BACKGROUND OF THE INVENTION

THIS invention relates to a method of operating a grinding mill and to a grinding mill system.

SAG mills are usually installed as a first processing step for mineral ore after it has been stockpiled at a mine. The SAG mill includes a rotating grinding chamber 18 and a plurality of liners/mill liners 20 which are located inside the grinding chamber 18. These liners 20 have a generally tooth-like formation/profile and are configured to lift a portion of the material 22 located inside the grinding chamber 18 (as well as grinding balls which may also be located inside the chamber), as the grinding chamber 18 rotates. The lifted material 22 (and grinding balls) would then fall onto the material 22 located at the bottom of the grinding chamber 18, thereby grinding the material 22 into smaller particle sizes. Reference is in this regard made to Figures 1 A-D.

Historically, to set the charge cell limits of a SAG mill (i.e. the amount of material (e.g. ore) contained inside a grinding chamber of the SAG mill), operators would typically focus on a charge cell value that maximizes throughput. However, this charge cell value is usually unknown, difficult to determine and changes over time. As a result, a lot of academic research has been dedicated in order to try and determine this charge cell value (i.e. for maximizing throughput). Reference is in this regard made to the following publication:

Van der Westhuizen, Andre P and Powell, Malcolm S., Milling curves as a tool for characterizing SAG mill performance, SAG 2006, Department of Mining Engineering, University of British Columbia, I- 217 to 1-231.

Through various research, it has been determined that a curve of a SAG mill throughput (e.g. in tonnage per hour (TPH)) versus a charge cell value (e.g. percentage filling of material in the chamber - abbreviated as “CC” in the drawings) has an inverse parabolic shape as shown in Figure 2A, as well as in the above-mentioned publication (see specifically Figure 8 of the publication). As mentioned, a lot of research has gone into identifying the apex of this curve (i.e. the charge cell value which results in maximum throughput). However, this curve is not static and changes all the time, which makes it very difficult to identity the apex. More specifically, the curve can move more towards the left or right as shown in Figure 2B (see curves 102 and 104). As a result, although the specific charge cell value which results in maximum throughput might have been accurate originally (see reference numeral 106 on curve 102), it may end up later producing a throughput which is unsatisfactorily low (see reference numeral 108 on curve 104).

Reference is made to the following prior art documents, namely BR102020004854 A2, WO81/01373 A1 , W02002/000072 A1 ,

W02021/045720 A1 and US2013/0008985. However, none of these prior art documents address the issues mentioned above.

The Inventors wish to address at least some of the problems mentioned above.

SUMMARY OF THE INVENTION

In accordance with a first aspect of the invention there is provided a method of operating or monitoring a grinding mill, wherein the method includes: i. obtaining/receiving a first variable which relates to a first property of a current grinding mill operation of the grinding mill; ii. obtaining/receiving a second variable which relates to a second property of the current grinding mill operation of the grinding mill; iii. utilising the first variable and the second variable in order to determine/calculate a maximum charge cell amount for the grinding mill which would result in a stable throughput/output where the throughput/output of the grinding mill remains at a pre- determined/desired target throughput/output level or higher than the pre-determined/desired target throughput/output level.

The step (iii) may include utilising the first variable and the second variable, together with historical information on what a measured throughput/output of a particular historical grinding mill operation was at different times over a period of time, and for each measured throughput/output what its associated first variable and second variable were, in order to determine/calculate the maximum charge cell amount for the grinding mill which would result in a stable throughput/output where the throughput/output of the grinding mill remains at the pre-determined/desired target throughput/output level or higher than the pre-determined/desired target throughput/output level .

The grinding mill may be a SAG (semi-autogenous) grinding mill.

The grinding mill may be an AG (autogenous) grinding mill.

The method may include utilising the first variable and the second variable, together with the historical information, in order to determine/calculate a minimum charge cell amount for the grinding mill.

The method may include managing/controlling an amount of material contained in a grinding chamber of the grinding mill during operation so that it does not exceed the maximum charge cell amount. The managing/controlling step may include managing or controlling an inflow of material into the grinding chamber.

The method may include obtaining/receiving a third variable which relates to a third property of the current grinding mill operation of the grinding mill. Step (iii) may include utilising the first variable, the second variable and the third variable, together with the historical information, in order to determine/calculate the maximum charge cell amount, wherein the historical information includes information on what a measured throughput/output of a particular historical grinding mill operation was at different times over a period of time, and for each measured throughput/output what its associated first variable, second variable and third variable were.

The first variable may be any one selected from: a material property of material located inside a grinding chamber of the grinding mill or a material which is to be loaded into the grinding chamber, or a physical or operational property of the grinding mill.

The second variable may be any one selected from: a material property of material located inside the grinding chamber of the grinding mill or a material which is to be loaded into the grinding chamber, or a physical or operational property of the grinding mill.

The first variable may relate to a state of liners which are located inside the grinding chamber. The second variable may relate to a hardness of the material located inside the grinding chamber or material which is to be loaded into the grinding chamber. The method may also include obtaining/receiving a third variable which relates to a particle size/granulometry of material located inside a grinding chamber or material which is to be loaded into the grinding chamber. Step (iii) may then include utilising the first variable, the second variable and the third variable, together with the historical information, in order to determine/calculate the maximum charge cell amount, wherein the historical information includes information on what a measured throughput/output of a particular historical grinding mill operation was at different times over a period of time, and for each measured throughput/output what its associated first variable, second variable and third variable were. The third variable may change/vary over a period of time. The step of obtaining/receiving the third variable may include receiving/obtaining a new/updated third variable on a continual basis over the period of time. Step (iii) may include obtaining/calculating a new maximum charge cell amount when a new/updated third variable has been obtained/received.

The first variable and the second variable may both change/vary over a period of time. The step of obtaining/receiving the first variable may include receiving/obtaining a new/updated first variable on a continual basis. The step of obtaining/receiving the second variable may include receiving/obtaining a new/updated second variable on a continual basis. The method may include obtaining/calculating a new maximum charge cell amount by utilising the new/updated first and second variables together with the historical information.

A rate at which the first variable changes over time may be lower than a rate at which the third variable changes over time. A rate at which the second variable changes over time may be lower than the rate at which the third variable changes over time.

Step (iii) may include utilising a calculation module which is configured to use the first, second and third variables as inputs in order to determine/calculate the maximum charge cell amount. The method may include configuring the calculation module by utilising the historical information. The method may include automatically managing/controlling the amount of material contained in the grinding mill during operation so that it does not exceed the maximum charge cell amount, by using a control module.

The material may be ore.

The first variable may relate to a state of liners which are located inside the grinding chamber and the second variable may relate to a hardness of the material located inside the grinding chamber or material which is to be loaded into the grinding chamber, and wherein the method may also include obtaining/receiving a third variable which relates to a particle size/granulometry of material located inside a grinding chamber or material which is to be loaded into the grinding chamber, and wherein step (iii) may include utilising the first variable, the second variable and the third variable, together with the historical information, in order to determine/calculate the maximum charge cell amount, wherein the historical information includes information on what a measured throughput/output of a particular historical grinding mill operation was at different times over a period of time, and for each measured throughput/output what its associated first variable, second variable and third variable were, and wherein the third variable may change faster in time than both the first variable and the second variable, whereby the method may further includes generating a graphical illustration of the maximum charge cell amount versus a range of third variable values, by assuming that the first and second variables remain unchanged/constant for a certain period of time, and utilising the graphical illustration in order to set cell charge limits for the grinding mill.

In accordance with a second aspect of the invention there is provided a grinding mill system for performing a grinding mill operation, wherein the system includes: a grinding chamber; and a calculation module which is configured to i. obtain/receive a first variable which relates to a first property of the grinding mill operation, ii. obtain/receive a second variable which relates to a second property of the grinding mill operation, and iii. utilise the first variable and the second variable, in order to determine/calculate a maximum charge cell amount for the grinding chamber which would result in a stable throughput/output where the throughput/output of the grinding mill remains at a pre- determined/desired target throughput/output level or higher than the pre-determined/desired target throughput/output level. A “module”, in the context of the specification, includes an identifiable portion of code, computational or executable instructions, or a computational object to achieve a particular function, operation, processing, or procedure. A module may be implemented in software, hardware or a combination of software and hardware. Furthermore, modules need not necessarily be consolidated into one device.

The calculation module may be configured to utilise the first variable and the second variable, together with historical information on what a measured throughput/output of a particular historical grinding mill operation was at different times over a period of time, and for each measured throughput/output what its associated first variable and second variable were, in order to determine/calculate the maximum charge cell amount for the grinding chamber which would result in a stable throughput/output where the throughput/output of the grinding mill remains at a pre- determined/desired target throughput/output level or higher than the pre- determined/desired target throughput/output level.

The grinding mill system may be a SAG (semi-autogenous) grinding mill system.

The grinding mill system may be an AG (autogenous) grinding mill system.

The system may include a control module which is configured to manage/control the amount of material contained in the grinding chamber during operation, so that it does not exceed the maximum charge cell amount.

The calculation module may be configured to utilise the first variable and the second variable, together with the historical information, in order to determine/calculate a minimum charge cell amount for the grinding mill.

The calculation module may be configured to: obtain/receive a third variable which relates to a third property of the grinding mill operation, and utilise the first variable, the second variable and the third variable, together with the historical information, in order to determine/calculate the maximum charge cell amount, wherein the historical information includes information on what a measured throughput/output of a particular historical grinding mill operation was at different times over a period of time, and for each measured throughput/output what its associated first variable, second variable and third variable were.

The first variable may be any one selected from: a material property of material located inside a grinding chamber of the grinding mill or a material which is to be loaded into the grinding chamber, or a physical or operational property of the grinding mill.

The second variable may be any one selected from: a material property of material located inside the grinding chamber of the grinding mill or a material which is to be loaded into the grinding chamber, or a physical or operational property of the grinding mill.

The first variable may relate to a state of liners which are located inside the grinding chamber. The second variable may relate to a hardness of the material located inside the grinding chamber or material which is to be loaded into the grinding chamber. The calculation module may be configured to: obtain/receive a third variable which relates to a particle size/granulometry of material located inside a grinding chamber or material which is to be loaded into the grinding chamber, and utilise the first variable, the second variable and the third variable, together with the historical information, in order to determine/calculate the maximum charge cell amount, wherein the historical information includes information on what a measured throughput/output of a particular historical grinding mill operation was at different times over a period of time, and for each measured throughput/output what its associated first variable, second variable and third variable were.

The third variable may change/vary over a period of time and the calculation module may be configured to: obtain/receive a new/updated third variable on a continual basis over the period of time, and obtain/calculate a new maximum charge cell amount when a new/updated third variable has been obtained/received.

The first variable and the second variable may both change/vary over a period of time. The calculation module may be configured to: obtain/receive a new/updated first variable on a continual basis; obtain/receive a new/updated second variable on a continual basis; and obtain/calculate a new maximum charge cell amount by utilising the new/updated first, second and third variables.

The system may include a control module which is configured to manage/control the amount of material contained in the grinding chamber during operation, so that it does not exceed the maximum charge cell amount.

The system may include one or more sensors for measuring the first and/or second variables.

The first variable may relate to a state of liners which are located inside the grinding chamber and the second variable may relate to a hardness of the material located inside the grinding chamber or material which is to be loaded into the grinding chamber, and wherein the calculation module is configured to: obtain/receive a third variable which relates to a particle size/granulometry of material located inside a grinding chamber or material which is to be loaded into the grinding chamber, wherein the third variable changes faster in time than both the first variable and the second variable, utilise the first variable, the second variable and the third variable, together with historical information on what a measured throughput/output of a particular historical grinding mill operation was at different times over a period of time, and for each measured throughput/output what its associated first variable, second variable and third variable were, in order to determine/calculate the maximum charge cell amount, and generate and display a graphical illustration of the maximum charge cell amount versus a given particle size/granulometry of material on a display screen, by assuming that the first and second variables remain unchanged/constant for a certain period of time.

In accordance with a third aspect of the invention there is provided a method of operating or monitoring a grinding mill, wherein the method includes: i. obtaining/receiving a first variable which relates to a state of liners which are located inside a grinding chamber of the grinding mill and wherein the first variable remains constant/unchanged for a certain period of time; ii. obtaining/receiving a second variable which relates to a hardness of the material located inside the grinding chamber or material which is to be loaded into the grinding chamber and wherein the second variable remains constant/unchanged for said certain period of time; iii. obtaining/receiving a third variable which relates to a particle size/granulometry of material located inside a grinding chamber or material which is to be loaded into the grinding chamber and wherein the third variable changes over said certain period of time; iv. determining/calculating a maximum charge cell amount for the grinding mill for each of a range of third variable values which would result in a stable throughput/output where the throughput/output of the grinding mill remains at a pre-determined/desired target throughput/output level or higher than the pre-determined/desired target throughput/output level, by taking into account the first variable and the second variable; v. determining/calculating the maximum charge cell amount for the grinding mill at various points in time within the said period of time, by taking into account the specific third variable value at each point in time.

The method may include controlling the charge cell amount inside the grinding chamber over the said period of time by ensuring that it does not exceed the maximum charge cell amount as determined/calculated for each point in time.

Step (iv) may include determining/calculating the maximum charge cell amount for the grinding mill for each of said range of third variable values which would result in a stable throughput/output where the throughput/output of the grinding mill remains at a pre-determined/desired target throughput/output level or higher than the pre-determined/desired target throughput/output levelby taking into account the first variable, the second variable and historical information on what a measured throughput/output of a particular historical grinding mill operation was at different times over a period of time, and for each measured throughput/output what its associated first variable, second variable and third variable were.

The method may include displaying the determined/calculated maximum charge cell amount for each of said range of third variable values graphically on a display screen.

In accordance with a fourth aspect of the invention there is provided a method of operating a grinding mill, wherein the method includes: i. obtaining/receiving a first variable/parameter which relates to a material property of material located inside the grinding chamber or material which is to be loaded into the grinding chamber, or a physical or operational property of the mill; ii. obtaining/receiving a second variable/parameter which relates to a material property of material located inside the grinding chamber or material which is to be loaded into the grinding chamber, or a physical or operational property of the mill; and iii. utilising the first variable/parameter and the second variable/parameter in order to determine/calculate an indication (hereinafter referred to as the “quantity indication”) as to a quantity/amount of material which should be contained in the grinding chamber during operation, in order to reduce the likelihood of a throughput/output of the mill dropping below a pre-determined minimum/desired throughput/output level.

The grinding mill may be a SAG (semi-autogenous) grinding mill. Alternatively, the grinding mill may be an AG (autogenous) grinding mill.

The method may include managing/controlling the quantity/amount of material contained in the grinding chamber during operation, based on the quantity indication obtained in step (iii) above.

The first variable/parameter may relate to a state of liners which are located inside the grinding chamber. The second variable/parameter may relate to a hardness of the material located inside the grinding chamber or material which is to be loaded into the grinding chamber. The method may also include obtaining/receiving a third variable/parameter which relates to a particle size/granulometry of material located inside a grinding chamber or material which is to be loaded into the grinding chamber. Step (iii) may include utilising the first variable/parameter, the second variable/parameter and the third variable/parameter in order to determine the quantity indication. The material may be ore (e.g. mineral/mining ore).

The method may include determining a maximum material quantity/amount which should be allowed to be contained inside the grinding chamber during operation, by utilising the first variable/parameter, the second variable/parameter and the third variable/parameter. The managing/controlling step may include managing/controlling the quantity of material inside the grinding chamber during operation such that the quantity does not exceed the allowed maximum material quantity/amount.

The managing/controlling step may include managing or controlling an inflow of material into the grinding chamber.

The third variable/parameter may change/vary over a period of time. The step of obtaining/receiving the third variable/parameter may include receiving/obtaining a new/updated third variable/parameter on a continual basis over the period of time. Step (iii) may therefore include obtaining/calculating a new quantity indication when a new/updated third variable/parameter has been obtained/received.

The first variable/parameter and the second variable/parameter both change/vary over a period of time. The step of obtaining/receiving the first variable/parameter may include receiving/obtaining a new/updated first variable/parameter on a continual basis. The step of obtaining/receiving the second variable/parameter may include receiving/obtaining a new/updated second variable/parameter on a continual basis. The method may include obtaining/calculating a new quantity indication by utilising the new/updated first and second variables/parameters in step (iii).

A rate at which the first variable/parameter changes over time may be lower than a rate at which the third variable/parameter changes over time. A rate at which the second variable/parameter changes over time may be lower than the rate at which the third variable/parameter changes over time. Step (iii) may include utilising a calculation module which is configured to use the first, second and third variables/parameters as inputs in order to estimate the maximum material quantity/amount which should be allowed to be contained inside the grinding chamber during operation.

The method may include configuring the calculation module by utilising historical information on what a measured throughput/output of a particular mill was at different times over a period of time, and for each measured throughput/output, what

(a) a state of liners was which were located inside a grinding chamber of the mill at the time,

(b) a hardness of the material located inside a grinding chamber at the time or material which was in the process of being fed into the grinding chamber at the time, and

(c) a particle size/granulometry of material located inside the grinding chamber at the time or material which was in the process of being fed into the grinding chamber at the time.

The method may include automatically managing/controlling the quantity/amount of material contained in the mill during operation, based on the quantity indication obtained in step (iii), by using a control module.

In accordance with a fifth aspect of the invention there is provided a grinding mill system which includes: a rotatable grinding chamber; and a calculation module which is configured to i. obtain/receive a first variable/parameter which relates to a material property of material located inside the grinding chamber or material which is to be loaded into the grinding chamber, or a physical or operational property of the mill, ii. obtain/receive a second variable/parameter which relates to a material property of material located inside the grinding chamber or material which is to be loaded into the grinding chamber, or a physical or operational property of the mill, and iii. utilise the first variable/parameter and the second variable/parameter in order to determine/calculate an indication (hereinafter referred to as the “quantity indication”) as to a quantity/amount of material which should be contained in the grinding chamber during operation, in order to reduce the likelihood of a throughput/output of the mill dropping below a pre- determined/desired minimum desired throughput/output level.

The grinding mill system may be a SAG (semi-autogenous) grinding mill system. Alternatively, the grinding mill system may be an AG (autogenous) grinding mill system.

The system may include a control module which is configured to manage/control the quantity/amount of material contained in the grinding chamber during operation, based on the quantity indication obtained by the calculation module.

The first variable/parameter may relate to a state of liners which are located inside the grinding chamber. The second variable/parameter may relate to a hardness of the material located inside the grinding chamber or material which is to be loaded into the grinding chamber.

The calculation module may be configured to: obtain/receive a third variable/parameter which relates to a particle size/granulometry of material located inside a grinding chamber or material which is to be loaded into the grinding chamber, and utilise the first variable/parameter, the second variable/parameter and the third variable/parameter in order to determine the quantity indication. The calculation module may be configured to determine a maximum material quantity/amount which should be allowed to be contained inside the grinding chamber during operation, by utilising the first variable/parameter, the second variable/parameter and the third variable/parameter. The control module may be configured to manage/control the quantity of material inside the grinding chamber during operation such that the quantity does not exceed the allowed maximum material quantity/amount.

The third variable/parameter may change/vary over a period of time. The calculation module may therefore be configured to: obtain/receive a new/updated third variable/parameter on a continual basis over the period of time, and obtain/calculate a new quantity indication when a new/updated third variable/parameter has been obtained/received.

The first variable/parameter and the second variable/parameter may both change/vary over a period of time. The calculation module may therefore be configured to: obtain/receive a new/updated first variable/parameter on a continual basis; obtain/receive a new/updated second variable/parameter on a continual basis; and obtain/calculate a new quantity indication by utilising the new/updated first and second variables/parameters, and wherein the control module may be configured to manage/control the quantity/amount of material contained in the grinding chamber during operation, based on the new/updated quantity indication obtained by the calculation module.

The calculation module may be configured to utilise historical information on what a measured throughput/output of a particular mill was at different times over a period of time, and for each measured throughput/output, what: (a) a state of liners was which were located inside a grinding chamber of the mill at the time,

(b) a hardness of the material located inside a grinding chamber at the time or material which was in the process of being fed into the grinding chamber at the time, and

(c) a particle size/granulometry of material located inside the grinding chamber at the time or material which was in the process of being fed into the grinding chamber at the time.

The system may include one or more sensors for measuring the first and/or second variables/parameters.

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 1A shows a sectional view of a grinding chamber in accordance with the invention, which is underloaded with material (i.e. the charge cell value is too low);

Figure 1 B shows a sectional view of the grinding chamber shown in Figure 1A, which is correctly loaded with material (i.e. the charge cell value is good (i.e. not too low or too high));

Figure 1C shows a sectional view of the grinding chamber shown in Figure 1 A, which is overloaded with material (i.e. the charge cell value is too high);

Figure 1 D shows a side view of a liner which is secured/mounted to an inner surface of the grinding chamber;

Figure 2A shows a graphical illustration of a SAG mill’s throughput (in tonnage per hour (TPH) versus charge cell value (CC) (i.e. the amount of material contained inside the chamber (e.g. as a percentage)));

Figure 2B shows a graphical illustration of a SAG mill’s throughput (in tonnage per hour (TPH) versus charge cell value (CC) (similar to Figure 2A), where the solid line indicates an original curve for a particular SAG mill and the broken line indicates a curve for the same SAG mill, which has shifted as a result of changes in material properties and/or changes in a physical or operational property of the mill;

Figure 3 shows graphical illustration of a SAG mill’s throughput (in tonnage per hour (TPH) versus charge cell value (CC), where a certain desired/target throughput/output level (see the horizontal line “Objective TPH”), a lower charge cell value limit (the vertical line “LL”) and an upper charge cell value limit (vertical line “HH”) are indicated;

Figure 4 shows graphical illustration of a SAG mill’s throughput (in tonnage per hour (TPH) versus charge cell value (CC), where the inverse parabolic curve shown in Figure 3 has shifted, which results in the upper charge cell value corresponding with a throughput which is below the desired/target throughput/output level;

Figure 5 shows a graphical illustration of a SAG mill’s maximum charge cell value for a predetermined desired/target throughput versus granulometry, whereby the area below the curve indicates where the mill should provide a throughput of equal or more than the predetermined desired/target throughput (also referred to as the “stable operation”), while the area above the curve indicates where the mill could provide a throughput lower than the predetermined desired/target throughput (also referred to as the “unstable operation”);

Figure 6A shows a graphical illustration of an example of the grindability curve for a SAG mill (TPH (tonnage per hour vs CC (charge cell value));

Figure 6B shows a graphical illustration of a time series plot example for the TPH, SAG mill states and CC, respectively;

Figure 7 shows a graphical illustration of another time series plot example for the TPH, SAG Mill states and CC;

Figure 8A shows a graphical illustration of a histogram that shows the number of elements for a clusterization of the mineral data, with the aim of separating the mineral data in different families or clusters;

Figure 8B shows a graphical illustration of a box plot diagram with an example of a cluster description;

Figure 8C shows a graphical illustration of a box plot for one label of mineral data for all the different clusters;

Figure 9A shows a graphical illustration of the SAG mill weight over time for a specific study;

Figure 9B shows a graphical illustration of the SAG mill weight over time (same as Figure 9A), as well as the state of liners over the same period of time;

Figure 10 shows a graphical illustration of an example of a granulometry value versus time;

Figure 11A shows a graphical illustration of an example of the data used to generate a part of the grindability curve, for a selection of age of liners (3 or "old"), a mineral characteristics with an SPI near 120 and a granulometry of 50;

Figure 11 B shows a graphical illustration of a table that resume the properties for the sub context (SAG Mill liners age and mineral characteristics) that together with the granulometry defines the context for the grindability curve shown in Figure 1 1 A;

Figure 12A shows a graphical illustration of a recommended CC (charge cell value) versus the granulometry for a subcontext (SAG mill liners age plus mineral properties/characteristics) defined on Figure 12B;

Figure 12B shows a graphical illustration of a table that resumes/summarises the properties for the sub context (SAG Mill liners age and mineral properties/characteristics) that together with the granulometry defines the context for the grindability curve shown in Figure 12A;

Figure 12C shows a graphical illustration of the data for the context (see Figures 12A and 12B), that defines the explored regions of the grindability curve (similar to Figure 1 1 A); Figure 12D shows a graphical illustration of a fractional count of the SAG mill states, together with the counts of data versus the granulometry;

Figure 13 shows a graphical illustration of a graph for a particular sub context, of the recommended maximum CC versus the granulometry; and

Figure 14 shows a schematic layout of the SAG mill system in accordance with the invention.

DESCRIPTION OF PREFERRED EMBODIMENTS

As mentioned, historically, operators of SAG mills would typically focus on the charge cell value that maximizes throughput.

In contrast, the present invention focusses on the maximum charge cell value (i.e. a quantity/amount of material which should be contained in the grinding chamber during operation) at which a certain minimum/desired/objective/given/target throughput (e.g. TPH (tonnage per hour)) is obtained, which the Inventor believes is a more useful value, which is focused on the actual operation of the SAG mill. Reference is in this regard made to Figures 3 and 4. In Figure 3, the two vertical lines “LL” and “HH” indicate the lower and higher cell charge values for a control system, respectively, which have historically been used by the control system to try to keep the actual charge cell value within the boundaries given, and at the same time reaching the desired/objective/given/target TPH, , whereas the horizontal line (“Objective TPH”) indicates, as an example, a throughput of 4500 tonnage per hour. Figure 4 shows an example of where the throughput versus cell charge value curve may have shifted (e.g. as a result of a change in mineral properties/granulometry/state of liners), which results in the lower and higher cell charge values being misaligned and wherein if a SAG mill has a cell charge value close to the higher cell charge value, it would actually result in a throughput significantly less than 4500 TPH (see reference numeral 24). As mentioned, the present invention focusses on the maximum charge cell value at which a certain minimum/desired/objective/given/target throughput (hereinafter referred to as “desired/target throughput") is obtained (see the star indicated by reference numeral 26)

The present invention also provides a new methodology/process whereby the maximum charge cell value (i.e. at which a certain desired/target throughput is obtained) is plotted against a fast-changing variable, such as granulometry, fixed/specific material properties and the stage of wear of liners located inside a grinding chamber of the SAG mill 12. It should be noted that other physical or operational properties of the mill could also be used, such as the mineral hardness. The Inventors believe that this graph provides a very good guide for SAG mill operators. More specifically, the graph provides an operator with enough information to take action based on sudden changes, which is the usual case that leads to throughput losses (e.g. by explicitly showing if, due to a sudden change, a stable throughput (TPH) can be obtained by changing the charge cell limits to a new stable region from the graph).

In the present methodology, in accordance with the invention, SAG mill data is separated into groups. These groups may include different ranges of mineral strength/hardness, stage of liners age (i.e. state of wear) and granulometry. Other mineral properties and/or physical/operational properties of the mill may however also be used. The state of the liners (i.e. the liner’s age or amount of wear) can be referred to as the first variable/parameter which is used, while the mineral strength may be referred to as the second variable/parameter and the particle size/granulometry may be referred to as the third variable/parameter which are also used.

The material strength is, in turn, divided into different groups. In one example, the division can be based on mineral hardness as being the main differentiator. For example, the groups may be soft, medium soft, medium hard and hard. As mentioned, the liners 20 inside the grinding chamber 18 are a tooth-like pattern which cover the inside of the grinding chamber 18 and helps to lift the material 22 inside, to create a cascade of particles that, as it falls, reduces the particle size. These liners 18 wear down with use and are typically replaced approximately every six months. The stage/age of liners 18 (i.e. the state of wear) may be divided into a number of stages. Since these liners 18 may typically have a life of six months, the stage of the wear of the liner 18s can be divided into three stages: New, Medium and Old, each one corresponding to two months of use. The number of stages can obviously be increased (or even decreased to 2) if needed.

The granulometry measures the percentage of material 22 that passes through a sieve of a particular size. In this example the sieve may have a 1.5-inch (about 3.81cm) size. This percentage value can be reduced to the nearest integer, ranging from 20% to 90%, for example.

A common analysis of SAG Mill performance would have been to plot the throughput (TPH) versus the charge cell value, without considering these different categories/groups and search for the charge cell that maximizes throughput, which usually is not clear from the data.

In contrast, the Inventors constructed their analysis by plotting the throughput versus cell charge value for each group, which considered the mineral strength, the stage of liners and granulometry. This was done by using historical data which had been captured which included information on what a measured throughput/output of a particular SAG mill was at different times over a period of time, and for each measured throughput/output, what:

(a) the state of the liners was which were located inside a grinding chamber of the mill at the time,

(b) a hardness of the material located inside a grinding chamber at the time or material which was to be loaded/fed into the grinding chamber at the time, and (c) a particle size/granulometry of material located inside the grinding chamber at the time or material which was to be loaded/fed into the grinding chamber at the time.

In other words, historical information was obtained/extracted for each possible combination of (i) the state of the liners, (ii) the material hardness and (iii) the particle size.

By utilising this information, it was possible to plot the maximum charge cell value for a predetermined desired/target throughput (e.g. a desired/target TPH) (see the star indicated by reference number 26 in Figure 4) versus granulometry, for each subgroup of mineral strength and the state of the liners. In this regard, it should be appreciated that granulometry is typically the faster changing variable (it can change every few minutes) when compared to the stage of the liners which changes over months and the hardness which changes over hours. In this regard, see the example illustrated in Figure 5 which provides a graphical illustration of a SAG mill’s maximum charge cell value for a predetermined desired/target throughput (based on a specific state of liners and a specific material hardness) versus granulometry. The area below the curve indicates where the mill 12 should provide a throughput at the predetermined desired/target throughput TPH (also referred to as the “stable operation”). The area above the curve indicates where the mill 12 could provide a throughput lower than the predetermined desired/target throughput (also referred to as the “unstable operation”).

From the above, it should be clear that the SAG mill system 10, in accordance with the invention, is based on a statistical model (using historical data). In order to improve the system 10, machine learning could be implemented which can help to improve the accuracy of the system 10.

It should be appreciated that, in addition to the above, a lower limit for the charge cell value (e.g. when the cascade hits liners) can also be obtained in a similar manner. This can also then be incorporated in the control module 32 (described further below) for automisation purposes.

The stage of the liners, the mineral strength and the granulometry of the incoming material into the SAG mill 12 are all known. For example: the stage of the liners is known based on when last the liners were replaced (e.g. 1 month ago or 5 months ago); the mineral strength (as well as other information on the mineral content) is obtained from the mine (i.e. upstream). This data is known with about 5 hours of anticipation before entering the SAG mill 12, and each entry is valid for an hour (these timings may however be adapted as needed). The incoming mineral tags were analyzed, and an unsupervised clusterization of the data was performed (i.e. an unsupervised machine learning algorithm was used), revealing that with a few clusters the minerals that enter the mill can be classified. The most important mineral property to classify was the mineral hardness or SAG Power Index (SPI); and the granulometry is obtained from a camera split measurement which scans the material using image analysis just before it enters the SAG mill 12. In one example, the technology used in Belt Metrics™ from MotionMetrics™ could be implemented (https://www. motionmetrics.com/belt-metrics/).

As mentioned, granulometry is a measurement that characterizes the particle size of material 22 entering the SAG mill. It consists of the percentage that passes through a sieve of a particular size. In the present example, the sieve size was 1 .5 inches, which is used as the reference value. The values can range from 0 to a 100, and the lower the number, the coarser the particle size. This quantity typically has a random behavior and can change/vary relatively quickly, when compared to the change in the liner stage or the change in mineral hardness.

By using the three above-mentioned variables/parameters, it is possible to provide a recommendation of the maximum charge cell value for the operator, which should provide the pre-determined desired/target throughput (i.e. the target TPH). In this regard, it should be noted that there may still be some instances where the throughput drops below the pre-determined desired/target throughput. However, this only occurs for short periods of time. In other words, the present invention does not completely eliminate the possibility of the throughput dropping below the pre-determined desired/target throughput, but it does inhibit it to a large degree. This is mainly due to the fact that this approach consists of a simplification of the problem, and there are other relevant parameters that can cause a throughput drop, such as the water content inside the mill, the ratio between mineral and steel balls, the rotational speed, granulometric measurements corrections, etc.

The above calculations can, in one example, be implemented in a calculation module 30 which forms part of the system 10 in accordance with the invention.

It will be appreciated that the operator can also prepare for sudden changes in granulometry/mineral strength or stage of liners using this system 10.

In one example, the system 10 may include a control module 32 which is configured to control the amount of material 22 being fed into the SAG mill 12, by ensuring that the amount of material 22 inside the SAG mil 121 does not exceed the calculated/recommended maximum charge cell value.

From the above, it should be noted that a set of these three variables/quantities (liner state, hardness and granulometry) define a context for which the behavior of the SAG mill 12 defines a grindability curve on a plot of throughput vs charge cell value. This curve has an inverse parabolic shape, having a maximum for a particular value of the charge cell value. The precise shape of this grindability curve is however extremely difficult to determine (as mentioned before).

Each context was then populated with historical data, and given that most of the time the SAG mill 12 was operating with limited throughput, for each context the maximum charge cell at which the throughput was at a selected desired/target throughput level (i.e. a target TPH), was determined/detected (e.g. by using the calculation module 30), as if it were higher than this value, the behavior of the mill 12 would start to follow the right edge of the grindability curve, inevitably decreasing the throughput.

The Inventors believe that this methodology sets a new paradigm for the SAG mill, as all previous research focused only on the charge cell that maximizes the throughput. Furthermore, even if one is able to obtain the charge cell that maximizes the throughput, it might not be useful, as when the SAG Mill operates with limited throughput, it is likely that at that particular charge cell, the particle cascade might be hitting the SAG mill liners, which is undesirable as it does not reduce the particle size efficiently and it wears down the SAG mill liners, thereby reducing the liners’ useful life. The present invention therefore considers the fact that the mill 12 needs to work as heavy as possible without losing throughput. This recommended charge cell value is communicated to an operator, which then assists an operator in operating the SAG mill effectively. In a slight variation, the control/operation of the SAG mill can be automated by using the control module 32 to control the amount and rate of material 22 being fed into the SAG mill.

Furthermore, if one considers that two of the variables/parameters vary slowly (i.e. the SAG mill liners age last for two months and the dispatch information (i.e. the material hardness) is valid for an hour), for each subcontext that these quantities form, it is possible to visualize in a two- dimensional plot how the recommendation will vary with the granulometry, which is the fast-varying variable. This plot will be valid as long as the subcontext (i.e. the SAG mill liners age and the material hardness) remains unchanged. This information will allow an operator to determine, in real-time, what the maximum charge cell value (i.e. the amount of material inside the SAG mill) should be (i.e. given the sub-context). The operator can then use the information to adjust the rate/amount of material entering the mill so that the maximum charge cell value is not exceeded. As mentioned before, in another variation, the control of material being fed into the SAG mill can be implemented by the control module 32, in order to help automate the operation of the SAG mill 12.

Practical Experiment

The experiment had the following objectives:

1 ) Identify contexts given by the relationship between operational variables, ore properties, mill scanner data and SAG mill coating campaign periods that characterize the mill as the bottleneck of the plant; and

2) Identify recommendations and develop a model that help operators stabilize the SAG mill operation, setting the limits for the charge cell weight given the context the mill is passing through.

The data used in this experiment were:

• PI data tags for the SAG Mill variables (2019 to 2021 ). This included TPH, Charge cell value, high and low charge cell limits, RPM (i.e. the rotational speed of the grindignchamber), granulometry, etc., by minute.

• Dispatch data (hardness and mineral content of input raw material) (2019 to 2021 ), by hour. This data is known 5 hours in advance (however this lead time may vary).

Figure 6A shows a graphical illustration of an example of the grindability curve (TPH versus CC (charge cell value)). The parabolic shape indicates the limits of the grindability curve, where only states within the parabolic region are allowed. In this example, the TPH (tonnage per hour) of the plant is expressed as the horizontal line, and the vertical lines correspond to the CC (charge cell value) limits set by the operator (LL: Low Low, and HH: High High). In this graph, the black points 200 indicate a stable operation where the SAG mill is operating within the CC limits, and are within the grindability curve, therefore giving a constant TPH (“constant TPH” in this context refers to a desired/target throughput amount (i.e. a target TPH) (indicated by the line 202)). The star shaped dot 204 is the maximum charge cell value that allows a stable TPH of that value. In this example, this value is set as the HH limit, so the SAG mill can operate as heavy as possible, which is good operational regime. The white dots 206, correspond to CC values that go beyond the HH limit, and start exploring the edge of the grindability curve, forcing a reduction in TPH.

Figure 6B shows a graphical illustration of a time series plot example for the TPH, SAG mill states and charge cell (CC), respectively. The top graph shows the TPH where a horizontal line 208 indicates the objective TPH. The middle graph shows the SAG mill states, where 1 .5 corresponds to the state "Limited by High Tonnage", 1 corresponds to "Limited by Weight", and 0.5 to "Optimizing Tonnage".

The state “Limited by High Tonnage” corresponds to the operational state when the TPH of the plant is constant, and the CC can fluctuate within the operator setting limits. The state “Limited by Weight” corresponds to the operational state when the charge cell exceeds the allowed range for the CC, therefore decreasing the TPH, emptying the SAG mill, until the CC value is within the allowed limits. The state “Optimizing Tonnage” consists in a transition state between the first two, indicating the control system is acting.

The bottom graph shows the charge cell (CC) value, together with the LL and HH limits. From these graphs, it can be seen that whenever there is a TPH decrement (see reference sign 210), it was caused by the CC exceeding the CC HH limit (see reference sign 212). In those instances, the mill changes states, showing that the control system is acting to lower the CC value, until it reaches the allowed region state by the operator. This is an example of not correctly setting the CC limits, as they allow fluctuations on the TPH.

Figure 7 shows a graphical illustration of another time series plot example for the TPH, SAG mill states and charge cell (CC), respectively. Again, from these graphs, it can be seen that whenever there is a TPH decrement, it was caused by the CC exceeding the CC HH Limit. In those instances, the mill changes states, showing that the control system is acting to lower the CC value, until it reaches the allowed region state by the operator. In this example, after 6am, there is a significant change in the CC limits, that clearly allow for a more stable operation.

Figure 8A shows a graphical illustration of a histograms that show a number of elements for an unsupervised clusterization of the mineral data, with the aim of separating the mineral data in different families or clusters. In this example, 10 clusters were considered.

Figure 8B shows a graphical illustration of a box plot diagram with an example of a cluster description. To form the clusters, 8 labels for the data were used, and the graph shows a normalized box plot for each label, all of this for one particular cluster.

Figure 8C shows a graphical illustration of a box plot for one label of mineral data for all the different clusters. The label used in this example is the SAG Power Index (SPI), which indicates a measure of how much energy will be required to reduce the particle size with the SAG mill. For this case, 10 clusters were being considered. The SPI is the most discriminant label for these clusters, and it is closely related to the equipment under investigating.

Figure 9A shows a graphical illustration of the SAG mill weight for the data considered for this study. The instant SAG Mill weight consists of the weight of the mill, plus the mineral and steel balls inside the mill. As can be seen, the signal shows a toothlike pattern, indicating that the weight of the mill starts decreasing, until a point where it increases, to start decreasing again. This is due to the wearing of the liners inside the mill, and their eventual replacement. One set of liners usually lasts for six months.

Figure 9B shows a graphical illustration of the SAG Mill weight for the data considered for this study. As mentioned in Figure 9A, the SAG mill weight shows a toothlike pattern, due to the wearing of the liners inside, due to the mills operation. The thick black lines (see reference sign 230) represent a category that indicates the state of the liners, where 1 indicates that the liners are "new", 2 indicates that the liners are "medium", and 3 indicates that the liners are "old".

Figure 10 shows a graphical illustration of an example of a granulometry value versus time. This value consists of the passing percentage of the material through a sieve. Therefore, it is bound between 0 and 100, and a lower value indicates that a low percentage of material went through the sieve, with the significance of an overall higher particle size. On the contrary, a higher granularity value indicates that a high percentage of the material went through the sieve, being a signal of an overall lower particle size. There are several sieves sizes. However, in the present case, a 1 .5 inch sieve is presented, which is the standard reference value. Under normal conditions, this value is continuous, with a fast-changing pace.

Figure 11 A shows a graphical illustration of an example of the data used to generate a part of a grindability curve, for a selection of age of liners (3 or "old"), a mineral characteristic with an SPI near 120 and a granulometry of 50. Each point represents a data point with a 1 -minute aggregation. The different shadings (dark, medium and light) represent the different states of the mill. The dark dots (see reference sign 232) indicate "Limited by Weight", the light dots (see reference sign 236) indicate "Limited by High Tonnage", and the medium dots (see reference sign 234) indicate the highest CC at which a stable TPH is found for a given HH value. The black line 240 represents the top edge of the data, that represents the flat line (see reference sign 244) with the maximum TPH, and the data leaned to the right (see reference sign 246) represents in this case the edge of the grindability curve. Reference is in this regard again made to Figure 6A. The vertical line 242 represents the maximum CC at which a constant TPH is found, therefore being the recommended CC for this particular example.

Figure 1 1 B shows a graphical illustration of a table that resumes/summarises the properties for the sub context (SAG Mill liners age and mineral characteristics) that together with the granulometry defines the context for the grindability curve in Figure 1 1 A. Figure 12A shows a recommended CC vs the granulometry for a subcontext (SAG mill liners age plus mineral properties) defined in Figure 12B. Figure 12C shows the data for the context, that defines the explored regions of the grindability curve (please refer to the explanation in respect of Figure 1 1 A). Figure 12D shows a fractional count of the SAG mill states, together with the counts of data versus the granulometry. In Figure 12D, the dark area (see reference sign 250) indicate "Limited by Weight", the light area (see reference sign 252) indicate "Limited by High Tonnage", and the medium area (see reference sign 254) indicate the highest CC at which a stable TPH is found for a given HH value. From Figure 12D, it can be seen that as the granulometry value increases (i.e. as the particle sizes become smaller), the SAG mill spends more time in the state "Limited by High Tonnage", meaning that it had more frequently a stable TPH. The white segmented line (see reference sign 26) is the count of the number of datapoints versus the granulometry, showing that most of the data is near 50, with rare events on the tails.

Figure 13 shows a graphical illustration of a graph for a particular sub context, of the recommended maximum CC versus the granulometry. The thick line 270 represents the data extracted from the individual grindability curves and the segmented line 272 is an extrapolation. The main effect that can be seen from this graph is that, in the case of facing larger particles (smaller granulometry), there is the possibility of increasing the HH (upper charge cell value limit), as there are stable TPH regions.

A summary is hereby provided of the whole process/experiment:

• The SAG mill data was separated into groups, determined by different ranges of mineral strength, stage of liners age and granulometry.

• The material strength was divided into four groups, being the main differentiator the mineral hardness. For example, the four groups may be classified as soft, medium soft, medium hard and hard. • As the duration of the SAG mill liners are usually six months, the stage of the wear of the liners was divided into three stages: new, medium and old, each one corresponding to two months of use.

• The granulometry measures the percentage of material that passes through a sieve, in this case, a 1 ,5-inch size. This value was reduced to the nearest integer, ranging from 20 to 90.

• A common analysis of SAG mill performance was to plot the TPH vs the charge cell without considering these different categories and search for the charge cell that maximizes TPH, which usually was not clear from the data.

• The present analysis constructed the TPH vs charge cell graph for each group, that considered a mineral strength, a stage of liners and granulometry, and used that information to extract the maximum charge cell where a stable TPH was found.

• With this information, the maximum charge cell with a sable TPH vs the granulometry was plotted (granulometry is usually the fastchanging variable), for a subgroup of mineral strength and stage of liners.

• Given that the stage of the liners, the mineral strength and granulometry of the incoming material into the SAG Mill are known, a recommendation of the maximum charge cell can be calculated and communicated to an operator using this graph.

• Furthermore, the operator can prepare for sudden changes in granulometry/mineral strength or stage of liners using the present invention.

It should be noted that, although the example(s) set out above referred specifically to a SAG mill, it should be appreciated that it can also be applied to other types of grinding mills as well, such as AG (autogenous) grinding mills.