WO2012176211A1

WO2012176211A1 - A computer implemented interactive system for facilitating aluminium smelting analysis and optimization

Info

Publication number: WO2012176211A1
Application number: PCT/IN2011/000834
Authority: WO
Inventors: Jogesh MANKAR
Original assignee: Aditya Birla Science and Technology Co Ltd; Hindalco Industries Ltd
Current assignee: Aditya Birla Science and Technology Co Ltd; Hindalco Industries Ltd
Priority date: 2011-06-24
Filing date: 2011-12-07
Publication date: 2012-12-27
Anticipated expiration: 2013-12-24

Abstract

A computer implemented system and a method for facilitating aluminium smelting analysis arid optimization have been disclosed. The system analyses the various parameters of operating smelting cells in real-time and arrives at control actions to be taken to increase the performance of the cells. In addition, the system enables engineers/personnel to establish in advance various operating conditions for newly designed cells to do a test run before initiating the process in real-time. Also, the system ^provides a platform for operative training to engineers/personnel and also facilitates them in understanding the technical details of the aluminium smelting process and the impact of changes done to various parameters on the performance of the smelting cell.

Description

A COMPUTER IMPLEMENTED INTERACTIVE SYSTEM FOR FACILITATING ALUMINIUM SMELTING ANALYSIS AND OPTIMIZATION

FIELD OF THE INVENTION

The present invention relates to optimization and analysis of aluminum smelting process based on conditions and situations in a physical aluminum smelting cell.

DEFINITION OF TERMS USED IN THE SPECIFICATION

The term 'cascaded graphical user interface' in this specification relates to multimedia based graphical interfaces which provide multiple choices to users and open another graphical interface as a result of selecting a choice on the latter.

The term 'cascaded model' in this specification relates to models which are interdependent on an output from one or more predetermined models or common/shared inputs for their processing.

The term 'graphical user interface elements' in this specification relates to widgets that are used on a graphical user interface to perform predetermined operations including navigational operation, command input operation and data input/output operation. Typically, graphical user interface elements can be at least one of a menu, toolbar, icons, command buttons, text boxes, scrollbar and the like.

The term 'mathematical modeling' in this specification relates to the use of mathematical functions to derive optimum parameter values and analyze relationships between parameters and their effects on the system.

The term 'model' in this specification relates to an aluminium smelting based operational parameter which may be interdependent on other aluminium smelting based operational parameters. The model includes a set of input parameters and operations which determine the value of the model and in turn affect other interdependent models. For instance, the Superheat model, the value of Superheat is derived by performing predetermined operation on input parameters like bath temperature and liquidus temperature and the derived value in turn affects the value of the 'Ledge thickness model'. The term 'routine operation based model' in this specification relates to the model which captures and facilitates in processing of operational parameters like anode changing, alumina feeding and dissolution, metal tapping, crust breaking, bath transfusion, anode beam raising, anode effect termination, anode cover maintenance and the like for conducting daily aluminium smelting operations.

The term Overlapping values' in this specification relates to input parameter values and computed output values that are shared between two or more cascaded models for their processing. For instance, the output value computed for Superheat in the Superheat model forms a basis for computation of ledge thickness in the Ledge thickness model.

The term 'screen related operations' in this specification relates to the various operations that are applied to a content area for facilitating functions like clearing the screen, resetting the content on the screen and existing / quitting from the current content area.

These definitions are in addition to those expressed in the art.

BACKGROUND OF THE INVENTION

Aluminium, being the third most abundantly found element, is widely used in a range of household items, in construction and automobile and aircraft manufacture. In its natural form aluminium is found in bauxite and requires a process of smelting for separating the aluminium from its ore. Hall-Heroult smelting process is typically used for separating aluminium metal and oxygen gas in aluminium smelting cells or reduction cells.

However, many challenges are faced during the process of smelting / electrolysis. These challenges include:

• difficulties in operating the smelting cells continuously near the optimal operating point;

• the reduction cells being difficult to instrument;

• imprecise measurement of many inputs to the cell; and

• Finally, the control strategy used not being robust enough to optimize the cell's performance.

The reasons for these challenges are that firstly, the high temperatures make many typical instrumentation devices unusable, as the bath temperature exceeds 950°C and even the sidewall temperature can exceed 300°C. Secondly, strong magnetic fields in the smelting process cause significant distortion in electrical signals. Thirdly, the cell operation is noisy, the bubbling of the anode gas is chaotic and affects the overall cell voltage and waves in the metal pad also cause fluctuations in the cell voltage that are difficult to predict exactly. Fourthly, the area of greatest interest is the region between the anode face and the metal pad, but the electrolyte in this region is extremely corrosive and there are a limited options of materials that can be affordably manufactured that can withstand the bath for an extended period of time. Because of these difficulties, important quantities such as alumina concentration, bath temperature, anode cathode distance (the distance from the bottom of an anode to the top of a metal pad), metal pad depth and bath depth are not measured on a continuous basis. The only two variables monitored on a continuous basis are the cell voltage or the total voltage drop across the cell, and the line current or the total current passing through the cell.

These challenges can be overcome by having a better understanding of the effects of operational parameters on the cell, as well as by formulating improved control strategies to operate smelting cells closer to their optimal level, thus saving significant amounts of energy and money. Because of the immense amount of energy used in the aluminium reduction process, small operational improvements can produce significant savings. For example, a reduction in the cell voltage by 0.1 V on a single cell operating at 100,000 Amps can save about 90,000 kW/hours each year. At $0.05 per kilowatt-hour that will be equal to a saving of $4,500 per cell. In an established, highly competitive industry such as the aluminium industry, this is a significant amount of saving.

To optimize the smelting cell's operations and to maintain the smelting cells at optimum operating values the conventional systems used in the aluminium smelting process employ trial-and-error techniques. The trial-and-error techniques are highly complex and require engineers/personnel to manually operate the smelting cells at optimum temperature, test new cells by devising control strategies and learn the process parameter interdependencies on the performance of the cell. However, these techniques are time consuming, expensive and not suitable as they disturb the real-time operations and performance of smelting cells. Further, decisions made using a trial-and-error technique in a conventional aluminium smelting process are based on inadequate information and imprecise measurement data hence is not useful in optimizing the smelting process.

There is therefore felt a need for a system which can: • eliminate the trial-and-error testing and learning techniques;

• increase current efficiency, smelting cell life and reduce energy consumption;

• operate the smelting reduction cell at optimal set point values in real time;

• pre-test and validate newly designed cells and their control strategies without disturbing the real-time smelting process;

• provide hands-on training to the engineers/personnel without disturbing the real-time smelting process, provide guidelines for smelting process optimization and assist them in analysing the correlation between the various parameters and their impact on the smelting process as well as their optimum operating values; and

• advice the engineers/personnel to make better operational decisions.

OBJECTS OF THE INVENTION

It is an object of the present invention to provide a system which eliminates the trial-and- error testing and learning techniques.

It is another object of the present invention to provide a system which increases the current efficiency, smelting cell life and reduces energy consumption.

It is still another object of the present invention to provide a system which facilitates operation of smelting cells at optimal set point values in real time.

It is yet another object of the present invention to provide a system which facilitates pretesting and validation of newly designed smelting cells and their parameter values without disturbing the real-time smelting process.

One more object of the present invention is to provide a system which provides a platform for providing hands-on training to the engineers/personnel without disturbing the real-time smelting process.

Further, another object of the present invention is to provide guidelines for smelting process optimization and assisting the engineers/personnel in analysing the correlation between the various parameters and their impact on the smelting process.

Yet another object of the present invention is to provide advice to the engineers/personnel to make better operational decisions. SUMMARY OF THE INVENTION

The present invention envisages a computer implemented interactive system for facilitating aluminium smelting analysis and optimization, the system comprising:

• a computational unit adapted to receive a set of input values for a selected model and further adapted to operate in a plurality of modes wherein,

i. in an advisory mode the computational unit adapted to process the selected model based on predetermined process model data and the set of input values to compute at least one output value selected from the group consisting of predictions, control actions, optimum operating values, graphical analysis and operational recommendations; and ii. in a training mode the computational unit adapted to compute at least one output value selected from the group consisting of forecasting calculations to predict overall smelting process performance, technical information data and trend analysis based graphical output values to comprehend effects of the set of input values on the smelting process, based on predetermined process data and the set of input values; and

• a cascaded and intuitive graphical user interface comprising graphical user interface elements arranged in a predetermined format adapted to enable display of at least one section corresponding to a mode of operation, wherein the section is adapted to be navigated using the graphical user interface elements to display at least one model associated with the section for selection, the graphical user interface further adapted to accept and forward a set of input values corresponding^* to a model selection to the computational unit to receive and display the output values.

Typically, the system comprises a repository co-operating with the computational unit, wherein the repository adapted to store comprehensive process model data encompassing an aluminium smelting process including mathematical models linked to predetermined models, conservation laws, empirical correlation data pertaining to the aluminium smelting process and mass and energy balance data taken from a plurality of units in the smelting cells and from a plurality of components in the bath.

Preferably, the set of input values are selected from the group consisting of input values pertaining to input parameters associated with a model and commands representing operations to be performed on the input parameters, wherein the input parameters are selected from the group consisting of design parameters, bath composition and chemistry based parameters, physical and chemical properties based parameters and operating parameters, and the operations are selected from the group consisting of calculation operations, graph plotting operations, analysis operations, help operations, screen related operations, report generation operations, online data exchange with smelting cell operations, data retrieval operations, data display operations and data transfer operations.

"further, the models are selected from the group consisting of energy based models, mass balance based models, empirical correlation based models, cost based models, real time based models, routine operation based models and environment based models.

Still further, the computational unit comprises:

• processing means adapted to process the selected model based on a predetermined process model data and the input parameters values to compute at least one output value; and

• an analyser unit adapted to perform an iterative operation of modifying the input parameter values to obtain optimum output values which tend towards predetermined optimum operating values, the analyser unit having:

i. a comparator adapted to receive and compare the output values with predetermined optimum operating values and further adapted to provide a compared value; and

ii. second processing means adapted to modify the input parameter values based on the compared value and process the selected model based on a predetermined process model data and the modified input parameters values to compute an output value and further adapted to provide the output value to the comparator and still further adapted to iterate the operation till the modified input parameter values generate optimum output values.

Furthermore, the computational unit comprises fetching and storage means adapted to fetch and store predetermined overlapping values in a temporary storage for automatic processing of cascaded models. Additionally, the system comprises a process interface co-operating with the computational unit for linking the system and units of a macro aluminium smelter.

In accordance with this invention, there is provided a computer implemented method for facilitating aluminium smelting analysis and optimization, the method comprising the following steps:

• receiving a set of input values for a selected model corresponding to a mode of operation for an aluminium smelting process;

• processing the selected model based on the set of input values and corresponding data in the repository and computing at least one output value;

• comparing the output value with a predetermined optimum operating value and providing a compared value;

• iteratively modifying the input values based on the compared value to obtain an optimum output value which tends to the predetermined optimum operating value; and

• recommending a modified "set of input values corresponding to the optimum output value for aluminium smelting optimization and operating the smelting cell at optimum conditions to achieve the operational stability, optimum process conditions, maximum productivity, minimum energy consumption and minimum anode effects.

In accordance with this invention, the step of receiving a set of input values includes the steps of receiving input parameters values including design parameters, bath composition and chemistry based parameters, physical properties and chemical properties based parameters associated with a model, and commands representing operations to be performed on the input parameters values.

Typically, the step of processing the selected model includes the step of cascading predetermined overlapped values for simultaneous processing of cascaded models.

Preferably, the step of processing the selected model and computing at least one output value includes the steps of computing trends and output variable values for a model and displaying the trends and the output variable values for facilitating aluminium smelting analysis. BRIEF DESCRIPTION OF THE ACCOMPANYING DRAWINGS

Other aspects of the invention will become apparent by consideration of the accompanying drawings and their description stated below, which is merely illustrative of a preferred embodiment of the invention and does not limit in any way the nature and scope of the invention:

FIGURE 1 illustrates a schematic of the system for facilitating aluminium smelting analysis and optimization in accordance with the present invention;

FIGURE 2 is a screenshot of a sample graphical user interface showing the models of operational section in accordance with the present invention;

FIGURE 3 is a screenshot of a sample user interface showing the models of technical section in accordance with the present invention;

FIGURE 4 is a screenshot of a sample user interface showing a set of inputs, controls and outputs in the content area for 'Superheat' model selection in accordance with the present invention; -

FIGURE 5 is an exemplary graph generated on the graphical user interface showing analysis of Liquidus Temperature with other parameters in accordance with the present invention; and

FIGURE 6 is flowchart showing the steps involved in facilitating aluminium smelting analysis and optimization in accordance with the present invention.

DETAILED DESCRIPTION

The drawings and the description thereto are merely illustrative of a computer implemented interactive system to facilitate aluminium smelting analysis and optimization and only exemplify the invention and in no way limit the scope thereof.

The conventional trial-and-error techniques which are used for functioning the smelting cells closer to the optimum operating values, devising new control strategies for newly designed cells and enabling engineers/personnel to, analyse the correlation between changes of parameter values and their impact on the smelting cell performance are very expensive and time consuming. These techniques not only disturb the real-time operation of a cell but also do not enable the engineers/personnel to closely understand the correlations between parameters and the cell performance to formulate efficient strategies and take good operational decisions. The efficiency in decision making is further reduced as most parameters in real-time cannot be measured accurately.

Therefore, to overcome these shortcomings and to operate the cell at a high efficiency, minimize the number and duration of anode effects, maximize the life of the cell and minimize the amount of carbon consumption and energy consumption, the present invention envisages a computer implemented interactive system for aluminium smelting analysis and optimization. In accordance with this invention, the envisaged system provides a user friendly and intuitive interface which enables user/plant personnel to use the system with minimal training or assistance.

The proposed system uses mathematical modelling as the basis for its analysis and optimization. Mathematical modelling enables the present invention to determine the optimal values including the optimal set point for the cell, the best metal pad depth, the optimal alumina concentration, or the operating bath temperature. Mathematical modelling also facilitates the present invention in analysing significant changes in the cell operation, for example, the effect of changes in the size of the anode. Further, the mathematical modelling facilitates the present invention to develop and test new control algorithms without the threat of disturbing an operating smelting cell.

Thus, the present invention based on efficient mathematical modelling techniques performs aluminium smelting optimization by maintaining the cell at optimum process conditions so as to increase the current efficiency, increase cell life and reduce the energy consumption. Also, the proposed invention performs advanced process control, determines real time predictions and real time optimization for optimizing the aluminium smelting process without disturbing a real-time operating cell. For performing the optimization, the proposed system only accepts a limited set of input parameters like cell voltage and line current to accurately forecast the other parameter values to determine optimum values for a plurality of models for increasing the cell's performance and gaining economic benefits.

A well-trained, enlightened operating and technical team can maintain cells at best efficiency level on sustained basis, reduce anode effects, reduce downtime of cells, stabilize cells which in turn lowers emissions and improve specific consumptions of aluminium fluoride, soda and anode carbon. Therefore to provide well-trained operating and technical teams, the proposed system also provides engineers/personnel with an interactive graphical user interface and an efficient computational unit which emulates an operating environment and provides them hands-on training and assists them in performing aluminium smelting analysis including performing process parameters simulations, control strategy validations, training and prediction of cell behaviour on variations in cell process parameters and predictions of process performance. These analyses equip the engineers/personnel in making better operational decisions, in gaining better understanding of relationships between cell process parameters and Understanding interactions existing between a cells various parameters. The engineer/plant personnel can apply various values to input parameters of a cell to visualize how varying the process parameters will affect the cell operation. Then choose the best values to operate the cell so as to optimize the most important technical and economic results. The system also provides guidelines to engineers/personnel to take the control actions in the cell to maintain the cell at optimum process conditions. The system behaves as an "advisor" or "trainer" and is built around a mathematical model of the cell and conservation laws governing the process. The system includes various sections and their respective models which are dedicated to various aspects of the aluminium smelting process /reduction process to perform mathematical model based calculations for analysis and optimization without disturbing an operating cell.

The system can work in a standalone mode where it can recommend real time parameter values for optimizing the cell operations or facilitate operational control and training. Alternatively, the system can be connected to units of a macro smelter and can control the operation of the smelter by automatically adjusting the parameter values to operate the cell at optimum values.

Referring to the accompanying drawings, FIGURE 1 illustrates a schematic of a computer implemented system 100 which facilitates in conducting aluminium smelting analysis and optimization. The system f6 includes a repository 102 to store comprehensive process model data encompassing an aluminium smelting process. This process model data includes mathematical models linked to predetermined models, empirical correlation data pertaining to the aluminium smelting process and mass and energy balance data taken from a plurality of units in the smelting cells and also from a plurality of components in the bath. The repository 102 also includes conservation laws governing the aluminium smelting process and additional information on the process and the process parameters. Further, the system 100 includes a cascaded and intuitive graphical user interface 104. The intuitive graphical .user interface enables users to operate the system with ease and with very little training or assistance. The graphical user interface 104 comprises graphical user interface elements which are arranged in a predetermined format to enable display of at least one section corresponding to a mode of operation. The graphical user interface elements provides various controls which help to display information, navigate the displayed information, make selections and enable searching for a particular information.

The sections include namely an operational section and a technical section. These sections include a plurality of models which are selected from the group consisting of heat based models, mass balance based models, empirical correlation based models, cost based models, real time based models, routine operation based models and environment based models. For instance, the mass balance based models include AIF3 addition model and Na₂Co₃ addition model; Heat based models includes Superheat Model; and empirical correlation based models include Bath Ratio control model, ledge thickness model and the like. Also, provided is a help section to help users operate the system and provide information about the envisaged system.

These sections are adapted to be navigated using the graphical user interface elements via navigation means 106 to display at least one model associated with the section for selection. For instance, the navigation means 106 may include a drop down menu for displaying the list of models which are associated with a particular section or a scroll bar to scroll the list of models or a set of command/control buttons.

Referring to FIGURE 2 of the accompanying drawings, which is an exemplary screenshot of the graphical user interface 104 shows the graphical user interface elements of type menu and command/control buttons represented by reference numeral 200.

The operational section includes the following models which are categorized in the following groups:

• a basic parameters model group comprising:

o a smelting cell superheat model to predict liquid temperature, bath temperature, superheat and heat transfer;

o a current efficiency monitoring model to compute theoretical and actual aluminium production; and o a ledge thickness model to compute a desirable ledge thickness.

• a compound addition model group comprising:

o an aluminium fluoride addition model to decrease bath ratio to target bath ratio; arid

o a sodium carbonate addition model to increase bath ratio to target bath ratio.

• a smelting cell control model group comprising:

o a bath ratio control model to control bath ratio with reference to target bath ratio;

o an aluminium fluoride control model to control addition of aluminium fluoride to bath; and

o a base voltage control model to predict smelting cell voltage using a voltage model.

FIGURE 2 shows an expanded operational section and its various models.

The technical section is subdivided into models selected from the group consisting of a process engineering model, an aluminium fluoride control model, and a smelting cell information model. And the process engineering model is further subdivided as follows:

• a bath chemistry model;

• a theoretical and actual consumptions and productions model;

• a voltage model;

• an energy consumption and balance model;

• a fluoride evolution model;

• a cell performance model; and

• a current efficiency loss model.

FIGURE 3 shows an expanded technical section and its various models.

The operational section covers all aspects of the aluminium smelting process required in the aluminium smelting optimization by emulating and predicting process parameter values of a cell/pot in real time. The technical section on the other hand assists the users in conducting analysis on the various parameters of the cell and their impact on the cell's performance. The technical section enables users to get training, advice and guidelines to increase their proficiency in taking operational decisions and understanding the correlations between the smelting process parameters and choice of optimum operating values. Further, the graphical user interface elements include a content area and control buttons which display multimedia content pertaining to a selection of a model and enable performing of operations on the selected model respectively. FIGURE 4 of the accompanying drawings shows the content area which is generally represented by reference numeral 400 and the controls for manipulating the content generally represented by reference numeral 402 on an exemplary snapshot of the graphical user interface 104.

The navigation means 106 of the graphical user interface 104 enables users to navigate the sections, selection means 108 to perform a selection of at least one model and a set of input values and communication means 110 to accept a set of input values corresponding to the selection and forward the same to a computational unit 112. The selection means 108 can sense the selection of a model and a set of input values via a keypad based selection, click based selection or a touch based selection.

Typically, the set of input values are selected from the group consisting of input values provided by a user pertaining to input parameters associated with a model and commands representing operations selected by the user to be performed on the input parameters. The operations are selected from the group consisting of calculation operations, graph plotting Operations, analysis operations, help operations, screen related operations, report generation operations, online data exchange with smelting cell operations, data retrieval operations, data display operations and data transfer operations. Thus, for instance, for 'Ledge Thickness' model selection, based on the mathematical model associated with 'Ledge thickness' the set of input values provided by the user may include values for input parameters like bath ratio, CaF₂, MgF₂, A1₂0₃, room temperature, ledge-bath interface heat transfer co-efficient, shell long side and air heat transfer co-efficient and 'calculation operation' selection to predict the superheat value and generate optimum value for Ledge thickness.

In accordance with this invention, the input parameter values form the basis for the process calculation and are process variables whose values can be directly obtained from the aluminium smelting plant in real time. Alternatively, the input parameter values can be provided manually to the system by a user. The input parameters values are categorized as design parameters values, bath composition and chemistry based parameters values, physical and chemical properties based values and operating parameters values. Typically, the design parameters values are the fixed values and the bath composition parameter values are the discrete data which is obtained from the lab measurements once in typically, thirty two hours. Similarly, bath temperature is also a discrete parameter value which is measured once in typically, twenty four hours.

Still further, the system 100 includes a computational unit 112 which co-operates with the - graphical user interface 104 to receive a set of input values for a selected model via the communication means 110 and based on the selection operates in a plurality of modes.

In an advisory mode, the computational unit 112 processes the selected model based on corresponding process model data in the repository 102 and the set of input values to compute at least one output value selected from the group consisting of predictions, control actions, optimum operating values and graphical analysis.

In a training mode, the computational unit 112 computes at least one output value selected from the group consisting of forecasting calculations to predict overall smelting process performance, technical information data and trend analysis based graphical output values to comprehend effects of the set of input values on the smelting process, based on the process model data from the repository and the set of input values.

The computational unit 112 includes fetching and storage means (not shown in the figures) which fetches overlapping values including the input parameter values and the output values from interdependent models and stores these overlapping values in a temporary storage for automatic processing of cascaded models. This is particularly useful in scenarios where output value of one model serves as input to another model or a predetermined set of input parameter values are shared between models. This facilitates simultaneous processing of cascaded model and saves the extra computation required for re-processing/re-populating the input parameter values and reduces computation time. For instance, the bath ratio control model, A1F₃ addition model and Na₂Co₃ addition model are interdependent. For maintaining the bath ratio lab value at a target value, the value of bath ratio (which is the overlapping value) is fetched and stored for providing to the A1F₃ addition model and Na₂Co₃ addition model on their activation/selection to generate the amount of A1F₃ or Na₂C0₃ to be added to maintain/control the bath ratio at the target value.

The computational unit 112 further includes processing means 114 to process the selected model based on the predetermined process model data from the repository 102 and the input parameters values to compute at least one output value. This output values is given an analyser unit 116 which performs an iterative operation of modifying the input parameter values to obtain optimum output values which tend towards predetermined optimum operating values. The analyser unit 116 generates recommendations/feedback for the plant personnel to derive appropriate input parameter values to get optimized outputs. The analyser unit 116 also facilitates the plant personnel to perform analysis on a set of output values by supplying a set of inputs to the system.

The analyser unit 116 includes a comparator 118 which receives the output value from the processing means 114 and compares it with predetermined optimum operating values to provide a compared value. The 'compared value' indicates to the system whether the output value is optimized. The analyser unit 116 also includes second processing means 120 which modifies the input parameter values based on the compared value and processes the selected model based on a predetermined process model data from the repository 102 and the modified input parameters values to compute an output value. This output value is given to the comparator 118 to determine if output value generated tends towards the predetermined optimum operating values. The operation of the analyser unit 116 is iterated till the modified input parameter values generate optimum output values. The analyser unit 116 then provides the modified input parameter values as the recommended operating values for getting optimized outputs from the smelting cells. These modified input parameter values are the input values which need to be maintained in the smelting cell to operate the smelting cell/ smelting pot at optimum conditions to achieve the operational stability, optimum process conditions, maximum productivity, minimize energy consumption, minimize anode effects and minimize emissions.

The output values generated by the analyser unit 116 of the computational unit 112 are displayed on the content area of the graphical user interface 104 via the communication means 110 for enabling users to conduct aluminium smelting analysis and optimization. For instance, if a selection includes a 'superheat' model selection and a set of input values include 'calculate' operation selection along with bath temperature and base voltage values as the set of input parameters, then the computational unit 112 fetches the mathematical model corresponding to the 'superheat' model from the repository 102 and based on the mathematical model processes the received set of input values to calculate optimum value for bath temperature via the analyser unit 116. This optimum value of bath temperature can be used by the engineers/personnel to adj st the bath temperature of the operating cell in real time to optimize the process.

FIGURE 5 of the accompanying drawings shows an exemplary graph generated as 'graphical analysis' type of output value which is displayed on the content area of the graphical user interface 104 for the Superheat model selection by the computational unit 112 on selection of the 'graph plotting operation' by a user. The graph shows a variation of Liquidus Temperture (TL) with other parameters associated with the Superheat model, wherein the variation can be plotted for different input parameters by selecting at least one control button represented by reference numeral 500. FIGURE 5 shows the TL with R at various A1₂0₃ concentrations to enable a user to arrive at the optimum values of TL, R and A1₂0₃, on selection of the 'TL Vs R at different A1₂0₃' control button.

Furthermore, the system 100 includes a process interface 122 which is communicably coupled to the computational unit 112 for linking the system 100 and units of a macro aluminium smelter. Using the process interface 122 the system can automatically receive a set of measured inputs and predict values of other input parameters to generate output values to operate the smelter at optimum values in real-time.

The various sections, their corresponding models and functions will be explained in detail hereinafter.

The operational section is seen in FIGURE 2 with its various models including superheat, A1F3 addition, Na2C03 addition, Bath Ratio control, Base voltage control, current efficiency and ledge thickness. These models can be navigated using at least one of a menu type of graphical use interface element or command buttons /control buttons type of graphical interface element or the like. This section is designed for the operational personnel facilitating them in taking daily control actions for example for bath ratio control and the like. This section and its models are structured to easily incorporate new knowledge and/or experimental results about specific relations or process parameters, which become available during the processing for enhancing the cell model authenticity level/ accuracy.

In accordance with this invention, these models are linked to predetermined mathematical models in the repository 102 to perform the desired calculations. The functions of the models of the operational section are explained hereinafter. OPERATIONAL SECTION Superheat model:

The liquidus temperature is the temperature at which freezing of bath starts. It sets the lowest temperature at which a Hall-Heroult cell can be operated without a precipitate forming. The bath temperature needs to be kept sufficiently above the liquidus temperature i.e. bath must have sufficient superheat to provide heat to the solution for alumina additions. A low liquidus temperature permits a low operating temperature. This is desirable because, all being equal, lowering the operating temperature by 1°C improves Current Efficiency (CE) by about 0.18%. The bath temperature is the temperature of the liquid electrolyte measured during pot normal operation. The difference between the bath temperature and liquidus temperature is the superheat. *t

This model is used for predicting superheat, hence heat transfer and thereby ledge profile or ledge thickness. Layout of the graphical user interface 104 for superheat is shown in FIGURE 4. The computational unit 112 on selection of this model extracts the corresponding mathematical model from the repository 102 and receives/predicts a set of inputs parameters like the weight % of Aluminium fluoride (A1F₃), Aluminium Oxide (A1₂0₃), Calcium Fluoride (CaF₂), Magnesium Fluoride (MgF₂), Lithium Fluoride (LiF), Lithium Hexafluoroaluminate (Li₃AlF₆), KF, Bath Ratio, Base Voltage and measured Bath temperature and uses these to provide as output value predicted values for the liquidus temperature and the bath temperature. And processes the value of superheat based on the bath temperature predicted value and the liquidus temperature predicted value. The processing is initiated after receiving the inputs and selection of the control button 'calculate'. This model via the computational unit 112 also predicts superheat based on 'bath temperature measured' and 'liquidus temperature predicted'. The exemplary graph can be seen in FIGURE 5. The graph provides users with the variation of Liquidus Temperature (TL) with Bath Ratio (R), Aluminium fluoride (AIF3), Aluminium Oxide (A1₂0₃), Calcium Fluoride (CaF₂) and Magnesium Fluoride (MgF₂). This model also provides users with the guidelines to arrive at optimum operating region. For instance, from the graph of TL with R at various A1₂0₃ concentrations as seen in FIGURE 5, users can arrive at the optimum value of TL, R and A1₂0₃. A1F3 addition Model:

Bath ratio (NaF/AlF₃) is a typical and fundamental cell parameter to achieve the best performance of electrolytic cells. Bath ratio has a direct impact in cell temperature control, alumina solubility, ledge formation and current efficiency.

A1₂0₃ contains Na₂O=0.3wt% and CaO=0.02wt%. A1F₃ gets consumed by the following two reactions as per the predetermined mathematical model for this model:

3Na₂0 + 4A1F₃ = 2Na₃AlF₆ + A1₂0₃ (1)

3CaO + 2AlF₃ = 3CaF₂ + Al₂0₃ -— -(2) '

A1F₃ in the bath reacts with Na₂0 and CaO present in A1₂0₃ to produce Na₃AlF₆ and CaF₂ according to the above reactions. Hence, the A1F₃ in the bath gets reduced and thereby the bath ratio gets increased. Also, additional A1F₃ is required to neutralize the Na₃AlF₆ produced. Thus, to decrease the bath ratio (Lab Value) to the target bath ratio, A1F₃ is to be added and to increase the bath ratio (Lab value) to the target bath ratio Na₂C0₃ is to be added to the bath.

This model via the computational unit 112 provides the users with the amount of A1F₃ to be added to the bath in twenty four hours to maintain the target bath ratio i.e. to control the bath ratio. Also it provides the user with the A1F₃ to be added in three days and the like number of days based on the practise followed at a smelting plant.

This model is of utmost importance to the operational personnel to control the bath ratio. Na₂C0₃ addition Model:

This model via the computational unit 112 provides users with the amount of Na₂C0₃ to be added to the bath in twenty four hours to maintain the target bath ratio i.e. to control the bath ratio.

Na₂C0₃ added to the bath reacts with A1F₃ present in the bath to produce NaF according to the following reaction and thereby increasing the bath ratio to bring it closer to the target value.

3Na₂C0₃ + 2A1F₃ = 6NaF + A1₂0₃ + 3C0₂ — (3)

Bath Ratio Control Model: This model via the computational unit 112 is used for the batf ratio control. With the bath ratio available from the lab, users can easily find out whether it is above or below the target bath ratio. If bath ratio lab value is more than the target value then users can visit the A1F₃ addition model and find out the amount of A1F₃ to be added to maintain/control the bath ratio to the target value.

On the other hand, if the bath ratio lab value is less than the target value, users can visit the Na₂C0₃ addition model and Find out the amount of Na₂C0₃ to be added to maintain/control the bath ratio to the target value. This model is of utmost importance to the operational personnel to control the bath ratio.

This model predicts xsAlF₃, utilizing bath ratio (R) and bath composition (CaF₂, MgF₂, A1₂0₃) as input variables and for a particular value R, predicts XSA1F₃. To adjust the R (reduce R), users have to feed the new value of R as input parameter for which new value of XSA1F₃ can be predicted. The difference in new value of XSA1F₃ and previous XSA1F₃ gives the amount of XSA1F₃ to be added to achieve the desired value of R.

For example,

To reduce R as per the predetermined mathematical model for this model:

For the Lab value of R=1.2, the computational unit 112 predicts Excess A1F3 Wt% =8.22 and Excess A1F3 in Kg = 140.17. For the Target value of R=1.121, the computational unit 112 predicts Excess A1F3 Wt% =10.77 and Excess A1F3 in Kg = 183.67. Therefore difference in Excess A1F3 in Kg = 183.67 - 140.17 = 43.5 Kg.

A1F₃ Neutralization reactions (1) and (2) are mentioned above in A1F3 addition models. Na₃AlF₆ produced by the reaction (1) is 6.46 (Kg/24h). The amount of A1F₃ required for neutralizing this Na₃AlF₆ = 0.87 (Kg 24h).

A1F₃ neutralized by both these reactions is 8.03 (Kg/24h) The Excess A1F3 addition = Excess A1F₃ in Kg from the bath ratio + A1F₃ neutralized by both these reactions. Hence, Excess A1F3 addition = 52.42 (Kg/24h) and Excess A1F3 addition = 70.22 (Kg/3 days).

To increase R to the target bath ratio, Na2C03 is to be added. If the lab bath ratio is 1.0 and the target bath ratio is 1.121, then 41.0418 Kg of Na2C03 is to be added.

Base voltage control model: This model via the computational unit 112 provides users with the prediction of cell voltage using voltage model. The cell voltage is a sum of bath voltage, voltage drop in anode, voltage drop in cathode, external voltage drop (voltage drop in busswork) and internal voltage drop. The bath voltage is a sum of ohmic bath voltage (Vohrn) and back EMF (Ebemf). The ohmic bath voltage is a sum of bubble voltage drop and resistive voltage drop across the anode- cathode distance. Back EMF (Ebemf) is a sum of decomposition potential and surface overvoltages [anode surface overvoltage, anode concentration overvoltage (due to concentration gradients at the anode surface) and cathode concentration overvoltage]. Internal Voltage drop is the sum of Diamond drop, Clamp drop, Rod stub drop, Stub carbon drop, Copper rod drop. External Voltage drop is the sum of cathode ring bus drop and anode ring bus drop. Voltage model calculates all these voltages which contribute to the pot voltage.

This model provides a 'plot graphs' control button option which on the content area plots the values which facilitate users to perform the following analysis:

• Cell voltage with alumina concentrations at different ACD values;

• Energy consumption with alumina concentrations at different ACD values;

• Energy consumption with alumina concentrations at different current intensity values;

• Cell voltage with alumina concentrations at different bath ratio [R] values;

• Cell voltage with alumina concentrations at different bath ratio [R] and corresponding bath temperature values;

• Cell voltage with alumina concentrations at different current density values;

• Cell voltage with alumina concentrations at different bath temperature values;

• Cell voltage with alumina concentrations at different superheat values;

• Pseudoresistance with alumina concentrations at different ACD values; and

• Pseudoresistance with alumina concentrations at different current density values.

For instance, as per the predetermined mathematical model for this model, for a set of input parameters such as amperage=66000 amps, anode area = 2829 cm2, number of anodes = 26, anode current density = 0.8973 A/cm2, anode cathode distance (ACD) = 0.048m, internal busbar drop= 0.15597 volts, external busbar drop= 0.125 volts, voltage drop in cathode= 0.332 volts, voltage drop in anode= 0.142, bath ratio=1.121, CaF2=6.1, MgF2=0.47, A1203=3; the various drops contributing towards the total cell voltage are decomposition potential or equilibrium potential -1.222, reaction limited current density= 0.005 A/cm2, concentration limited current density= 3.189 A/cm2, concentration overvoltage at anode= 0.017 volts, concentration overvoltage at cathode = 0.041 volts, bubble layer thickness under anode = 0.492 cm, bubble overvoltage= 0.303 volts, back emf = 1.810 volts, polarized cell potential= 2.113 volts, voltage drop in ACD = 1.483 volts, ohmic voltage drop in bath = 1.786 volts, bath voltage = 3.596 volts, cell voltage = 4.351 volts.

The objective of increase in aluminium production is achieved by increasing the amperage. In that case, take for instance the amperage is to be increased to 76000 amps with maintaining the same cell voltage =4.351 volts, the probable combinations of parameters are anode area = 3500 cm2, number of anodes = 44, anode current density = 0.493506 A/cm2, keeping the bath composition same.

This model also enables the users to plot a graph showing the variation of cell voltage with the A1₂0₃ concentration at different ACD to enable users to derive the optimum A1₂0₃ concentration and ACD to maintain cell voltage at the optimum and thereby facilitates optimum energy consumption.

Current Efficiency model:

Plant personnel have to feed the actual aluminium production for one cell per day. This model via the computational unit 112 provides these personnel with the actual aluminium production for n cells n days, theoretical aluminium production for one cell per day, theoretical aluminium production for n cells n days and the current efficiency of the cell in percentage.

Ledge Thickness model:

Superheat controls the formation of side-ledge arid bottom freeze (ridge). The carbon sides of the cell must be protected by a layer of frozen bath to prevent erosion. To maintain a frozen ledge, sufficient heat must be extracted through the wall to drop the temperature (across the boundary layer at the bath-ledge interface) from that of the adjacent liquid to the liquidus temperature.

One of the reasons for the cell failure is the diminishing of the side-ledge. This is one of the most important models. This model via the computational unit 112 provides users with the desired ledge thickness produced on the side-wall. For the desired ledge thickness, it provides users with the set of optimum bath composition, temperatures and heat transfer coefficients. For instance, for a set of input variables such as bath ratio = 1.121, CaF2 = 6.1, MgF2=0.47, A1203 =2.8, room temperature = 70 degC, ledge-bath interface heat transfer coefficient = 1.2 (KW/m2K), shell long side and air heat transfer coefficient = 0.024 (KW/m2K) superheat predicted is 10.7599 degC and the ledge thickness is 7.70 cm.

Ledge thickness can be increased by reducing the superheat which is adjusted by changing the bath composition (bath ratio, CaF2, MgF2, A1203) or by reducing the room temperature or by increasing the shell long side and air heat transfer coefficient.

With the reduction in bath composition such as CaF2, MgF2, A1203; superheat decreases and thereby ledge thickness increases. With Ihe decrease in bath ratio superheat increases and thereby ledge thickness decreases. Hence this model is of utmost importance to maintain the desired side-ledge thickness by arriving at the optimum set of parameters described in this example.

TECHNICAL SECTION

This section is designed for the technical aficionados, who want to understand the smelting process in detail. Layout of the user interface 104 with the navigation pertaining to technical section is shown in FIGURE 3. As can be seen from this section's snapshot in FIGURE 3 this section comprises process models, advisory control, and additional information.

The process model section comprises models such as bath chemistry, theoretical and actual productions and consumptions, voltage, energy consumption or energy balance, fluoride evolution, cell performance and current efficiency loss. The Bath chemistry model comprises various bath properties such as liquidus temperature, excess A1F₃, saturation concentration of alumina in the bath, electrical conductivity of bath, electrical resistivity of bath, bath density and viscosity, aluminium density and viscosity, aluminium bath density difference, vapour pressure of bath, surface tension of bath, bath composition based on actual weight of bath and the 100 kg of bath. Each of these models and their functions are explained as follows:

Bath Chemistry model:

This model via the computational unit 112 provides users with the various bath properties. The bath (cryolite+additives) influences many key parameters and processes in the Hall- Heroult electrolysis. Liquidus temperature model:

This model via the computational unit 112 provides users with the liquidus temperature using various correlations. Also it provides users with bath temperature. Superheat is predicted based on the predicted liquidus temperature and the measured bath temperature. The user can select the different correlations from the drop down menu of this model.

This model provides users with the 'plot graphs' control buttons to analyse the variations of liquidus temperature with chemical composition (various bath components such as A1F3, xsAlF3, CaF2, MgF2, A1203, Bath ratio (R) and the like). A click on the 'plot graphs' button enables the computational unit 112 to present the graphical analysis to users on the content area of the user interface 104. Typically, the graph plots TL with R at various alumina concentrations and provides users with the operating region to maintain of TL, R and A1203 at optimum values.

Maximum alumina solubility or saturation concentration of alumina in the bath model:

This model via the computational unit 112 is used for point feeder setting, formation and dissolution of sludge. Above the saturation concentration, alumina does not dissolve and sludge is formed. This model via the 'plot graphs' control button enables users to analyse the variation of saturation concentration of alumina with chemical composition (various bath components such as A1F3, XSA1F3, CaF2, MgF2, A1203, Bath ratio (R) and the like) and the bath temperature. The 'plot graph' contains multiple control button having author names whose theory can be used to plot the graph. A click on the buttons denoted by authors' names opens the graph on the content area and facilitates users to perform analysis. The graph of Cs with R at various Tsuperheat values provides users with the operating region to maintain Cs, R and Tsuperheat at optimum values.

Electrical conductivity of bath model:

The electrical conductivity of the cryolite electrolyte is an important factor in cell voltage, cell voltage settings, energy consumption and therefore important factor in power/energy efficiency. The unit of the electrical conductivity is Siemens per meter (S/m or S/cm). Siemens (reciprocal Ω or eventually mho) is the unit of electrical conductance. This model via the computational unit 112 provides users with the electrical conductivity of bath (k) using various correlations. The 'plot graph' contains multiple control button having author names whose theory can be used to plot the graph. A click on the buttons denoted by authors' names opens the graph on the content area and facilitates users to analyse the variation of electrical conductivity of bath (k) with chemical composition (various bath components such as A1F3, XSA1F3, CaF2, MgF2, A1203, Bath ratio (R) and the like) and the bath temperature. The graph also enables analysis of k with R at various A1₂0₃ values and provides users with the operating region to maintain of k, R and A1₂0₃ at optimum values.

Electrical Resistivity of bath model:

This model via the computational unit 112 is used for cell voltage settings. This model provides users with the electrical resistivity of bath using various correlations.

The 'plot graph' contains multiple control button having author names whose theory can be used to plot the graph. A click on the buttons denoted by authors' names opens the graph on the content area and facilitates users to analyse the variation of electrical resistivity of bath (p) with chemical composition (various bath components such as A1F3, XSA1F3, CaF2, MgF2, A1203, Bath ratio (R) and the like) and the bath temperature.

The graph also enables analysis of electrical resistivity with R at various Al₂0₃ concentrations, using which users can arrive at the optimum value of electrical resistivity, R and A1₂0₃ concentrations.

Aluminium bath density difference model:

Density of bath and aluminium are essential, for instance for magneto-hydrodynamic behaviour of the aluminium pad under the bath layer. The density difference must be large enough -to assure good separation of the aluminium pad from the bath layer. This density difference must be larger than 0.2 g/cm3 in order to prevent mixing and to maintain good separation between the metal pad and the electrolyte layer. This model via the computational unit 112 provides the user with the aluminium bath density difference.

Bath and aluminium viscosity model:

The viscosity values of the liquid aluminium phase and the molten bath affect the movement of metal and electrolyte, the dissolution and sedimentation of alumina particles and the release of gas bubbles from the surface have electromagnetic effects. This model via the computational unit 112 provides the user with the Bath viscosity. A click on the 'plot graphs' control button provides users with the graph of bath viscosity with R at various A1₂0₃ values on the content area of the user interface 104 and provides users with the operating region to maintain bath viscosity, R and A1₂0₃ at optimum values. The graph also provides bath viscosity with R at various Tsuperheat values to users to maintain the operating region for bath viscosity, R and Tsuperheat at optimum values.

Vapour Pressure model:

The vapour phase of the liquid cryolite contains several species, NaF and NaAlF₄. The vapour pressure is responsible for the fluorine evolution of the cell. This model via the computational unit 112 provides users with the vapour pressure of bath.

A click on the 'plot graphs' control button gives users the variation of vapour pressure of bath with chemical composition (various bath components such as A1F3, XSA1F3, CaF2, MgF2, A1203, Bath ratio (R) and the like) and the bath temperature. This is the graph of vapour pressure of bath with R at various A1203 concentrations and provides users with the operating region to maintain vapour -pressure of bath, R and A120₃ concentrations. Also the graph of vapour pressure of bath with R at various Tsuperheat values provides users with the operating region to maintain vapour pressure of bath, R and Tsuperheat values.

Surface Tension model:

The surface tension is responsible for wetting of the carbon anodes and the anode effect. This model via the computational unit 112 provides users with the surface tension of bath. On clicking 'plot graphs' control button provided on the user interface 104, users get the variation of surface tension of bath with chemical composition (various bath components such as A1F3, xsAlF3, CaF2, MgF2, A1203, Bath ratio (R) and the like) and the bath temperature.

The graph of surface tension of bath with R at various A1₂0₃ concentrations provides users with the operating region to maintain surface tension of bath with R at various A1₂0₃ concentrations at optimal values. The graph of surface tension of bath with R at various superheat values also provides users with the operating region to maintain surface tension of bath with R at various superheat values at optimal values.

Theoretical and- Actual Productions and consumptions model:

This model interface via the computational unit 112 provides users with the following values to increase the theoretical and actual productions and consumptions amperage: • Theoretical aluminium, C02, A1203 consumption, carbon consumption.

• Actual aluminium, C02, CO, A1203 consumption, carbon consumption.

• Aluminium production for the specified number of pots.

• Specific productions and consumptions i.e. for 1 kg of aluminium.

• Specific production of C02, CO.

• Specific consumption of A1203 and carbon.

• A1F3 consumption and Cryolite production.

• Specific A1F3 consumption and specific Cryolite production.

• Electrolytic volume production of C02 and CO.

• Mass and volumetric production of 02 in case of inert anodes.

For instance, as per the predetermined mathematical model for this model, for a set of input variables amperage = 65 KAmps, fractional current efficiency = 0.95, time of production = 24 hours, number of cells =1, the outputs are theoretical aluminium production = 523.54 Kg, theoretical C02 production = 640.38 Kg, theoretical A1203 consumption = 989.20 Kg, theoretical carbon consumption = 174.72 Kg, actual aluminium production = 497.36 Kg, actual C0₂ production = 576.34 Kg, actual A1203 consumption = 939.74 Kg, actual carbon consumption = 174.72 Kg, specific (for 1 Kg of Al) C02 production = 1.16 Kg, specific CO production = 0.08 Kg, specific A1203 consumption = 1.92 Kg, specific carbon consumption = 0.35 Kg, actual cryolite production by neutralization reactions (1) and (2) = 6.60 Kg, theoretical cryolite production = 6.95 Kg, specific cryolite production = 0.01 Kg, actual A1F3 consumption = 5.28 Kg, theoretical A1F3 consumption = 5.56 Kg, specific A1F3 consumption = 0.01 Kg. »

Voltage model:

Voltage model is already explained in the section of bath voltage control model (operational section). A click on the 'plot graphs' control button provided on this user interface opens a content area which facilitates users to conduct various analyses.

Energy Consumption or Energy Balance model:

This model interface via the computational unit 112 provides users with the enthalpy/energy (KJ/mol) for the following:

• To heat components (C, A1203).

• For the forward and the backward reaction. • To heat alumina.

• To convert gamma alumina to alpha alumina.

• Formation of CO, C02, A1203.

• The reaction and the total enthalpy to produce aluminium.

For instance, as per the predetermined mathematical model for this model, for a set of input variables such as heat of formation of Al = 0 (KJ/mol), heat of formation of C02 = 393.522 (KJ/mol), heat of formation of CO = 110.527 (KJ/mol), heat of formation of A1203 = 1675.69 (KJ/mol), heat of formation of carbon = 0 (KJ/mol); the outputs are enthalpy to heat carbon = 14.44 (KJ/mol), enthalpy to heat A1203 = 116.18 (KJ/mol), enthalpy to heat A1203 to ambient temperature = 8.07 (KJ/mol), enthalpy to heat A1203 to bath temperature = 105.98 (KJ/mol), enthalpy to heat A1203 from ambient temperature to bath temperature = 97.91 (KJ/mol), enthalpy of reaction = 549.51 (KJ/mol), total enthalpy i.e. energy consumption to produce Al = 619.01 (KJ/mol).

Energy efficiency model:

The model interface via the computational unit 112 provides users with the following:

• Energy equivalent of voltage theoretically required.

• Energy equivalent of various voltages actually required (KWh KgAl).

• Power equivalent of various voltages actually required (KW).

• Theoretical and actual energy requirements (KWh/KgAl).

• Energy efficiency of the process from the actual energy required and the theoretical amount of energy required by the process.

For instance, as per the predetermined mathematical model for this model, for a set of input variables such as cell voltage = 4.351, amperage . = 66000 amps, voltage drop in cathode= 0.332 volts, voltage drop in anode= 0.142, the outputs obtained are external voltage drop= 0.47 volts, internal voltage part entering into the cell= 3.88 volts, voltage required to heat alumina = -0.40 volts, voltage required to heat anodes = -0.05volts, voltage required for the formation of C02 = 1.37 volts, voltage required for the formation of CO= 0.39 volts, voltage required for reaction to take place = -1.93 volts, voltage consumption= voltage required for reaction + voltage required to heat alumina + voltage required to heat anodes = -2.38 volts, voltage equivalent of heat loss or heat production = 1.50 volts, theoretical energy required = 6.42 (KWh/KgAl), actual energy required = 13.65 (KWh KgAl), hence the energy efficiency - (theoretical energy required) / (actual energy required) = 47.00 %. The objective of energy efficiency to be increased to 49.88% can be achieved by reducing the cell voltage to 4.1 volts.

Fluoride Evolution model:

Fluoride evolution happens by means of particulate and the gaseous. This model interface via the computational unit 112 provides users with the amount of fluoride evolved by the volatilization of bath, entrainment of bath and hydrolysis of bath (gaseous fluoride is produced by hydrolysis of bath and hydrolysis of pot vapor).

Under the 'plot graphs' control buttons there are sub controls provided named as volatilized bath, entrained bath, hydrolysis bath and hydrolysis pot fume. A click on the "volatilized bath" button provides users with a content area showing a graph with volatilized bath analysis. Similarly, a click on the button named as "FVP Vs A1203 at different R" opens the graph in the content area showing the variation of amount of volatilized bath (FVP) in (KgF/ton Al) with A1203% at different R. Similar trends pertaining to entrained bath, hydrolysis of bath, hydrolysis of pot fume caii be obtained from the specified control buttons on the user interface of this model.

Cell Performance model:

This model interface via the computational unit 112 provides users with the cell performance in terms of actual energy consumption in the process. The user interface of this model provides users with many control buttons. A click on the button named as "EC Vs A1203 at different ACD" opens the graph showing the variation of energy consumption with A1203% at different ACD values. Similarly trends pertaining to energy consumption can be obtained from the different buttons provided on the user interface of this model.

Current Efficiency Loss model:

This model interface via the computational unit 112 provides users with the saturation concentration of aluminium in the bath and the current efficiency. The loss in CE is due to the back reaction of aluminium with C02 in the bath. This model is based on this CE loss. The CE is mainly dependent on the saturation concentration of aluminium in the bath (C*A1) which depends on R, Tbath, CaF2 and A1203 concentrations.

The user interface of this model provides users with many control buttons. A click on the button named as "CE Vs CS at Al" opens the graph showing the variation of current efficiency with saturation concentration of aluminium in the bath. Similarly, trends pertaining to current efficiency can be obtained from the different specified buttons provided.

A1F3 control model:

The model interface via the computational unit 112 provides users with the amount of A1F3 to be added to the bath to compensate the losses from the fluoride emission and the A1F3 neutralization.

Material Properties model: -

This model interface via the computational unit 112 gives the material properties of all the materials (raw materials, lining materials, products) used and produced in aluminium smelter.

Heat transfer coefficients model:

This model interface via the computational unit 112 provides users with the convective heat transfer coefficients at various interfaces in the smelter.

Cell Geometry model:

This model interface via the computational unit 112 provides users with the anode volume, anode mass, bath mass in ACD, bath mass above ACD, total bath mass, metal volume and metal mass before tapping, metal volume and metal mass after tapping.

Bath Composition model:

Basis is 1705 Kg of bath model: The actual weight of bath is typically near around 1705 kg. The wt% composition of bath is 100 wt%. This model interface via the computational unit 112 provides users with the actual weight in kgs, mole fraction, mass fraction, molar concentration (kmol/m3) and mass concentration(kg/m3) for all the components in the bath. Basis is 100 Kg of bath model: For a given bath composition in wt% on 100kg basis, this model interface via the computational unit 112 provides users with the molefraction, mass fraction, molar concentration (kmol/m3) and mass concentration(kg/m3) for all the components in the bath.

In accordance with this invention, there is provided a computer implemented method for facilitating aluminium smelting analysis and optimization, the method comprising the following steps as seen in FIGURE 6:

• receiving a set of input values for a selected model corresponding to a mode of . operation for an aluminium smelting process, 1000; processing the selected model based on the set of input values arid corresponding data in the repository and computing at least one output value, 1002;

comparing the output value with a predetermined optimum operating value and providing a compared value, 1004;

iteratively modifying the input values based on the compared value to obtain an optimum output value which tends to the predetermined optimum operating value, 1006; and

recommending a modified set of input values corresponding to the optimum output value for aluminium smelting optimization and operating the smelting cell at optimum conditions to achieve the operational stability, optimum process conditions, maximum productivity, minimum- energy consumption and minimum anode effects, 1008.

Preferably, the step of processing the selected model and computing at least one output value includes the steps of computing trends and output variable values for a model and displaying the trends and the output variable values for facilitating aluminium smelting analysis.

TECHNICAL ADVANTAGES

The technical advantages of the present invention include providing a computer implemented system and method for conducting aluminium smelting analysis and optimization.

In accordance with one aspect, the proposed system is an advisory system which receives real-time operating conditions of smelting cells and generates control actions like advisory information, recommendations and forecasts pertaining to the operating region of the various desired operating process parameters to optimize the performance of the cells including increasing current efficiency,, increasing cell life, reducing energy consumption and the like. Also, system facilitates in prior testing of process control operations and in establishing operating conditions for newly designed cells. The system provides advisory information for taking control action of various operating units of the smelting process for example for Bath ratio control, Bath composition and Ledge thickness.

In another aspect, the proposed ¾ystem is a training system which provides the engineers/ personnel with a simulated environment to operate and control the system, study and analyse the effects of various parameters on the smelting process. The training provided by the system helps the engineer/ personnel to get all the necessary information pertaining to a particular process/sub process in smelting. The system also enables the engineers/personnel to conduct trend analysis to determine the optimum operative regions for various parameters. Moreover, the system is very comprehensive and encompasses the entire aspects of aluminium smelting process. Further, the system is user friendly and requires very less training or assistance for its operation. The system is based on mathematical modelling and hence provides accurate predictions/ recommendations for analysis and optimizations without disturbing an operating cell.

Still further, in its various modes of operation the system can be interfaced to a physical macro smelter for generating control actions for controlling the smelting process and operating it at optimal values. Alternatively, system can be a standalone unit which accepts offline inputs of the operating measurements of a smelting cell and generates control actions giving the optimal values which need to set for achieving the optimal operating point. .

Thus, the proposed system encompasses the various aspects of the smelting process and facilitates in optimizing the technical and economic results for smelting cells and performs process optimization, process parameters simulations, control algorithms validations, advanced process control, training and prediction of smelting cell behaviour on variations in smelting cell process parameters, predicting process performance, real time predictions and real time optimizations. Furthermore, the system assists in making better operational decisions, facilitating better understanding of relationships between smelting process parameters and analysing interactions existing between various smelting process parameters.

While considerable emphasis has been placed herein on the components and component parts of the preferred embodiments, it will be appreciated that many embodiments can be made and that many changes can be, made in the preferred embodiments without departing from the principles of the invention. These and other changes in the preferred embodiment as well as other embodiments of the invention will be apparent to those skilled in the art from the disclosure herein, whereby it is to be distinctly understood that the foregoing descriptive matter is to be interpreted merely as illustrative of the invention and not as a limitation.

Claims

CLAIMS:

1. A computer implemented interactive system for facilitating aluminium smelting analysis and optimization, said system comprising:

i. in an advisory mode said computational unit adapted to process said selected model based on predetermined process model data and said set of input values to compute at least one output value selected from the group consisting of predictions, control actions, optimum operating values, graphical analysis and operational recommendations; and ii. in a training mode said computational unit adapted to compute at least one output value selected from the group consisting of forecasting calculations to predict overall smelting process performance, technical information data and trend analysis based graphical output values to comprehend effects of said set of input values on the smelting process, based on predetermined process model data and said set of input values; and

• a cascaded and intuitive graphical user interface comprising graphical user interface elements arranged in a predetermined format adapted to enable display of at least one section corresponding to a mode of operation, wherein said section is adapted to be navigated using said graphical user interface elements to display at least one model associated with said section for selection, said graphical user interface further adapted to accept and forward a set of input values corresponding to a model selection to said computational unit to receive and display said output values.

2. The system as claimed in claim 1, wherein said system comprises a repository cooperating with said computational unit, wherein said repository adapted to store comprehensive process model data encompassing an aluminium smelting process including mathematical models linked to predetermined models, conservation laws, empirical correlation data pertaining to the aluminium smelting process and mass and energy balance data taken from a plurality of units in the smelting cells and from a plurality of components in the bath.

3. The system as claimed in claim 1, wherein said set of input values are selected from the group consisting of input values pertaining to input parameters associated with a model and commands representing operations to be performed on said input parameters, wherein said input parameters are selected from the group consisting of design parameters, bath composition and chemistry based parameters, physical and chemical properties based parameters and operating parameters, and said operations are selected from the group consisting of calculation operations, graph plotting operations, analysis operations, help operations, screen related operations, report generation operations, online data exchange with smelting cell operations, data retrieval operations, data display operations and data transfer operations.

4. The system as claimed in claim 1, wherein said models are selected from the group consisting of energy based models, mass balance based models, empirical correlation based models, cost based models, real time based models, routine operation based models and environment based models.

5. The system as claimed in claim 1, wherein said computational unit comprises:

• processing means adapted to process said selected model based on a predetermined process model data and said input parameters values to compute at least one output value; and

• an analyser unit adapted to perform an iterative operation of modifying said input parameter values to obtain optimum output values which tend towards predetermined;Optimum operating values, said analyser unit having:

i. a comparator adapted to receive and compare said output value with predetermined optimum operating values and further adapted to provide a compared value; and

ii. second processing means adapted to modify said input parameter values based on said compared value and process said selected model based on a predetermined process model data and the modified input parameters values to compute an output value and further adapted to provide said output value to said comparator and still further adapted to iterate the operation till said modified input parameter values generate optimum output values.

6. The system as claimed in claim 1, wherein said computational unit further comprises fetching and storage means adapted to fetch and store predetermined overlapping values in a temporary storage for automatic processing of cascaded models.

7. The system as claimed in claim 1, wherein^' said system comprises a process interface cooperating with said computational unit for linking said system and units of a macro aluminium smelter.

8. A computer implemented method for facilitating aluminium smelting analysis and optimization, said method comprising the following steps:

• processing said selected model based on said set of input values and corresponding data in said repository and computing at least one output value;

• comparing said output value with a predetermined optimum operating value and providing a compared value;

• iteratively modifying said input values based on said compared value to obtain an optimum output value which tends to said predetermined optimum operating value; and

• recommending a modified set of input values corresponding to said optimum output value for aluminium smelting optimization and operating the smelting cell at optimum conditions to achieve the operational stability, optimum process conditions, maximum productivity, minimum energy consumption and minimum anode effects.

9. The method as claimed in claim 8, wherein the step of receiving a set of input values includes the steps of receiving input parameters values including design parameters, bath composition and chemistry based, parameters, physical properties and chemical properties based parameters associated with a model, and commands representing operations to be performed on said input parameters values.

10. The method as claimed in claim 8, wherein the step of processing said selected model includes the step of cascading predetermined overlapped values for simultaneous processing of cascaded models.

1. The method as claimed in claim 8, wherein the step of processing said selected model and computing at least one output value includes the steps of computing trends and output variable values for a model and displaying said trends and said output variable values for facilitating aluminium smelting analysis.