Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
AN ARC SMELTING SYSTEM AND METHOD OF MONITORING THE LENGTH OF AN ELECTRODE IN SAID SYSTEM
Document Type and Number:
WIPO Patent Application WO/2017/182902
Kind Code:
A1
Abstract:
The invention relates to an arc smelting system 10 and a method of monitoring, or obtaining an indication of, a length of an electrode of the arc smelting system 10. The system 10 includes at least one electrode; a current measuring arrangement which is configured to measure the current in the electrode during operation; and a processing arrangement. The processing arrangement is configured to determine a variation of the current in the electrode measured by the current measuring arrangement over a period of time, and derive a current variation value which is indicative of a length of the electrode. The derivation of the current variation value is based on at least the determined variation of current over the period of time.

Inventors:
LE ROUX DANIEL JACUES (ZA)
Application Number:
PCT/IB2017/051903
Publication Date:
October 26, 2017
Filing Date:
April 04, 2017
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
GLENCORE OPERATION SOUTH AFRICA (PROPRIETARY) LTD (ZA)
International Classes:
H05B7/152; H05B7/09
Foreign References:
JPS57130398A1982-08-12
KR20140007083A2014-01-17
JPH038290A1991-01-16
CN204679109U2015-09-30
JPH06260281A1994-09-16
US7095777B22006-08-22
US3872231A1975-03-18
US6178191B12001-01-23
Other References:
HOCKADAY, C., ELECTRODE LENGTH DETERMINATOR [INTERVIEW, 29 December 2015 (2015-12-29)
Attorney, Agent or Firm:
SPOOR & FISHER et al. (ZA)
Download PDF:
Claims:
CLAIMS

1. A method of monitoring, or obtaining an indication of, a length of an electrode of an arc smelting system, wherein the method includes:

measuring the current in/running through the electrode during operation;

determining, by using a processor, a variation of the measured current over a period of time;

deriving, by using a processor, a current variation value which is indicative of the electrode length, wherein the derivation of the current variation value is based on at least the determined variation of current over the period of time.

2. The method of claim 1 , which is for monitoring the lengths of, or obtaining an indication of the lengths of, two electrodes of an arc smelting system, wherein the method includes:

measuring the current in each electrode during operation;

determining a variation of the measured current for each electrode over a period of time;

deriving a current variation value for each electrode, wherein the derivation of the current variation value is based on at least the determined variation of current for the particular electrode over the period of time.

3. The method of claim 1 which includes comparing, by using a processor, the derived current variation value with a reference current variation value, in order to obtain an indication of the electrode length

4. The method of claim 3, wherein the reference current variation value is a current variation value for an electrode with a known length.

5. The method of claim 3, which includes

measuring the current in the electrode at specific time instances/intervals over a period of time; and

determining, by using a processor, the variation in current by calculating the difference in current measured at successive time instances/intervals.

6. The method of claim 5, wherein the derived current variation value is calculated by using the following formula:

wherein:

Ivaris the derived current variation value;

n is the total amount of data points from which the derived current variation value is to be calculated;

/i is the measured current at time interval/instance /; and

tn-to refers to the time span/period over which the data points were recorded.

7. The method of claim 6, wherein the arc smelting system is a

ferrochrome smelter.

8. The method of claim 5, which includes not deriving any current variation values during a time period/instance when a tap change at a transformer, which is operatively connected to the electrode, occurs.

9. An arc smelting system which includes:

at least one electrode;

a current measuring arrangement which is configured to measure the current in the electrode during operation; and

a processing arrangement which is configured to

determine a variation of the current in the electrode measured by the current measuring arrangement over a period of time, and

derive a current variation value which is indicative of a length of the electrode, wherein the derivation of the current variation value is based on at least the determined variation of current over the period of time.

10. The system of claim 9, which includes two electrodes.

1 1. The system of claim 10, wherein the current measuring arrangement is configured to measure the current in each electrode during operation.

12. The system of ciaim 11 , wherein the processing arrangement is configured to

determine a variation of the measured current for each eiectrode over a period of time; and

derive a current variation vaiue for each electrode which is indicative of the length of the eiectrode, wherein the derivation of the current variation vaiue is based on at feast the determined variation of current over the period of time.

13. The system of ciaim 9, wherein the processing arrangement is configured to identtfy/determine a change in eiectrode length of the eiectrode by comparing the derived current variation value with a reference current variation vaiue.

14. The system of ciaim 9, wherein the current measuring arrangement is configured to measure the current in the eiectrode at specific time

instances/intervals over a period of time.

15. The system of claim 14, wherein the processing arrangement is configured to determine the difference/amount of variation in current by calculating the difference in current measured at successive time

instances/intervals.

16. The system of claim 15, wherein the current variation vaiue is calculated by using the following formula:

wherein:

is the derived current variation value;

n is the total amount of data points from which the derived current variation value is to be calculated;

recorded.

17. The system of claim 16, wherein the system is a ferrochrome smelter.

18. The system of claim 16, wherein the system is an electric arc furnace.

19. A method of identifying/determining a change in electrode length of an electrode of an arc smelting system, wherein the method includes:

measuring the current in the electrode during operation;

determining, by using a processor, a variation of the measured current over a period of time;

deriving, by using a processor, a current variation value which is indicative of the electrode length, wherein the derivation of the current variation value is based on at least the determined variation of current over the period of time; and

comparing, by using a processor, the derived current variation value with a reference current variation value.

20. An arc smelting system which includes:

at least one electrode;

a current measuring arrangement which is configured to measure the current in the electrode during operation; and

a processing arrangement which is configured to

identify/determine a change in electrode length of the electrode by

determining a variation of the measured current over a period of time,

deriving a current variation value which is indicative of a length of the electrode, wherein the derivation of the current variation value is based on at least the determined variation of current over the period of time, and

comparing the derived current variation value with a reference current variation value.

Description:
AN ARC SMELTING SYSTEM AND METHOD OF MONITORING THE LENGTH OF AN ELECTRODE IN SAID SYSTEM

BACKGROUND OF THE INVENTION

THIS invention relates to an arc smelting system and a method of identifying/determining a change in electrode length of an electrode which forms part of the arc smelting system.

In the smelting process, a carbon based electrode is submerged in the burden with a tip of the electrode at an unknown depth. The purpose of the electrodes is to conduct electricity and provide the necessary energy required by the process. One or more electrodes can be used in the process. The electrodes can either be pre-manufactured and replaced as needed or continuously formed (known as Soderberg electrodes). Effective replenishment of the electrodes is a critical function and affects the stability and efficiency of the production process (electrodes are carbon based and therefore also participate in the reaction). An ideal electrode length exists which not only ensures effective and efficient smelting, but also continuance of the electrode forming process by maintaining appropriate average electrode currents (in the case of Soderberg electrodes).

Maintaining optimal electrode lengths have proven to be one of the most difficult (and potentially performance inhibiting) activities faced daily by production personnel. Due to the nature of the electrical circuit employed, it is difficult to determine electrode lengths by observing "dashboard" parameters such as electrode current and resistance. Only in extreme cases of electrode breakages (i.e. shortening of the electrode), will these parameters give an indication of the occurrence, but using them to maintain lengths have however proven to be a futile exercise. Various attempts have been made at solving this problem, with none of them providing an acceptable solution (typically either due to the financial implications, modifications needed to be made to existing installations and/or the accuracy of the method).

Numerous methods are based on measuring the weight of the electrode. These methods are based on the fact that a longer electrode will necessarily weigh more than a shorter electrode and accordingly the exact length can be calculated assuming a certain electrode mass density. Since its advent, the method has also been improved upon by eliminating potential sources of variability. Initially the weight of the entire electrode column was used. This method was improved on by eliminating weight disturbances arising from components associated with the electrode mantel (flexible bus tubes, friction with roof seals, water flow etc.), by placing load cells beneath the slipping clamps, thereby isolating the weight of the electrode (steel casings, paste and baked electrode section) (Sidorski, E., 2006. Arc furnace electrode length determination. United States of America, Patent No. 7095777). A further alteration of this method is to compensate for plasma forces by taking into account electrode currents (Allen, C. & Motter, J., 1975. System for determining electrode length. United Sates of America, Patent No. US3872231).

An alternative method is to make use of a consumption model. When used in conjunction with daily slipping values the theoretical electrode length can be determined. Mintek uses such a method to determine electrode lengths (their consumption model is based on power input) (Hockaday, C, 2015. Electrode length determinator [Interview] (29 December 2015)). The accuracy of the model is obviously dependent on the accuracy of consumption data. Consumption varies with excess carbon conditions, slag regime etc., making this method very inaccurate. This method is also incapable of detecting electrode breakages.

A patent exists where the length is determined using the potential difference between two points on the furnace shell (one point being as close as possible to the electrode being measured) and the electrode current (Kalgraf, K., Morkesdai, G. & Tronstad, R., 2001. Method for determination of the tip position of consumable electrodes used in electric smelting furnaces. United States of America, Patent No. 6178191).

The inventor wishes to address at least some of the problems identified above. SUMMARY OF THE INVENTION in accordance with a first aspect of the invention there is provided a method of monitoring, or obtaining an indication of, a length of an electrode of an arc smelting system, wherein the method includes:

measuring the current in/running through the electrode during operation (i.e. during an arc smelting operation);

determining a variation of the measured current over a period of time; deriving a current variation value which is indicative of the electrode length, wherein the derivation of the current variation value is based on at least the determined variation of current over the period of time.

The electrode is typically elongated, e.g. it may have an elongate body.

The method may more specifically be for monitoring the lengths of, or obtaining an indication of the lengths of two, preferably three, electrodes of an arc smelting system. The method may then include:

the measuring the current in each electrode during operation;

determining a variation of the measured current for each electrode over a period of time;

deriving a current variation value for each electrode, wherein the derivation of the current variation value is based on at least the determined variation of current for the particular electrode over the period of time.

The method may include comparing, preferably by using a processor, the derived current variation value with a reference current variation value, in order to obtain an indication of the electrode length. The method may include determining the reference current variation value. The reference current variation value may be a current variation value for an electrode with a known length. The reference current variation value is hereinafter referred to as the "setpoint value".

In accordance with a second aspect of the invention there is provided a method of identifying/determining a change in electrode length of an electrode of an arc smelting system, wherein the method includes:

measuring the current in the electrode during operation;

determining, preferably by using a processor, a variation of the measured current over a period of time;

deriving, preferably by using a processor, a current variation value- which is indicative of the electrode length, wherein the derivation of the current variation value is based on at least the determined variation of current over the period of time; and

comparing, preferably by using a processor, the derived current variation value with a reference current variation value ("setpoint value").

The method, in accordance with the first or second aspect of the invention, may more specifically include measuring the current in the electrode at specific time instances/intervals over a period of time. The method, in accordance with the first or second aspect of the invention, may then further include determining, preferably by using a processor, the variation in current by calculating the difference in current measured at successive time instances/intervals.

The derived current variation value, in accordance with the first or second aspect of the invention, may be calculated by using the following formula:

wherein:

War is the derived current variation value;

n is the total amount of data points from which the derived current variation value is to be calculated;

/ i is the measured current at time interval/instance /; and

t n -t 0 refers to the time span/period over which the data points were recorded.

It should be clear that "data points" refer to the time instances when the current in the electrode was measured.

The determining of the variation of the measured current may be implemented by a processor. The processor may form part of a computing arrangement such as a computer, or it may be a programmable logic controller (PLC).

The derived current variation value, in accordance with the first or second aspect of the invention, may be calculated by a processor. The processor may form part of a computing arrangement such as a computer, or it may be a programmable logic controller (PLC). The determining of the variation of the measured current and the calculation of the derived current variation value may be implemented by the same processor.

The arc smelting system mentioned above may be an arc smeiter. The arc smelter may be a ferrochrome smeiter.

The arc smelting system may be an electric arc furnace.

The method may include not deriving any current variation values during a time period/instance when a tap change at a transformer, which is operativeiy connected to the electrode, occurs. More specifically, the method may include not deriving a current variation value during a time period when a tap change at a transformer, which is operativeiy connected to the electrode, is occurring. in accordance with a third aspect of the invention there is provided an arc smelting system which includes:

at least one electrode;

a current measuring arrangement which is configured to measure the current in the electrode during operation; and

a processing arrangement which is configured to

determine a variation of the current in the electrode measured by the current measuring arrangement over a period of time, and

derive a current variation value which is indicative of the electrode length, wherein the derivation of the current variation value is based on at Ieast the determined variation of current over the period of time.

The electrode is typically elongated, e.g. it has an elongate body.

The system may include two, preferably three, electrodes. The current measuring arrangement may be configured to measure the current in each electrode during operation. Alternatively, the system may include three current measuring arrangements, one for each electrode.

The processing arrangement may be configured to

determine a variation of the measured current for each electrode over a period of time; and derive a current variation value for each electrode which is indicative of the length of the electrode, wherein the derivation of the current variation value is based on at least the determined variation of current over the period of time.

The processing arrangement may be configured to identify/determine a change in electrode length of the electrode by comparing the derived current variation value with a reference current variation value ("setpoint value"). The reference current variation value ("setpoint value") may be a current variation value for an electrode with a known length. The processing arrangement may be configured to determine a reference current variation value ("setpoint value"). The reference current variation value ("setpoint value") may be a current variation value for an electrode with a known length.

In accordance with a fourth aspect of the invention there is provided an arc smelting system which includes:

at least one electrode;

a current measuring arrangement which is configured to measure the current in the electrode during operation; and

a processing arrangement which is configured to

identify/determine a change in electrode length of the electrode by

determining a variation of the measured current over a period of time,

deriving a current variation value which is indicative of the electrode length, wherein the derivation of the current variation value is based on at least the determined variation of current over the period of time, and

comparing the derived current variation value with a reference current variation value ("setpoint value").

In accordance with the third or fourth aspect of the invention, the current measuring arrangement may be configured to measure the current in the electrode at specific time instances/intervals over a period of time. The processing arrangement may then be configured to determine the difference/amount of variation in current by calculating the difference in current measured at successive time instances/intervals. The current variation value, in accordance with the third or fourth aspect of the invention, may be calculated by using the following formula:

wherein:

Ivar is the derived current variation value;

n is the total amount of data points from which the derived current variation value is to be calculated;

/,· is the measured current at time interval/instance /; and

t n -to refers to the time span/period over which the data points were recorded.

The processing arrangement may include a processor. The processing arrangement may include a computing arrangement such as a computer, and/or it may include a programmable logic controller (PLC).

The arc smelting system, in accordance with the third or fourth aspect of the invention, may be an arc smelter. The arc smelter may be a ferrochrome smelter

The arc smelting system, in accordance with the third or fourth aspect of the invention, may be an electric arc furnace.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention will now be described, by way of example, with reference to the accompanying diagrammatic drawings. In the drawings:

Figure 1 shows a schematic layout of a hardware configuration of an arc smelting system in accordance with the invention;

Figure 2a shows a graphical representation of a calculated current flowing through an arc smelting system (i.e. a so-called electrode current) over a period of time, for a 2.3m electrode;

Figure 2b shows a graphical representation of a calculated current flowing through an arc smelting system (i.e. a so-called electrode current) over a period of time, for a 3.0m electrode;

Figure 3a shows a graphical representation of resistance versus current variation through a first electrode of an arc smelting system; Figure 3b shows a graphical representation of resistance versus current variation through a second electrode of an arc smelting system; Figure 3c shows a graphical representation of resistance versus current variation through a third electrode of an arc smelting system;

Figure 4a shows a graphical representation of a change in current variation in a first electrode as a function of change in

Figure 4b shows a graphical representation of a change in current variation in a second electrode as a function of change in the amount of tap changes

Figure 4c shows a graphical representation of a change in current variation in a third electrode as a function of change in the amount of tap changes

Figure 5a shows a graphical representation of a time-based comparison between a normal current variation and a reference state current variation model for electrode 1 ;

Figure 5b shows a graphical representation of the effect of tap changes on current variation;

Figure 5c shows a graphical representation of a functioning of a reference state current variation model;

Figure 6a shows a graphical representation of the impact tap changes have on an absolute difference between subsequent electrode current values and

Figure 6b shows a graphical representation of the impact tap changes have on an absolute difference between subsequent electrode current values

when calculations are not made during tap changes. DESCRIPTION OF PREFERRED EMBODIMENTS

Reference is again made to the drawbacks mentioned in the background of the invention. Due to the absence of anything more accurate, average resistance and current readings have been used to gauge electrode lengths, even though it was not accurate. From a control point of view, slipping rates were adjusted to maintain certain average electrode currents or resistance set points, in times of furnace instability, it was not uncommon for production personnel to melt down the furnace bed to expose the electrode tips and visually measure the lengths, more often than not being surprised by the actual lengths.

From these experiences, the observation has been made that although the average current in the electrodes are the same, the way in which the particular average is reached differs. It was found that a longer electrode tends to operate in a more stable manner when observing electrode current with only minor fluctuations in the value. Shorter electrodes on the contrary exhibit much more volatility. Figures 2a and 2b provide an illustration of this occurrence. Furthermore, it has also found that there is no real relationship between current variation and average current/resistance (see Figures 3a-c).

The present invention utilises the volati!ity aspect mentioned above, in order to monitor the lengths of electrodes in an arc smelting system.

The present invention therefore relates to an arc smelting system, more specifically a ferrochrome smelter, which is configured to monitor the length of one or more electrodes used in the smelting process, in order to monitor/obtain an indication of the length of the electrodes, a specific algorithm, which takes into account the variation of current through the electrode over time, is implemented by a processing arrangement, in order to calculate a specific current variation value (which may also be referred to as the " electrode current variation").

The calculated current variation value is then compared against a setpoint value. The setpoint valuetypically refers to an ideal/optimal value for the calculated current variation value when the electrode is at a specific/desired length. The setpoint valueis typically obtained by using the system to calculate the current variation value when the length of the electrode is known (e.g. when a new electrode is first used in the smelting system or after the length of an electrode has been physically measured). Once the setpoint vaiuehas been caicu!ated, this value can be compared against any calculated current variation values which are calculated in the future during the operation of the smelting system.

!n the drawings, reference numeral 10 refers generally to an arc smelting system (e.g. a ferrochrome smelter) in accordance with the invention. Three single phase transformers 12.1-12.3 are used to deliver power to three electrodes located in a submerged-arc furnace of the system 10. It will however be appreciated that any number of electrodes can be used. Each electrode is typically elongated (i.e. it has an elongate body).

Multi ratio current transformers 14.1-14.3 are operatively connected on the primary side of furnace transformers 12.1-12.3 using reference current transformers 11.1 - 11.3. This configuration of transformers (11.1 - 11.3 and 14.1 - 14.3) allows for the conversion of primary currents at 16.1 - 16.3 to an indication of electrode currents.

Transformers 12.1 - 12.3 and 14.1 - 14.3 are multi-tap transformers. Tap positions on transformers 12.1 - 12.3 and 14.1 - 14.3 are controlled in unison with a motor-driven tap changer.

The current transformers 14.1-14.3 are connected in a delta configuration (which may be similar to the connection between the electrodes).

The current transformers 14.1-14.3 are each connected to a current transducer (e.g. 4-20mA) 20.1-20.3 which is configured to generate an analogue stgnal which is used as an input to a programmable logic controller (PLC) 22. More specifically, the transducers 20.1-20.3 are each connected to an input channel 19.1-19.3 of the PLC 22. A digital output of the PLC 22 is connected to a computer 24 (or other computing unit), e.g. via an Ethernet connection.

It will be appreciated that the multi-ratio current transformers 14.1-14.3, reference current transformers 11.1 - 11.3 and associated transducers 20.1- 20.3 together form a current measuring arrangement which is configured to measure the current in the electrodes during operation and relay the details thereof to the PLC 22 for further processing. It will be appreciated that alternative current measuring arrangements could be implemented in order to measure the current in the electrodes. Furthermore, the current measuring arrangements could be connected directly to a computer instead of a PLC. Either the PLC or the computer 24 can then be used in order to calculate the said current variation value for each electrode. The PLC 22 and/or computer 24 can therefore effectively form a processing arrangement which is configured to calculate the said current variation value for each electrode and to monitor/obtain an indication of the length of the electrodes, based on the said calculation.

The current variation value for each electrode can be calculated by using the following formula:

wherein:

Ivar is the derived/calculated current variation value;

n is the total amount of data points from which the derived current variation value is to be calculated;

/,· is the measured current at time interval/instance /; and

t n -to refers to the time span/period over which the data points were recorded.

From the above equation it can be noted that the current variation value is time based. Furthermore, the absolute differences are first calculated and then divided by the total time in which the data collection transpired. This method may typically be applicable for various sampling methods, which depend on the type of PLC used. Averaging instantaneous values will typically only be accurate when sampling at a fixed rate.

Once the current variation value for a particular electrode has been calculated, the PLC 22 or the computer 24 derives an indication of the electrode length by comparing the calculated current variation value with a setpoint value. As mentioned above, the setpoint valuecan typically be obtained by calculating a current variation value in the same manner when the electrode length is known, e.g. at the beginning when a new electrode is used in the system 10. This reference value may differ from system to system as well as from PLC to PLC. It is advised that in determining a reference value that data measured over a time will be used and optimized for a specific system 10. In the case of a short electrode, due to the volatility of the current variation value, the numerator of the equation becomes a large value and subsequently also the current variation value.

Experiment

Investigation Methodology

The following investigation methodology was followed for data collection, analysis and results verification. In this experiment, the effect of electrode length was eliminated by assuming the length remains constant from one 30 minute period to the next. The effect of tap changes on current variation was then analysed from one 30m in interval to the next.

Data Collection

Data from a specific furnace was obtained for a period of 13 days. For every 30 minutes in this time interval {624 data sets in total) the following data were captured:

● Current variation for each electrode; and

● The amount of tap changes per transformer.

Data Analysis

A filter was placed on an average power input of the system (e.g. a 40MW filter) to eliminate periods where the furnace was off, since it could potentially have a significant effect on the current variation value.

● Current variation on electrode E1 was correlated to the sum of tap changes at Transformer 3 and 1 (Denoted as TX312).

Current variation on electrode E2 was correlated to the sum of tap changes at Transformer 1 and 2 (Denoted as TX123).

● Current variation on electrode E3 was correlated to the sum of tap changes at Transformer 2 and 3 (Denoted as TX231).

Model Verification

From the data analysts a straight-!ine regression fit (see equation 2 below) was determined. The line passes through the origin, since no change in the factor being investigated will necessarily result in no change to the current variation:

For any instance (/) the actual current variation value can now be adjusted, using the known relationship between current variation and the number/sum of the tap changes, to represent the "adjusted current variation value".

If Ivar.o exhibit less fluctuations from one period to the next, it means it is consistently giving a more reliable result. The degree of fluctuation is calculated using the following formula (sum of absolute changes between intervals):

Equation 5: Calculation of Ivar fluctuation to compare models

The range of the current variation (difference between maximum and minimum value for a specific interval) was also used as an additional measure of accuracy. The smaller the range, the narrower the band within which the value was fluctuating.

Equation 6: Calculation of Ivar range for a specific interval

Results

As illustrated in Figures 4a-c, the results indicated that there is a correlation between the amount of tap changes being made and the fluctuations in the current variation value. As shown in these figures, a straight-line regression fit was obtained for each electrode (see the dotted lines and corresponding equations in the top right corners of the graphs), !n this regard, a good correlation can be noted at all three electrodes, with very similar slopes in all instances.

The correlation between the change in total tap changes and change in current variation can be used to derive an improved model (referred to as the "adjusted state current variation model"). When forcing the individual regression lines through the origin, the accompanying slopes are 0.0014, 0.0016 and 0.0015 respectively, which indicates that all three current variations react similar to changes in the amount of tap changes per interval. For this experiment, it was decided to make use of the average of ail three electrodes (i.e. 0.0015).

For comparison sake, the aim was to have the adjusted current variation value as close as possible to what we have become accustomed to. For that reason the average amount of tap changes per 30min interval (per transformer pair) was first calculated for the data collected and served as the reference state. The average amount of tap changes was calculated as 17 and was used to substitute If, for example, the total amount of tap changes equated to 40, then Δχ would be 23. Equation 4 was then used to calculate the adjusted current variation value (on a 30min basis) for all the data collected. Table 1 below summarizes the daily averages. With regards to equation 4, and keeping in mind that an increased amount of tap changes result in a higher calculated value, the mdX term adjusts the actual value to an "adjusted value (i.e, what would the have been, had there

only been 17 tap changes). The second and third columns for each electrode provide an indication of fluctuations observed (calculated using equations 5 and 6, respectively.

Table 1 : Current Variation: Normal current variation value (i.e. lvar as calculated by equation 1) versus the adjusted state current variation model

refers to the range of the adjusted current

variation value in electrode 1 (i.e. the maximum adjusted current variation value minus the minimum adjusted current variation value);

It should be appreciated that the same general definitions/references also apply to the six columns under electrodes 2 and 3, respectively.

Whenever the adjusted state current variation mode! outperformed the normal measurement (in terms of fluctuation), the cell in Table 1 is colored grey, in roughly 94% of the cases the adjusted state current variation model decreased the amount of fluctuations observed within the data. Figure 5a shows the actual current variation values for the 13 th day of the experiment, as well as the adjusted current variation values for the same intervals.

Figure 5b provides a graphical illustration of the effect of increased/decreased tap changes on normal current variation To remove some of the noise, a 4hr running average was used on the 30min- interva) data (electrode 3 was used in this example). It should be clear that the amount of tap changes has a significant effect on the normal current variation value.

Figure 5c illustrates the working of the adjusted state current variation model As the amount of tap changes increase, the normal current variation value is forced downwards (i.e. via the adjusted state current variation model) to compensate for the increased tap actions - which results in fewer fluctuations and a narrower operating range.

The adjusted state current variation mode!, which implements the correlation between the amount of tap changes and current variation, therefore improved the accuracy and reliability of the current variation value.

In another example, the instances where tap changes take place can be disregarded/removed from the calculations, in order to help improve the accuracy of the derived/calculated current variation value or the adjusted state current variation values. The system 10 may therefore be configured to remove/disregard data obtained during a tap change, when calculating a current variation value(s). In an alternative example, the system 10 may be configured to not take/log/process current measurements during a tap change. More specifically, the computer 24 or PLC may be configured to not take/log/process current measurements during a tap change. For example an execution of a program stored on the computer 24 or PLC, which implements the calculations, can be paused, while a tap change is occurring. Reference is in this regard specifically made to Figures 6a and 6b. Figure 6a shows the effect tap changes have on an absolute difference between subsequent eiectrode current values Figure 6b, on the other hand, shows how the effect of tap changes on the absolute difference between subsequent electrode current values

can be reduced.

From the above it is clear that this invention is particularly useful for identifying eiectrode lengths. It clearly indicates breakages or tip losses. Possible reasons for furnace unbalances can be identified and proper corrective actions be taken. Very long electrodes are also prone to operate at very low eiectrode current values, which historically caused a iot of confusion regarding length. This invention can then be used to determine whether the low electrode currents are due to long electrodes or not.

Accurate control of the relative length of individual electrodes and ultimately significant improvement in the accuracy of locus of electrode arc points gives direct rise to:

● Improvement in the radial and axial distribution of thermal energy throughout the cylindrical or rectilinear geometry of the high temperature reactor volume;

● Maximization of the available viable volume within the high temperature furnace reactor in which to conduct the high temperature reaction, offering improved metallurgical recovery efficiency and throughput benefits;

● An improved and less varying thermal profile in a baking zone of electrodes, which makes the electrodes more resilient to breakages.

The invention provides a significant reduction in the macroscopic movement of electrode elements, with a corresponding decrease in levels of induced mechanical and thermal stress. This results in significantly decreased mechanical and chemical wear rates (e.g. an aggregated reduction of over 15%).

Improved overall metallurgical process stability is provided as a result of the above stated advantages, thereby yielding superior recovery of target element species to their designated respective phases and a corresponding reduction in undesirable (or tramp) element species distribution to the target reaction phases of value.