh A1203~ which corresponds to a cell serviceO Thus in operation the anode
effect is always associated with a cell service, which, in contras* to normal
cell service, can be referr~d to as 'lanode effect service"~ ~
The aluminum 14 produced electrolytically, which collects on the ;
carbon bot~om of the cell, is generally tapped once a day from the cell, eOg
by conventional sucking devices. Generally the level of the liquid aluminum
14 is brought back to an optin~lm value for each type of cel10 Thls value
corresponds to the desired metal level, which can be the starting levelO
An important characteristic value in the operation of a cell is
its electrical base voltageO This is established empirically for each cell hav- ;
ing regard to its age, the condition of the carbon lining 11, the composition
of the electrolyte melt 10 as well as the cell current intensity and current ~;density. For the establishment of the base voltage regard is also had to the
hori~ontal extent of the cathode surface 15, which is influenced by the thick-
ness of the lateral ledge 24
From the base voltage the base resistance of the cell can be cal-
culated according to the following equation:
Ro = U0 - 1065
J ~
Ro is the ohmic base resistance in ohms, U0 the base voltage in ~
volts, 1065 the back electromotive force in volts and J the instantaneous ;
cell current intensity in ampsO
For the actual voltage to equal the base voltage, the interpolar
distance must have an optimum value. If the cell is so operated that the
horizontal extent of the cathode surface 16 remains unchanged, thongeneral-
ly the rise in level of the liquid alumin~m above the carbon bottom is equal
to the burning away of the anodes at their lower face. The cell is designed
,
-5-
1 0~ 4 2
so that these condi~ions are reachedO If then in these circumstances one
wants to tap the metal to an extent which will return the metal level to a
starting level, then it is sufficient Just to tap a height of liquid metal -;
that corresponds to the burning away of the anodesO
In practice the actual interpolar distance is from time to time,
eDgO between two tapping operations, larger or smaller than the optimum inter-
polar distanceO The departures are substantially caused by irregular rise in -
.
the level of the liquid aluminum above the carbon bottom, by ~Tegular burning
away of the anodes at their lower face~ and by variation in the horizontal ex-
tent of the cathode surface 16 as a consequence of alteration of the thickness
of the lateral ledge 240 If in this circumstance oLetaps the metal in a cell -~
exactly by the amount which corresponds to the burning away of the anodes,
~e~o_~,s~
~hen one does not reach the desired metal level in the cell,~b*~ the cell is
over or under tapped, that is to say one has tapped too much or too little
metal
My invention relates to a method of tapping aluminum from a cell for
electrolytic recovery of aluminum in accordance with automatic determination
of the metal height to be tapped.
Below, an advantageous example of a method according to my invention
is describedO
By a computer the cell voltage U and the cell direct current intens-
lty J are sampled at regular time intervals~ e.gO every lO to 60 seconds~ and
thc o~ll re~i~tancc i9 calculated from th:i9 according to the equation:
Rinst = U -Jl~65
Ringt is the instantaneous ohmic resistance in ohms, U is the instan-
taneous cell voltage in volts, ]~65 the back electromotive force in volts and
J the cell direct curren~ inten9ity in amperesO
Simultaneously, e.g. with the help of a potentiometer arranged on
the anode beam, the level of the anode beam of the cell is read off by the com-
puterO J~ RinSt and the value of the level of the anode beam are stored in the `
~ 2~42
compu~er. The ~alues for Ri t calculated by the computer are
smoothed over a predetermined period of time, e.g. 10 minutes,
and are compared at regular time intervals, e.g. every 10 to 15
minutes, with the base resistance R of the cell. If the com-
puter notices a difference ~R between the smoothed value and
R , and if this difference exceeds a limiting value previously
given to the computer and stored in it, then an order is issued
by the computer, in accorda~ce with which the anode beam is
raised or lowered, until the instantaneous resistance is sub-
stantially equal to the base resistance of the cell. This ad-
justment is carried out 30 to 60 minutes after a normal cell
service. The values of the level of the anode beam before and
after each movement are read by the computer, for lnstance by
means of the potentiometer, and are stored in the computer.
This procedure ensures in particular that the inter-
polar distance is substantially at the optimum value during
the time ranges when the measurements of anode beam level are
taken on which is based the calculation of ~B, which will now
be discussed.
The metal height to be tapped from the cell is cal-
culated according to the equation
~.
-Ln which H is the metal height to be tapped from the cell in
millimetres, J the mean direct current of the cell in kiloamps,
t the time in hours which elapses between two successive tapping
operations, f a proportionality factor. The dimensions of f are
mm , and this factor provides for the conversion of kiloamp
KA.h
hours into millimetres of anode burning away. Usually f is in
the range 0.0056 to 0.0063. ~B in millimetres is a difference
between two levels of the anode beam.
To calculate the difference ~B of the levels of the
~ - 7 -
~7%~Z
anode beam, reference is made to the following two values o~
level. The first value is taken 30 to 60 minutes after the
normal service following the previous tapping operation. The ~ ~
second value is taken 30 to 60 minutes after the last normal . ;
service before the ne~t tapping operation. The times at which -- `
the two values of
'' ;'
- 7a - ~
level are taken need not be equally far in tim~ ~rom the respective normal
service. From the two values the clifference~ B is obtained.
Figure 2 illustrates in a diagram the tlmes at which the level of ~ ~
the anodè beam ls taken. 50 is the instant of a previous tapping operation, ~-
51 the instant of the first norm~l cell service following it. At the instant
52, which lies 30 to 50 minutes after the first normal service 51, the first
val~e of the level of the anode beam is taken At 53 occurs the next tapping
norm~l operation~ At the instant 54 occurs the la9t normal service before
the tapping operation 53, At the instant 55, which lies 30 to 60 minutes after
the norm~l cell service 54, the second value of the level of the anode beam i
taken Between the instants 52 and 54 there can lie furSher normal cell ser-
vices or anode effect services (not indicated 1~ Figure 2)o
It is to be noted that~ during the tim~ range 30 to 60 minutes after
à normal service, the actual resistance (Rin9t in ohms) of the cell establish-
ed by adjustment of the anode beam should not depart from tlle base resistance
(Ro in ohms) by more than about - 1 x 10 6JQ This restriction is necessaryg ;
~ -6
bec~use if the limiting value of - 1 x 10 J is exceeded, the departure of
the actuàl interpolar distance from the optimum interpolar distance ceases to
be negligible. If the limiting value is not exceeded at each instant of tak-
C 20 ing the level of the anode ~beam~ then one can reckon that the two measure-
ments of the level have been undertaEcen at substantially equal interpolar dis-
tance. The taking of the level9 of the anode beam at the mentioneA determined
tim~s after a normal service on the cell is important, because during this
time the alumina concentration in a cell reaches its maximum value. ~uring
this time the influence of alumina concentration on the cell voltage is prac-
tically negligible. `
The term ~B in the formula H = Jm t f ~ ~B has its origin in
the results of possible variation in the thicknes~ of the lateral ledge of
the cell. If the differencea B is equal to zero, then one can conclude that
there is no alteration in the ledge thickness. All other values of ~B indic-
.. ~ .
-8-
`: :
iO~YZ~)~Z ':
ate an alteration of the ledge thickness. The presence of ~
~B in the formula means that in a long run, that is over a ;
length of time covering several successive tapping operations,
the level of the aluminum is repeatedly returned to an optimum
level, and hence the thickness of the lateral ledge cannot
undergo progressive change, but only minor irregular changes ~ --
around a mean value. - ;
If, during the first taking of the level of the
anode beam, an anode effect occurs, or manipulations are carried
out on the cell, which disturb the taking of the level (e.g.
erroneously carried out movements o the anode beam), this
value of the level should not be employed for obtaining the
difference ~B. In this circumstance either ~B is arbitrarily
set equal to zero, or the provision o the level of the anode
beam is again undertaken 30 to 60 minute.s after the next follow- ,
ing normal service.
If during the second taking (55) of level of the
anode beam an anode effect occurs, or manipulations are carried
out on the cell, which disturb the taking of the level, this
value also should not be made use of in forming the diference.
~B in this case is again arbitrarily set as nil, or the level
oE the anode beam 30 to 60 minutes after the previous normal
service i8 provided from data stored ln the computer.
When the value of the height of me~al to be tapped
has been calculated according to the invention, then the tapping
i5 carried out.
The accuracy o the tapping depends on the devices
available. In order to increase the accuracy of tapping, the
device, a conventional suction device for example, ran be con-
trolled by a computer. When the computer starts the tapping,
while observing the metal level lowers the anode beam so as to `
keep the ohmic resistance of the cell constant, and interrupts
,~ _ g - ':
.. .
:1 V72~4Z `
the tapping operation as soon as the metal level has been re~
duced by the given height.
The advantages of the method according to the ln- :~
vention lie in the fact that the metal height to be tapped is
automatically determined and that this determination gives ;:
more accurate results in comparison with existing ~ ~
,,.,'`: ': ., .
- ~
'. '
'" '~
:, .
- 9a - ;
iO~Z~z "~, ,
technique. If one always tap9 the metal height calculated according to the ~ -
invention, a uniform tapping is ensured, and thus an excessive or insufficient
tapping from the cell is avoided. Thus a uniform cell operation is achieved,
which leads to improve~ent of the current efficienc~- and of the specific elec-
trical energy consumption.
-10- ,
