US3345273A - Method of and apparatus for indicating anode positions - Google Patents

Description

@et 3, 1967 R. v. BROWN 3,345,273

METHOD OF AND APPARATUS FOR INDICATING NODE POSITIONS Filed Feb. 4, 1964 ROBERT V. BROWN ATTORNEYS United States Patent Oiice 3,345,273 Patented Oct. 3, 1967 3,345,273 METHOD OF AND APPARATUS FOR INDICATING ANODE POSITIONS Robert V. Brown, Portland, Oreg., assignor to Reynolds Metals Company, Richmond, Va., a

corporation of Delaware Filed Feb. 4, 1964, Ser. No. 342,505 7 Claims. (Cl. 204-1) ABSTRACT F THE DISCLOSURE The yrelative position of an anode and a cathode in a multiple anode reduction cell is indicated by an arrangement which measures low frequency voltage variations which are superimposed on the DC voltage applied across the cell. These variations are developed when the anode is too close to the cathode thereby causing overload of the anode.

This invention relates to a method of and apparatus for indicating the anode condition of an electrolytic reduction cell and more particularly to a method of and apparatus for detecting and indicating one or more low or overloaded, carbons or anodes in a multiple anode alumina reduction cell.

As is well known in the industry, carbon from the anode of a reduction cell is converted into carbon dioxide gas during the electro-chemical reduction process, so that the carbon blocksor anodes must be replaced periodically. Generally, it is presently the' practice to selectively replace one or more of the carbon blocks in each plural anode cell during each work shift. A carbon block replacement schedule is followed so that the average age of the carbon blocks in each section of the cell is maintained substantially constant. Where a pot line includes a large number of multiple anode cells, for instance one hundred and forty lcells each having twenty-four carbons, anode positioning and replacement is a constant costly process that permits much error and inefficient cell operation.

Generally two adjacent blocks of a cell are seldom of the same size and neither the stub nor ,the top ofan adjacent block can serve as a reference` mark whereby a new carbon block can be set to give the desired thickness of molten cryolite beneath the carbon blocks. The obvious method of setting a new carbon precisely according to the measured distance between the anode rod clamp and bottom of the old carbon is seldom used due to the extra time and labor required. Many times this method also requires repositioning and in general has proven neither economical nor accurate.

When the operator simply approximates the proper position when installing new carbons, satisfactory results are sometimes obtained although costly misplacements occur frequently and the cell operator must spend a lot of time searching for misplaced carbons. Additionally, after a carbon block or anode has been set, its position may be inadvertently disturbed during normal operation, and at times conditions arise within the cell that lead to waves in the bath or a humping or thickening of the molten aluminum layer under one or more carbons. This condition may drastically reduce the thickness of the molten cryolite layer under the carbon anodes-leading to an upset condition.

When a carbon is set high such that the layer of molten cryolite is thicker than the average, the carbon will not pick up its share of the current load and the pot voltage will tend to run high. Usually this does not create undesirable effectsin the cell. On the other hand, if a carbon is set low or if for any of the reasons previouslyV discussed, the thickness of the molten cryolite layer is reduced under a carbon, that carbon picks up more than its share of the current and thus becomes overloaded. If the maladjustment is relatively small the pot does not generally become unstable and the increased current load on the carbon tends to burn away the carbon more rapidly resulting in a self-adjustment of the carbon to the thickness of the cryolite layer. However, if the maladjustment is large the carbon may pick up two to four times its normal current load.

This may lead to numerous adverse effects. First of all, the additional current will tend to heat the carbon and the steel stub. As the stub becomes red it will become visible to the operator and may be noticed, in which case the operator will raise said carbon by a small amount. If the operator fails to detect the heated stubs of the low carbon, the heating process will continue until the iron lstub actually melts and the carbon is disconnected from the anode rod. Once this has occurred the operator may not notice the condition until the carbon is due for a changeout or he has occasion to raise the set of carbons in the cell. The heating of the stub to its melting point generally occurs at a much faster rate than can be matched by the inherent tendency to self-adjustment.

A second phenomenon may be experienced under low carbon conditions. The one low carbon drawing its large amount of current appears at times to create a disturbance in the liquid aluminum layer such that a number of neighboring carbons are periodically subjected to extreme overloads of current as though waves were flowing in the liquid aluminum layer. Under this condition a number of adjacent carbons may show over-heating and redness of the stubs, making it difficult for an operator to determine which carbon should. be raised in order to cure the trouble. The production rate of the cell is reduced during the period of the upset.

The inventor has discovered that the voltage wave form across a multiple-block-anode alumina reduction cell comprises a substantial direct current wave and a random low frequency alternating current wave of a fluctuating amplitude superimposed thereon that varies as a function of the degree of overloading of one or more carbons in the cell. Experiments clearly varify that when one or more of the carbon blocks are closer to the molten aluminum layer than the majority of blocks in the cellV the amplitude variations of the superimposed alternating curent (AC) are of sufficient magnitude to be detected and'indicated on a suitable meter. When all the carbon blocks are inproper relationship to the molten aluminum layer the AC signal strength is comparatively small and the reading on the meter is proportionately lower. This variation in amplitude of the AC signal may be used to accurately indicate when the individual carbon blocks of a cell should be adjusted, with the result that cell upsets due to low carbon blocks are prevented. As a result the cell requires less attention by the operator and operatesrnore eficiently.

The magnitude of the AC component may be sensed by various means and interpreted by the operator and/ or utilized to actuate alarm means. For example, an AC voltage component in a frequency range downwardly of about 20 cycles per second may be generated by a highly upset cell condition and cause a DC Voltmeter to oscilate over a range of 0.2 volt -or more, with the amplitude of the oscillations decreasing as the carbon adjustment is improved. Thus the amplitude of the oscillations may be observed by an operator on an expandedscale voltmeter, or an on-line digital computer may be used'to calculate the amplitude from a set of consecutive voltage readings and issue pre-selected instructions to an operator.

Accordingly it is an object of this invention to provide a method of and apparatus for indicating a low, or overloaded, carbon condition in a reduction cell utilizing an AC signal superimposed on the cell DC voltage wave form to actuate suitable indicator means. A further object of this invention is to provide a method of and apparatus for detecting low carbon anodes in a multipleanode-reduction cell by sensing a variation in the amplitude of an AC signal superimposed upon the DC voltage wave form of the cell, and utilizing said AC signal amplitude variations to indicate a low carbon condition.

A method of detecing low carbon anodes in a multiple anode reduction cell in accordance with this invention may include the steps of detecting the variations in amplitude of any AC voltage superimposed on said DC voltage; land relating said amplitude variations of said AC voltage to a low carbon anode condition in said cell.

The amplitude variations of said AC voltage caused by a low carbon anode may be directly observed on a suitable meter. Also, the AC voltage superimposed on the cell DC voltage may be amplified, integrated, and the resulting output utilized to actuate suitable loW carbon indicator means. Preferably, the frequency spectrum `of the cell voltage utilized is l-20 cycles, the remainder of the voltage signal being eliminated by standard low pass filters, with about 20 cycle cutoff which is of course variable and determined by the characteristics of the detection means. Any variation in amplitude of the AC voltage of l-20 cycles per second superimposed on the cell DC voltage is amplified and converted into a milliampere current proportional to the AC voltage component in the frequency range of 1-20 cycles. Of course, a broader or narrower frequency spectrum may be utilized if desired.

In accordance with the principles of this invention, apparatus for accomplishing this and many other objects may include means for measuring variations in the voltage of a reduction cell; means for amplifying and converting said variations in said voltage (within a frequency range of approximately 1 20 cycles per second) into a varying current proportional to said voltage variations; and utilizing said variations in current for indicating a low carbon `anode condition.

These and many other objects and advantages of this invention will become apparent from the following detailed description when read in view of the appended drawing wherein:

FIGURE 1 is a schematic illustrating a low carbon indicator arranged for use with a multiple-anode reduction cell.

FIGURE 2 is a wave diagram illustrating the reduction cell voltage Wave form and having an AC voltage component caused by a low carbon impressed thereon.

VFIGURE 3 is a wave diagram illustrating variations in an output current proportional to variations in the voltage wave of FIGURE 2.

Referring now to the drawings, an important type of alumina reduction cell is shown schematically in FIG- URE 1. This type of cell, known in the industry by various n-ames such as prebaked, Niagara, etc., is particularly distinguished from other types of reduction cells by virtue of the anode. The anode in this type of cell cornprise a plurality of carbon blocks 11 each of which are individually connected to the positive side of a source of direct current electricity, represented in FIGURE 1 as an anode bus 12 and the negative side of said source by a cathode bus 13. Each carbon block 11 is individually adjustable with respect lto a carbon cathode 14, and the set of carbon blocks that comprises the anode is vertically adjustable with respect to the cathode 14 in such a manner that the position of each block relative to the other blocks of the set is unchanged.

Each carbon block 11 is connected through an iron stub 16 cast in the block to a copper anode rod 17 which in turn is clamped to anode bus 12 by a hand operated clamp 18.`Anode bus 12 is supported at each end by a Ibridge jack 19 attached to :cell frame 21 whereby the anode bus 12 is raised or lowered with respect to the carbon cathode 14 and a layer of molten aluminum 22 which overlies the cathode 14, eifecting a uniform increase or decrease respectively in the thickness of a layer of molten cryolite 23 disposed between the neighorbing surfaces of the carbon blocks 11 and the molten aluminum. Individual carbon blocks 11 may be raised or lowered by an operator with respect to the anode bus 12 and cathode 14 after loosening the appropriate clamp 13 with a conventional hand jack (generally used by the industry) that operates between clamp 18 and rod 17.

The electrical circuit through the cell consists in sequence of the anode bus 12, the anode rod 17, the stub 16, carbon iblock 11, the molten cryolite layer 23, molten aluminum layer 22, carbon cathode 14, current collection means 24 and cathode bus 13. The anode bus 12 and cathode bus 13 are connected respectively through a rectifier or other suitable source of D.C. power (not shown). The major resistance to electrical current ow resides in the molten cryolite layer 23 and the total cell current distributes itself among the plurality of parallel-connected carbon blocks 11 in inverse relation to the thickness of the molten cryolite layer 23 between each of said blocks andthe molten aluminum layer 22. No satisfactory method of measuring the thickness of the cryolite layer under each carbon block exists and generally, prior to this invention, the uniformity of the cryolite layer has been simply inferred or guessed by the operator from observations of the condition of the cell and of the current flow in the plurality of anode rods.

A D.C. voltmeter 26, having an expanded scale, is connected across the anode bus 12 and cathode bus 13 by means of conductors 27 and 28 respectively and measures the reduction cell voltage which during periods of good adjustment (as shown in FIGURE 2) remains substantially constant. As one of the carbon blocks 11 which has been set low, such as block 50 of FIGURE 1, heats up and gradually becomes overloaded the cell voltage begins to fluctuate with increasing magnitudes that may reach several hundred millivolts and continue until the proper adjustment of the carbon is made. As reduction continues the cell works toward an anode effect at which time the cell voltage rises rapidly. An anode effect lamp 29 connected in parallel with the meter 26 indicates this condition.

In a typical 60,000 ampere, 24 carbon, reduction cell, the cell voltage is approximately 4.5 volts. The expanded scale meter is provided to permit simultaneously the measurement of the cell voltage and visual observation of the magnitude of any fluctuations in cell voltage particularly any voltage variations of a low frequency which have been found to be significant because directly `attributable to one or Imore low caribons in a cell. The low carbon condition leads to a current overload and t-hus cell voltage variations. For more accurate detection and indication of a low carbon, a conventional amplifier-integrating circuit designated as a low-carbon indicator 31 is connected in parallel with the lmeter 26 and measures the amplitude variations in the A C. voltage superimposed on the cell D.C. voltage Waveforms. As stated, Ithe A.C. voltage may vary from a few millivolts under good anode adjustment to several hundred millivolts under cell -upset conditions resulting from one or more low carbons.

By including amplied means in the detector 31 a signal may be provided to adjust the carbon while the eifect of the carbon maladjustment is still small and thus avoid an upset altogether. Further, an operator can readily determine from the strength of the signal at any time whether the cell carbon adjustment is excellent, good, or just satisfactory. Additionally, the signal is made more easily readable and interpretable by inclusion of integrating .means in the detector circuit to reduce pulsations arising from the low frequency portions of the voltage wave.

As stated, a very satisfactory signal can be obtained from a portion of the frequency spectrum of the cell voltage wave up to about cycles per second. A voltage wave of this frequency range is easily amplified and integrated to provide a highly responsive, substantially pulsation free, indicator signal. A conventional low-pass filter 30 is used to select the desired signal since a sharply dened frequency cut-off is not needed to insure elimnation of su-bstantially all of the 60 cycle and higher components of the voltage wave.

Utilizing conventional circuitry the indicator 31 may also convert the ce'll voltage wave in the desired frequency range of up to 20 cycles per second into an output current in milliamps proportional to the amplitude of the A.C. component in said frequency range. This wave form is shown in FIGURE 3 and may be recorded on a strip chart. This output current is measured by an expanded scale ammeter 32 that also turns on a low carbon light 33 connected in series with a suitable voltage source when the output current exceeds a determined value. Referring to FIGURE 3, the output current remains substantially constant during periods of good carbon adjustment fbut rises rapidly during poor carbon adjustment or when the cell goes into an anode effect. It is `to be noted that both lamps 29 and 33 are turned on during an anode effect while only lamp 33 is turned on during periods of low carbon upset and that lamp 33 goes out after the low carbon is properly adjusted.

Although a preferred embodiment of the invention has been described in detail, the invention is subject to many modifications and changes within the principles of the invention which is to be limited only by scope of the appended claims.

What is claimed is:

1. A method for determining the presence of a low carbon anode in an electrolytic reduction cell for the reduction of aluminum dissolved in a molten electrolyte the cell having multiple carbon anodes and a cathode to which DC voltage is applied, the method comprising the steps of:

electrically connecting across the anodes and cathode detection means sensitive to low frequency voltage variations in the neighborhood of 1 to 20 cycles per second superimposed on the applied DC voltage; and utilizing said detection means to indicate the occurrence of said voltage variations caused by a low carbon.

2. The method of claim 1, wherein the indicating is proportional to voltage variations exceeding a predetermined value.

3. An indicator for determining a low anode condition in an electrolytic reduction cell having relatively adjustable anode and cathode elements across which a DC voltage is applied, the indicator comprising:

' detection means to be connected electrically across s-aid anode and cathode elements and sensitive to low frequency voltage variations in the neighborhood of 1 to 20 cycles per second superimposed on the applied DC voltage; and

a display device responsive to said detection means to indicate the occurrence of voltage variations caused by a low anode condition.

4. An indicator as set forth in claim 3, wherein said detection means includes:

means for amplifying and converting said voltage variations into a varying current proportional to the voltage variations; and wherein said display device is responsive to the varying current to provide an output indicative of a low anode condition.

5. An indicator -as set forth in claim 4, wherein said display device is actuated when said varying current exceeds a predetermined Value.

6. A low carbon indicator for an electrolytic reduction cel-l having a plurality of carbon anodes adjustably positioned relative to a cathode and across which anodes and cathode DC voltage is applied, the indicator comprising:

detection means to be connected electrically across the anodes and cathode and sensitive to low frequency voltage variations in the neighborhood of l to 20 cycles per second superimposed on the applied DC voltage,

said detection means including a low-pass filter and means for amplifying and converting the voltage variations at the output of said filter into a varying current proportional to the variations of the filter output,

and a ldisplay device responsive to the varying current to provide an output indicative of a low carbon condition.

7. Apparatus for sensing low frequency voltage variations arising during operation of an electrolytic reduction cell and indicative of an upset condition therein, said cell having relatively adjustable anode and cathode elements across which a DC voltage is applied, comprising:

detection means to be connected electrically across said anode and cathode elements, said detection means being sensitive to voltage variations in the neighborhood of 1 to 20 cycles per second superimposed on the applied DC voltage; and

a display device responsive to said detection means to indicate the occurrence of said voltage variations,

References Cited kUNITED STATES PATENTS 2,904,490 9/ 1959 Hanssen 204-244 2,93 0,746 3/ 1960 Cooper 204-244 2,93 3,440 4/ 1960 Greenfield 204t7 JOHN H. MACK, Primary Examiner. T. TUNG, Assislant Examiner1

Claims (1)

1. A METHOD FOR DETERMINING THE RESENCE OF A LOW CARBON ANODE IN AN ELECTROLYTIC REDUCTION CELL FOR THE REDUCTION OF ALUMINUM DISSOLVED I A MLTEN ELECTROLYTE THE CELL HAVING MULTIPLE CARBON ANODES AND A CATHODE TO WHICH DC VOLTAGE IS APPLIED, THE METHOD COMPRISING THE STEPS OF: ELECTRICALLY CONNECTING ACROSS THE ANODES AND CATHODE DETECTION MEANS SENSITIVE TO LOW FREQUENCY VOLTAGE VARIATIONS IN THE NEIGHBORHOOD OF 1 TO 20 CYCLES PER SECOND SUPERIMPOSED ON THE APPLIED DC VOLTAGE; AND UTILIZING SAID DETECTION MEANS TO INDICATE THE OCCURENCE OF SAID VOLTAGE VARIATIONS CAUSED BY A LOW CARBON.

Images