EP0814181B1 - Process for regulating the alumina content of bath in cells for producing aluminium - Google Patents

Description

TECHNICAL AREA

The present invention relates to a method for precise regulation of the content alumina in igneous electrolysis tanks for the production of aluminum according to the Hall-Héroult process, in order not only to maintain the Faraday performance at a high level, but also to reduce emissions from fluorocarbon gases which are particularly harmful and polluting for the environment and that following the anomalies of operation of the tanks electrolysis known as the anode effect.

STATE OF THE ART

In recent years we have gradually automated the operation of aluminum production tanks, first to improve walking regularity and thereby the energy balance and Faraday performance, but also, with an ergonomic and ecological aim, for limit painful human intervention and increase the efficiency of capture of fluorinated effluents.

One of the essential factors to ensure the regularity of walking of a aluminum production tank by alumina electrolysis dissolved in a molten electrolysis bath based on cryolite, is the maintenance of a content suitable as alumina dissolved in this electrolyte and therefore the adaptation at any time of the quantities of alumina introduced into the bath the alumina consumption of the tank.

Thus, an excess of alumina creates a risk of fouling of the bottom of the tank by undissolved alumina deposits that can turn into plaques hard electrically insulating part of the cathode. This then promotes formation in metal of very horizontal electrical currents strong which, by interaction with magnetic fields stir the sheet of metal and cause instability of the bath-metal interface. Conversely, an alumina defect causes the appearance of the anode effect, resulting in a loss of production and a sharp increase in the voltage across the tank, which can increase from 4 to 30 or 40 volts. This overconsumption of energy also has the effect of degrading the energy efficiency of the tank but also the Faraday yield following the redissolution of aluminum in the bath and the elevation temperature of the electrolysis bath.

The need to keep the content of alumina dissolved in the electrolyte within precise and relatively narrow limits, and therefore to introduce alumina with the greatest regularity possible, has therefore led those skilled in the art to develop automatic processes. supply and regulation of alumina in electrolytic cells. This need has become an obligation with the use of so-called "acid" electrolysis baths (with a high AlF _{3 content} ) which makes it possible to lower the operating temperature of the cell from 10 to 15 ° C (approximately 950 ° C at instead of 965 ° C usually) and thus achieve Faraday yields of at least 94%. Indeed it is then essential to be able to adjust the alumina content in a very precise and very narrow concentration range (1% to 3.5%), taking into account the decrease in the solubility rate of alumina linked to the new composition as well as the lowering of the bath temperature.

Direct measurement of the alumina content of baths by analysis of samples taken periodically that have not proven to be sufficient industrially, most of the known industrial processes have resorted to an indirect evaluation of the alumina contents according to a parameter electrical representative of the alumina concentration of said electrolyte. This parameter is generally the variation of resistance R across the tank supplied with a voltage U, including a counter-electromotive force e evaluated for example at 1.65 volts and crossed by a current I so that R = (U-e) / l.

By calibration we can draw a curve of variation of R as a function of the alumina content and by measuring R (at a frequency determined according to well known methods) we can know at any time the concentration of alumina [Al ₂ O ₃ ] . It is this detection principle that the document FR 1457746 (GB 1091373) adopts for controlling an alumina dispenser associated with a means for piercing the electrolyte crust frozen on the surface of the bath. Likewise US 3400062 implements a measurement of the variation in resistance of the bath using a pilot anode to detect an alumina defect and a tendency to the anode effect and thus act on the rate of introduction of alumina from a hopper fitted with a device for piercing the frozen electrolyte crust.

More recently, precise regulatory methods based on the control of alumina content between an upper limit and a lower limit have made subject of new patents including US 4126525 and EP 044794 (US 4654129), this the latter already being in the name of the plaintiff.

In the first of these patents, the range of alumina contents to be respected is between 2 and 8%. We feed the tank for a time t1 predetermined with an amount of alumina greater than its consumption theoretical until a fixed alumina concentration is obtained (by example 7% therefore a little lower than the maximum admissible of 8%), then we switches the power supply at a rate equal to the theoretical consumption during a predetermined time t2, the supply is finally stopped until appearance of the first symptoms of anode effect. We then resume the cycle power supply at a rate higher than the theoretical consumption. According to what process, and more precisely the results of its application examples, the concentration of alumina in the bath can vary during a cycle of 3 to 8%, which is still insufficient to regulate the alumina content of an acid bath in a range as low and narrow as 1 to 3 or 4%. This is what the process according to EP 044794 (US 4431491), in the name of the plaintiff who makes call on the side of the resistance measurement R at the terminals of the tank electrolysis to a second adjustment parameter, the slope P = dR / dt representative of the change in resistance R caused by a change voluntary alumina diet for a period of time determined. Indeed the only knowledge of the resistance R at the terminals of the electrolytic cell is not sufficient to precisely control the content in bath alumina and consequently to control the quantity or the frequency of anode effects, because the R parameter at bath temperature constant is a function of 2 variables, on the one hand the alumina content image the resistivity ρ of the bath, on the other hand the anode-metal distance (DAM). It is necessary therefore find another discriminating parameter which is obtained by the slope P = dR / dt, called resistance slope, the only parameter truly representative of depletion or enrichment of the bath with alumina. By creating by example a momentary undernourishment of the alumina bath compared to at theoretical consumption, there is an increase in resistivity ρ with the lowering of the alumina content of the bath according to a law of evolution known while at the same time the evolutionary DAM much more slow hardly changed.

It is on the adjustment of these 2 parameters R and dR / dt that the process is based. according to EP 044794 which can be summarized as follows: starting from a phase of undernourishment of the alumina bath, we order the transition to boost for a predetermined duration T if the resistance R exceeds the upper limit Ro + r where Ro is the set resistance and if the slope of resistance P is greater than a setpoint slope Po.

On the other hand if the slope P remains lower than the reference slope Po reference a sufficient content of alumina in the bath, the diet of undernourishment is preserved of the bath, but we give if necessary an order to descend the anodic frame or "tightening" to reduce the DAM and thus bring back R In the setpoint range Ro ± r.

Finally, starting from the supercharging phase of duration T, we go on to a rate of undernourishment at the end of this duration T and if R has become less than the lower limit Ro-r of the setpoint range, an order of rise of the anode frame or "loosening" to increase the DAM and bring R into the setpoint range Ro ± r.
We then start a new cycle.

This mode of regulation therefore makes it possible to maintain the alumina content of the bath in a narrow and weak range and thus obtain yields Faraday of around 95% with acid baths, simultaneously reducing and notably the quantity (or frequency) of the anode effects on the tanks that are counted in number of anode effects per tank and per day (EA / tank / day) under the name "anode effect rate".

On the older generations of side stitching tanks, the effect rate of anode was greater than 2 or even 3 EA / tank / day, while on the tanks more recent with punctual stitching this rate is between 0.2 and 0.5 EA / tank / day. AT this stage the overconsumption of energy and the loss of yield Faraday linked to the anode effects are weak and until recent years this level performance could be considered sufficient.

Recently, however, with the development of very large electrolytic cells high intensity and the search for ever higher performance especially in terms of Faraday and energy efficiency, but also with the consideration of pollution problems by fluorocarbon compounds (CFx), in particular by carbon tetrafluoride CF4, whose high absorption potential of infrared rays promotes the effect reduction, or even elimination, of the anode effects generating fluorocarbon gases has become a priority. In this regard, it is appropriate to remember that the anode effect is a phenomenon of ion electrolysis fluorides which occurs when there is a lack of oxygen ions in contact with anodes due in particular to a lack of alumina. Instead of producing according to the normal process of carbon dioxide and carbon monoxide, the tank produces fluorocarbon gases whose trapping by the usual means is impossible due to their chemical inertness and high stability.

PROBLEM

• la construction de nouvelles usines d'électrolyse mettant en oeuvre des cuves de très haute intensité en nombre toujours plus grand,
• l'extension des usines existantes sans accroissement, voire même avec diminution, des rejets gazeux fluorocarbonés.

The development of a precise regulation process for low alumina contents in the electrolysis bath ensuring a high Faraday yield (≥ 95%) with an anode effect rate of less than 0.05 EA / cell / day has become an essential goal for:

• the construction of new electrolysis plants using very high intensity tanks in ever increasing numbers,
• the extension of existing factories without increasing, or even decreasing, fluorocarbon gas emissions.

OBJECT OF THE INVENTION

The method according to the invention makes it possible to solve this pollution problem by lowering the anode effect rate on average to 0.02 EA / tank / day, that is to say well below the target rate of 0.05 EA / tank / day and a fortiori rate of 0.2 to 0.5 EA / tank / day of the prior art; this even improving the Faraday yield over 95%. The method of the invention uses the principle basic alumina regulation already described in EP 044794 (US 4431491) which implements 2 adjustment parameters, the resistance R and the slope of resistance P = dR / dt, which are compared with setpoints for trigger a change in alumina diet or an order displacement of the anode frame to correct the metal anode distance (DAM).

• la détermination de la résistance et de la pente à chaque fin de cycle de régulation et non plus seulement lorsque la résistance sort de la plage de consigne,
• le déclenchement d'une phase de suralimentation si la teneur en alumine mesurée par la pente de résistance devient très faible et cela quelle que soit la position de la résistance par rapport à la plage de consigne,
• enfin l'affinement des méthodes de détermination de la résistance R et surtout de la pente de résistance P, ainsi que l'utilisation de paramètres auxiliaires qui seront explicités plus loin, assurant à la fois une grande précision et une grande fiabilité au nouveau procédé de régulation.

The method according to the invention is however clearly distinguished from the method previously described by the fact that it implements an entirely different operating sequence at each regulation cycle, in particular with:

• determination of the resistance and the slope at the end of the regulation cycle and no longer only when the resistance leaves the set range,
• the triggering of a supercharging phase if the alumina content measured by the resistance slope becomes very low, regardless of the position of the resistance relative to the set range,
• finally the refinement of the methods for determining the resistance R and above all the resistance slope P, as well as the use of auxiliary parameters which will be explained below, ensuring both great precision and great reliability for the new method of regulation.

It is therefore thanks to the new operating sequence inside each cycle taking into account these different modifications, that the method according to the invention made it possible to divide the anode effect rate by an average of 10 obtained with the methods of the prior art chosen, however, from among the most systematically higher Faraday yields 95%.

DESCRIPTION OF THE INVENTION

(A) Au terme de chaque cycle de régulation i, on calcule la résistance moyenne R(i), la vitesse d'évolution de la résistance ou pente de résistance P(i), la vitesse d'évolution de la pente de résistance ou courbure C(i) et une prévision de la valeur de la pente de résistance à l'instant t(i+1) ou pente extrapolée PX(i)=P(i)+C(i)×T qui est une estimation de la future pente de résistance P(i+1) à la fin du cycle de régulation i+1;

(B) La valeur R(i) est comparée à une valeur de consigne Ro et des ordres de déplacement du cadre anodique sont donnés en conséquence, à savoir : diminution de la distance anode métal ou serrage, augmentation de la distance anode métal ou desserrage;

(C) L'alimentation en alumine est régulée en fonction des valeurs de la pente P(i), de la courbure C(i) et de la pente extrapolée PX(i), de préférence par rapport à des seuils de référence tels que Po, Co et PXo, de manière à compenser par anticipation les évolutions de la teneur en alumine.

More specifically, the invention relates to a process for regulating the alumina content of the bath in an aluminum production tank by electrolysis of alumina dissolved in a molten salt based on cryolite using an alumina supply at a rate modulated as a function of the value and of the evolution of the resistance R of the tank calculated from the difference in electrical potential measured at the terminals of the tank, alternating phases of undernourishment of alumina with the introduction of alumina at a rate slow CL (phase 1) and phases of supercharging in alumina with introduction of alumina in rapid CR or ultra-rapid CUR (phase 2) rate compared to a reference rate or theoretical rate CT corresponding to the average theoretical consumption of alumina of the tank, characterized by regulation cycles of duration T comprising at each cycle, the following sequence of operations:

(A) At the end of each regulation cycle i, the average resistance R (i) is calculated, the speed of evolution of the resistance or resistance slope P (i), the speed of evolution of the resistance slope or curvature C (i) and a forecast of the value of the resistance slope at time t (i + 1) or extrapolated slope PX (i) = P (i) + C (i) × T which is an estimate of the future resistance slope P (i + 1) at the end of the regulation cycle i + 1;

(B) The value R (i) is compared with a set value Ro and orders to move the anode frame are given accordingly, namely: reduction of the metal anode distance or tightening, increase of the metal anode distance or loosening ;

(C) The supply of alumina is regulated as a function of the values of the slope P (i), of the curvature C (i) and of the extrapolated slope PX (i), preferably with respect to reference thresholds such as Po, Co and PXo, so as to compensate in advance for changes in the alumina content.

• Si l'alimentation en alumine est en phase 1, les valeurs P(i), C(i) et PX(i) sont comparées respectivement aux seuils de référence Po, Co et PXo:
  • Si P(i) < Po et PX(i) < PXo, la phase 1 se poursuit
  • Si P(i) ≥ Po ou PX(i) ≥ PXo, on passe en phase 2 d'alimentation en alumine:
    • Si C(i) ≥ Co, la phase 2 commence par une alimentation en cadence ultra-rapide pour une durée prédéterminée ou calculée, suivie d'une alimentation en cadence rapide pour une durée prédéterminée ou calculée, le calcul des durées étant effectué en fonction des valeurs calculées en fin du cycle de régulation précédemment défini
    • Si C(i) < Co, l'alimentation en alumine passe directement en cadence rapide pour une durée prédéterminée ou calculée en fonction des valeurs calculées en fin du cycle de régulation précédemment défini.
• Si l'alimentation en alumine est en phase 2 :
  • la phase 2 se poursuit normalement selon la durée prédéterminée ou calculée à l'issue de la précédente phase 1.

According to an advantageous embodiment of the invention, the regulation of the alumina in step (C) is carried out under the following conditions:

• If the alumina supply is in phase 1, the values P (i), C (i) and PX (i) are compared respectively to the reference thresholds Po, Co and PXo:
  • If P (i) <Po and PX (i) <PXo, phase 1 continues
  • If P (i) ≥ Po or PX (i) ≥ PXo, we go to phase 2 of alumina supply:
    • If C (i) ≥ Co, phase 2 begins with an ultra-fast rate supply for a predetermined or calculated duration, followed by a fast rate supply for a predetermined or calculated duration, the durations being calculated as a function values calculated at the end of the previously defined regulation cycle
    • If C (i) <Co, the supply of alumina goes directly to rapid rate for a predetermined duration or calculated according to the values calculated at the end of the previously defined regulation cycle.
• If the alumina supply is in phase 2:
  • phase 2 continues normally according to the duration predetermined or calculated at the end of the previous phase 1.

During the development of the new process according to the invention, the Applicant was able to observe that it was possible to dramatically reduce the rate of anode effect by switching to a fast-rate feeding regime without waiting that the resistance R has left the setpoint range according to the prior art previously described as soon as the resistance slope P becomes very high, an indication of a very low alumina content in the bath (1 to 2%) and a very high risk of appearance of anode effect. Figure 1 in the appendix which represents the variation of the resistance R at the terminals of an electrolytic cell as a function of the alumina content of the bath for different increasing anode-metal distances DAM ₁ to DAM ₃ , clearly shows that regulating the alumina content of the bath between 1 and 3.5% we are in the best possible conditions, on the one hand to use acid electrolysis baths at lowered temperature guaranteeing excellent Faraday yields, on the other hand for detect the slightest variation in resistance since we place ourselves in the zone with the steepest slope of variation of R, that is to say in the zone with the greatest sensitivity. The counterpart of this double advantage implies a very rapid and quantitatively significant reaction capacity in terms of the alumina bath supply regime to prevent the very significant risks of triggering an anode effect which appear as soon as the alumina content of the bath is around 1%.

To solve this problem incompletely treated by the process of regulation of the closest prior art, which provides only for a calculation of slope value when resistance R exceeds a high reference threshold Ro + r, it turned out to be necessary not only to perform this slope calculation at the end of each regulation cycle, but also the calculation of the slope extrapolated planned for the end of the next cycle to compare them with thresholds and immediately trigger if necessary and in advance an acceleration of the feed rate in the case of a rapid increase in resistance as shown in the graph of the figure 2.

This new procedure for regulating the alumina content does not exclude the implementation of additional safety procedures.
Thus the regulation procedure is initiated only when the tank is in normal operating conditions (that is to say correctly adjusted, stable and excluding disturbing operations of operation or adjustment such as change of anode, casting of metal or specific regulation procedures) authorizing the transition to phase 1. In the case where the tank is not in normal operating conditions, the supply of alumina is at theoretical rate CT or waiting phase until what it finds normal operating conditions to go to phase 1.
Furthermore, if the supply phase 1 carried out in the normal framework of the regulation procedure is prolonged beyond a predetermined duration and if the number of release orders during this phase 1 exceeds a predetermined threshold of security, we detect that the bath is too rich in alumina and then we reduce very strongly or completely interrupt the supply of alumina to purge the bath of its excess alumina.

Conversely if the number of tightening orders during the same phase 1 exceeds a predetermined safety threshold, phase 2 is started whatever the values of the resistance slope and the extrapolated slope.

Finally if the curvature C (i) exceeds a predetermined safety threshold we initiates phase 2 of alumina supply whatever the values the resistance slope P (i) and the extrapolated slope PX (i).

• des modifications ont été apportées dans les méthodes de calcul des paramètres connus que sont R et P afin d'en augmenter la précision
• des paramètres complémentaires et nouveaux ont été mis en oeuvre pour en augmenter aussi la fiabilité.

In addition, when determining the adjustment parameters involved in the new regulation process:

• modifications have been made in the methods of calculating the known parameters that are R and P in order to increase the precision thereof
• additional and new parameters have been implemented to also increase reliability.

Thus for the calculation of the resistance R (i) at each end of the regulation cycle i duration T (between 10 seconds and 15 minutes) at the start of which gives any adjustment commands which modify the resistance level, divide the regulation cycle i into n elementary cycles of duration t (between 1 second and 15 minutes), we eliminate the first cycles during which the resistance level is modified by the anode frame setting operations and the average R (i) is calculated over the last n-a elementary cycles (a <n).

In this case, the average resistance r (k) of this elementary cycle is also calculated at the end of each elementary cycle k of duration t. These values r (k) are stored during the entire supply phase 1 for the calculation of the slope P (i) while keeping the last N values (N being a predetermined number).
Indeed, the resistance slope P (i), the extrapolated slope PX (i) and the curvature C (i) determined at the end of each regulation cycle i of duration T are calculated from the history of average resistances r (k) of the elementary cycles stored since the start of phase 1 of undernourishment within the limit of the last N values and this by any calculation method implementing a smoothing of the raw data r (k) with elimination of the variations of resistance due to adjustment orders of the anode frame.
The computation of the resistance slope and of the auxiliary parameters can be carried out by parabolic regression on the resistances, or by linear regression on the resistance variations, or by any other method equivalent to a nonlinear regression on the resistances.

Preferably, the method used to calculate the resistance slope P (i) consists of a linear regression on the variations of resistance or slopes instantaneous dr (k) = r (k) -r (k-1) which is calculated at the end of each cycle elementary k of duration t and after elimination of elementary cycles during which orders for adjusting the anode frame have been given. This linear regression on instantaneous slopes dr (k) is equivalent to a parabolic regression on the resistances r (k) after elimination of the variations resistance due to adjustment orders of the anode frame.

It should be remembered that the resistance changes along a curve and not along a straight line. The slope according to EP 044794 is calculated by performing directly a linear regression on the resistance values measured at regular interval. As the graph in Figure 3 shows, this leads to necessarily underestimate the real value of the slope. In addition, this default estimation error becomes all the more important as the curve evolution of R is more curved, that is to say that the resistance increases quickly. Thus according to EP 044794 when the resistance exceeds the high threshold of reference Ro + r of the regulation range, this variation can lead simply to give a tightening order to the anode frame and to extend feeding at a slow rate when the actual slope P (i) is in fact greater than the reference slope Po and an anode effect is then very close.

The new slope calculation method used in the implementation of the present invention is based on the principle of parabolic regression, which allows a much better approach to the actual rise curve resistance that a classic linear regression as shown in the diagram of Figure 3. If for considerations of complexity and means of calculation outside the scope of the invention, the applicant has not implemented exactly this type of regression for the slope calculation it nevertheless uses a method akin to a parabolic regression of calculating a linear regression line on the slopes instantaneous, and the value of the resistance slope P (i) is provided by the ordinate at time t (i) of the linear regression line on the slopes instant.

This new slope calculation procedure also provides additional and new information that is used as auxiliary adjustment parameters for optimizing content regulation alumina.

Knowledge of the linear regression line on instantaneous slopes allows to predict the value of the resistance slope for the cycle i + 1 or extrapolated slope PX (i) which is provided by the ordinate of the regression line extrapolated at time t (i + 1) = t (i) + T. This extrapolated slope value PX (i) is implemented to detect in advance a rapid rise in the resistance and decide to switch to the feed phase at a rate fast CR when this extrapolated slope PX (i) becomes greater than a slope extrapolated from PXo reference so that PX (i) ≥ PXo ≥ Po.

It is also very advantageous to use another auxiliary parameter which is the curvature C (i), i.e. the speed of evolution of the resistance slope P (i) given by the slope of the linear regression line on the slopes instantaneous, to trigger and modulate the supercharging itself according to the principle that a high curvature announces a sharp rise in the resistance. Thus exceeding the setpoint Co triggers a “CUR” ultra-fast rate feeding regime. For a curvature lower than Co the fast rate CR feeding regime ordered by the parameters P (i) and PX (i) is considered sufficient to cause R (i) to drop and avoid an anode effect.

Note that the reference thresholds Po, PXo and Co can take different predetermined or calculated values according to the operating conditions of the tank (acidity of the bath, temperature, resistance for example).

As an indication, for a 400,000 Amp (400 kA) tank, the value of the reference slope Po is between 10 and 150 pΩ / s, that of the extrapolated slope of reference PXo, is between 10 and 200 pΩ / s, and that of the reference curvature Co is between 0.010 and 0.200 pΩ / s ² ..

R'= R x 400/l'

P' = P x 400/l'

C'= C x 400/l'.

All these operating characteristics valid for a tank of intensity = 400 kA, are easily transposable to tanks of lower intensity knowing that the previous values of resistance R, slope P and curvature C can be defined in relative value compared to at the intensity I '<l going through these tanks so that

R '= R x 400 / l'

P '= P x 400 / l'

C '= C x 400 / l'.

The invention will be better understood from the detailed description of its use. implemented below.

EXAMPLE OF IMPLEMENTATION

The method according to the invention was implemented for several months on prototypes of electrolytic cell with precooked anodes fed under 400,000 Amps under the following conditions:

Alumina is introduced directly into the molten electrolysis bath in doses successive of constant mass by several insertion orifices, maintained permanently open by a crust breaker. For this purpose we will use advantageously a device for the occasional supply of alumina to electrolytic cells as described in EP 044794 (= US 4431491) or also in FR 2527647 (= US 4437964) in the name of the plaintiff.

The resistance R is calculated every tenth of a second from the intensity I and voltage U measurements at the terminals of the tank according to the conventional relationship: R ohm = U volt - 1.65 I Ampere

Po = 66 pΩ/s

PXo= 110 pΩ/s

Co = 0,065 pΩ/S²

An integrating computer makes it possible to determine the average values of the resistances r (k) every 10 seconds or instantaneous resistances r (k) within a regulation cycle i of duration T = 3 minutes and after elimination if necessary of the first values of the regulation cycle corresponding to the period of the setting orders of the anode frame which modify the resistance level, it calculates the average resistance R (k) of the cycle and the average slopes dr (k) = r (k) - r ( k-1) for the remaining duration of the cycle then determines by linear regression on the values dr (k) memorized since the start of phase 1 within the limit of the last N = 360 values, the slope P, the extrapolated slope PX and the curvature C = dP / dt. Then the comparison of the values P, PX and C thus calculated with the respective reference values causes the triggering, by means of the control-command chain, of the appropriate orders to the distributor-doser of alumina. These reference values are in this case:

Po = 66 pΩ / s

PXo = 110 pΩ / s

Co = 0.065 pΩ / S ²

CL cadence lente =   CT - 25% soit 173 Kg Al₂O₃/heure utilisée dans la phase d'alimentation 1.

CR cadence rapide =   CT + 25% soit 288 Kg Al₂O₃/heure

CUR cadence ultra-rapide = 4 CT soit 920 Kg Al₂O₃/heure utilisées dans la phase d'alimentation 2.

The average hourly consumption of alumina for a tank of 400,000 amperes is of the order of 230 kg of Al ₂ O ₃ / hour corresponding to the reference rate or theoretical rate of CT supply. In relation to this theoretical rate, we define for example:

CL slow rate = CT - 25% or 173 Kg Al ₂ O ₃ / hour used in feed phase 1.

CR fast rate = CT + 25% or 288 Kg Al ₂ O ₃ / hour

CUR ultra-fast rate = 4 CT or 920 Kg Al ₂ O ₃ / hour used in the feeding phase 2.

a) On a trouvé au terme du cycle i de durée T = 3 minutes

R(i) = 5,924 µΩ

P(i) = 26 pΩ/s

PX(i) = 31 pΩ/s

C(i) = 0,028 pΩ/s²

La phase 1 d'alimentation se poursuit.

b) Au terme du cycle i+1 les valeurs de P(i+1) et PX(i+1) restant inférieures aux seuils de référence Po= 65 pΩ/s et PXo = 110 pΩ/s, la phase 1 d'alimentation se poursuit.

c) Au terme du cycle i+2 on a trouvé :

R(i+2) = 5,936 µΩ

P(i+2) = 71 pΩ/s

PX(i+2) = 75 pΩ/s

C(i+2) = 0,022 pΩ/s²

ce qui déclenche le passage en phase 2 d'alimentation en cadence rapide CR pour une durée de 12 minutes (durée calculée proportionnellement à la pente au terme du cycle considéré selon la relation expérimentalement définie : durée en minutes = 0,083 x P(i) + 6 arrondie à la minute supérieure soit dans le cas présent : 0,083 x 71 + 6 ∼ 12 minutes).

d) La phase 2 d'alimentation se poursuit jusqu'au début du cycle i+7 où l'on repasse en phase 1 d'alimentation.

e) Au terme du cycle i+7 on trouve :

R(i+7) = 5,898 µΩ

P(i+7) = 7 pΩ/s

PX(i+7) = 10 pΩ/s

C(i+7) = 0,017 pΩ/s²

la phase 1 d'alimentation se poursuit.

f) Au terme des cycles i+8 et i+9 les valeurs des pentes P(i+8) et P(i+9) et des pentes extrapolées PX(i+8) et PX(i+9) restant inférieures à leur seuil de référence Po et PXo respectifs la phase 1 d'alimentation se poursuit.

g) Au terme du cycle i+10 on a trouvé :

R(i+10) = 5,917 µΩ

P(i+10) = 108 pΩ/s

PX(i+10) = 120 pΩ/s

C(i+10) = 0,067 pΩ/s²

la phase 2 d'alimentation est enclenchée avec tout d'abord alimentation en cadence ultra-rapide pendant une durée prédéterminée de 2 minutes (la durée d'alimentation en CUR est généralement fixée entre 1 et 5 minutes pour assurer un rechargement rapide du bain en alumine sans risquer toutefois sa saturation et par suite l'encrassement de la cuve). Après 2 minutes la phase 2 d'alimentation passe en cadence rapide pour une durée calculée de 15 min [ 0,083 x P(i+10) + 6 arrondi à la minute supérieure ].

h) Au bout de (2+15)= 17 minutes c'est-à-dire au cours du cycle i+16 on repasse en phase 1 d'alimentation.

i) Au terme du cycle i+16 les valeurs de P(i+16) et de PX(i+16) restant inférieures aux seuils de référence Po et PXo la phase 1 d'alimentation se poursuit et plus généralement la régulation du taux d'alumine dans le bain d'électrolyse selon les règles précédemment définies.

The tank being under normal operating conditions and the supply being in phase 1, a typical sequence for regulating the alumina level is as follows:

a) We found at the end of cycle i of duration T = 3 minutes

R (i) = 5.924 µΩ

P (i) = 26 pΩ / s

PX (i) = 31 pΩ / s

C (i) = 0.028 pΩ / s ²

Phase 1 of feeding continues.

b) At the end of cycle i + 1 the values of P (i + 1) and PX (i + 1) remaining below the reference thresholds Po = 65 pΩ / s and PXo = 110 pΩ / s, phase 1 of feeding continues.

c) At the end of cycle i + 2 we have found:

R (i + 2) = 5.936 µΩ

P (i + 2) = 71 pΩ / s

PX (i + 2) = 75 pΩ / s

C (i + 2) = 0.022 pΩ / s ²

which triggers the transition to phase 2 feeding at a rapid rate CR for a duration of 12 minutes (duration calculated in proportion to the slope at the end of the cycle considered according to the experimentally defined relationship: duration in minutes = 0.083 x P (i) + 6 rounded up to the nearest minute, i.e. in this case: 0.083 x 71 + 6 ∼ 12 minutes).

d) Phase 2 of feeding continues until the start of cycle i + 7 where we return to phase 1 of feeding.

e) At the end of cycle i + 7 we find:

R (i + 7) = 5.898 µΩ

P (i + 7) = 7 pΩ / s

PX (i + 7) = 10 pΩ / s

C (i + 7) = 0.017 pΩ / s ²

phase 1 of feeding continues.

f) At the end of cycles i + 8 and i + 9 the values of the slopes P (i + 8) and P (i + 9) and the extrapolated slopes PX (i + 8) and PX (i + 9) remaining less than their reference threshold Po and PXo respective phase 1 supply continues.

g) At the end of cycle i + 10 we have found:

R (i + 10) = 5.917 µΩ

P (i + 10) = 108 pΩ / s

PX (i + 10) = 120 pΩ / s

C (i + 10) = 0.067 pΩ / s ²

phase 2 of feeding is started with first of all supply in ultra-rapid rate for a predetermined duration of 2 minutes (the duration of feeding in CUR is generally fixed between 1 and 5 minutes to ensure rapid recharging of the bath in alumina without risking its saturation and consequently the clogging of the tank). After 2 minutes, phase 2 of feeding goes into rapid cadence for a calculated duration of 15 min [0.083 x P (i + 10) + 6 rounded up to the nearest minute].

h) At the end of (2 + 15) = 17 minutes, that is to say during the cycle i + 16, we return to phase 1 of feeding.

i) At the end of the cycle i + 16 the values of P (i + 16) and PX (i + 16) remaining below the reference thresholds Po and PXo phase 1 of supply continues and more generally the regulation of the rate alumina in the electrolysis bath according to the rules defined above.

The implementation of the process being thus specified, after more than 6 months application in 400,000 amp prototype tanks using a bath cryolite-based electrolysis containing 12% excess AIF3, therefore marked acid character, at a temperature of 950 ° C, the alumina content a been permanently set between 1.5% and 3.5% with a central value of 2.1%.

At the same time, the average Faraday yield was 95.6% and the effect rate anode of 0.018 EA / tank / day.

Claims (22)

1. A process for control of the alumina content of the bath in a cell for production of aluminum by electrolysis of alumina dissolved in a molten cryolite-base salt, the said process employing alumina feed at a rate modulated as a function of the value and change of the resistance R of the cell as calculated from the difference of electric potential measured at the cell electrode terminals, phases of alumina underfeeding with introduction of alumina at a slow rate CL (phase 1) being alternated with phases of alumina overfeeding with introduction of alumina at a fast rate CR or ultrafast rate CUR (phase 2) compared with a reference rate or theoretical rate CT corresponding to the mean theoretical rate of alumina consumption of the cell, characterized by control cycles of duration T, comprising the following sequence of operations in each cycle:

   (A) At the end of each control cycle i, the mean resistance R(i), the rate of change of resistance or resistance slope P(i) and the rate of change of the resistance slope or curvature C(i) are calculated and a prediction is made of the value of the resistance slope at time t(i+1) or extrapolated slope PX(i) = P(i) + C(i) x T, which is an estimate of the future resistance slope P(i+1) at the end of control cycle i+1.

   (B) The value R(i) is compared with a setpoint value Ro, and on this basis there are transmitted the following commands to move the anode frame position: shorten the anode-metal distance or "pot squeeze", or lengthen the anode-metal distance or "pot unsqueeze".

   (C) The alumina feed is controlled as a function of the values of the slope P(i), curvature C(i) and extrapolated slope PX(i) in such a way as to compensate for variations in alumina content by anticipating them.
2. A control process according to claim 1, characterized in that the alumina feed in stage (C) is controlled as a function of the values of slope P(i), curvature C(i) and extrapolated slope PX(i) relative to reference setpoints Po, Co and PXo.
3. A control process according to claim 1, characterized in that the alumina feed in stage (C) is controlled under the following conditions:

   If the alumina feed is in phase 1, the values P(i), C(i) and PX(i) are compared respectively with the reference setpoints Po, Co and PXo:

   If P(i) < Po and PX(i) < PXo, phase 1 continues;

   If P(i) ≥ Po or PX(i) ≥ PXo, a changeover to alumina feed phase 2 takes place:

   If C(i) ≥ Co, phase 2 begins with an ultrafast feed rate for a predetermined or calculated time, which is followed by feed at fast rate for a predetermined or calculated time, the calculation of times being performed as a function of the values calculated at the end of the previously defined control cycle;

   If C(i) < Co, the alumina feed changes directly to fast rate for a predetermined time or a time calculated as a function of the values calculated at the end of the previously defined control cycle.

   If the alumina feed is in phase 2:

   phase 2 continues normally for the predetermined time or the time calculated at the end of the preceding phase 1.
4. A control process according to claim 1, characterized in that the control procedure is authorized only when the cell is in normal operating conditions, or in other words is correctly controlled, stable and free of actions that would perturb operation or control, such as change of anode, tapping of metal or specific control procedures, and in that the control procedure begins with a phase 1 of alumina underfeeding.
5. A control process according to one of claims 1 to 4, characterized in that, at the end of alumina feed phase 2, the cell returns to phase 1, provided the cell is in normal operating conditions.
6. A control process according to one of claims 1 to 5, characterized in that, at the end of phase 2, the alumina feed changes over to theoretical rate or to stand-by phase if the cell is not in normal operating conditions, then resumes phase 1 as soon as the cell has recovered normal operating conditions.
7. A control process according to claims 1, 2 or 3, characterized in that, if the duration of phase 1 exceeds a predetermined time, and if the number of "pot unsqueeze" commands during this phase 1 exceeds a predetermined safety setpoint, it is detected that the bath is too rich in alumina, and so the alumina feed is reduced very drastically or is completely stopped in order to purge the bath of its excess alumina.
8. A control process according to claims 1, 2 or 3, characterized in that, if the number of "pot squeeze" commands during a same phase 1 exceeds a predetermined safety setpoint, alumina feed phase 2 is initiated regardless of the values of resistance slope and extrapolated slope.
9. A control process according to claims 1, 2 or 3, characterized in that, if the curvature exceeds a predetermined safety setpoint, alumina feed phase 2 is initiated regardless of the values of resistance slope and extrapolated slope.
10. A control process according to claim 1, characterized in that each control cycle i of duration T between 10 seconds and 15 minutes is divided into n elementary cycles k of duration t between 1 second and 15 minutes.
11. A control process according to claims 1 or 10, characterized in that the resistance R(i) calculated at the end of each control cycle of duration T is the mean resistance over the last n-a elementary cycles of the control cycle, i.e., the first a elementary cycles of the control cycle during which the control system can transmit commands to adjust the anode frame position to modify the resistance level are eliminated.
12. A control process according to claims 10 or 11, characterized in that the mean resistance r(k) of the elementary cycle is calculated at the end of each elementary cycle k of duration t, and in that the successive values r(k) are stored in memory.
13. A process according to claim 12, characterized in that the values r(k) are stored in memory during phase 1, subject to a limit of the last N values.
14. A control process according to claims 12 or 13, characterized in that the resistance slope P(i), extrapolated slope PX(i) and curvature C(i) determined at the end of each control cycle i of duration T are calculated from the history of the mean resistances r(k) of the elementary cycles by any method capable of smoothing the raw data r(k) while eliminating the resistance variations due to commands to adjust the anode frame position.
15. A control process according to claims 1 or 14, characterized in that the resistance slope P(i) and auxiliary parameters PX(i) and C(i) are calculated by parabolic regression over the resistances or by linear regression over the resistance variations, or by any other method equivalent to nonlinear regression over the resistances.
16. A control process according to claims 1, 14 or 15, characterized in that the method used for calculating the resistance slope P(i) and the auxiliary parameters consists of a linear regression over the instantaneous slopes dr(k) = r(k) - r(k-1) after elimination of the cycles during which commands to adjust the anode frame position were transmitted.
17. A control process according to claims 1 or 16, characterized in that the value of the resistance slope P(i) corresponds to the ordinate at the instant t(i) of the line of linear regression over the instantaneous slopes.
18. A control process according to claims 1 or 16, characterized in that the predicted value of the resistance slope for the cycle i+1 or extrapolated slope PX(i) corresponds to the ordinate of the regression line extrapolated to the instant t(i+1) = t(i) + T.
19. A control process according to claims 1 or 16, characterized in that the value of the curvature C(i) is given by the slope of the line of linear regression over the instantaneous slopes.
20. A control process according to claims 2 or 3, characterized in that the reference setpoints Po, PXo and Co may assume different predetermined values or values calculated according to the operating conditions of the cell.
21. A control process according to claims 2 or 3, characterized in that, for a cell operating at 400 kA, the reference slope Po is fixed between 10 and 150 pΩ/s, the extrapolated reference slope PXo is fixed between 10 and 200 pΩ/s and the reference curvature Co is fixed between 0.010 and 0.200 pΩ/s².
22. A control process according to claims 1, 2, 3 or 21, characterized in that the operating characteristics of resistance R, resistance slope P, extrapolated slope PX and curvature C, which are valid for a cell of current I = 400 kA, can be transposed to cells of lower or higher current I', according to the relationships:

    R' = Rx400/I'

    Pt = Px400/I'

    PX' = PX x 400 / I'
    and

    C' = Cx400/I'.

Images