The present invention refers to the control of operation of rotary mills in the mining industry, more specifically an electronic system and method for the on-line estimate of the balls filling level and the loading level in a rotary mill.
Grinding is the last step of the comminution process. In this step, the size of the ore particles is reduced through a combination of impact and abrasion, whether dry or moist. The equipment mostly used at industrial level for milling is the horizontal rotary mills.
There are different types of rotary mills, such as rod, ball, autogenous (AG) and semiautogenous (SAG). All of them characterize for a tumbling action allowing the comminution of the material. Structurally, each type of mill consists of a cylindrical shell provided with renewable liners and a load of grinding media. The length and diameter of the mill determine the volume and, therefore, the capacity of the equipment. All types of mill can be used for wet or dry miffing by modifying the feeding and discharge areas.
SAG grinding circuits are constituted by a primary crushing step, followed by a SAG mill commonly associated with a stone crusher (pebbles) and then a ball mill. These circuits are those mostly used today because of the advantages they have compared to conventional grinding circuits, which were constituted by three stages of crushing, followed by a rod mill and a ball mill.
While global energy consumption in [kWh/ton] of product is comparable for both systems, tending to be a little higher for SAG systems, experience has shown that a line incorporating the SAG grinding will require a lower capital cost than the traditional line. There is also a decrease in maintenance cost due to the elimination of secondary and tertiary crushing stages.
SAG mill operation involves reducing the consumption of grinding media, compared with the costs associated with the consumption of balls and rods in a conventional circuit which also have a higher wear of coatings.
SAG mills are large-capacity equipment developed before the need to process larger flows due to lower grades of valuable minerals. Currently, there are SAG mills up to 12.19 [m] diameter and 20 [MWh] power in operation, where economies of scale can be used. SAG mills are more efficient than rod mills, in which there has been no success in increasing its size beyond 6.71 [m] long, because of excessive breakage and locking of bars when increasing its length.
The load inside a mill is usually called kidney, due to the characteristic form taken by it when the equipment is in motion. In the behavior of the kidney in a mill rotating counter clockwise, where the load rises on the right side to a point where it falls again, this point is known as the load shoulder. At the bottom, the load foot is located, which characterizes for a chaotic movement of the landing load, which dissipates the remaining energy of the fall to be raised again.
Two regimes of fall are defined, which will depend on the particle position and the mill's speed of operation, one known as cataract for free fall and another one as cascade for collapsing on other bodies.
To prevent slippage of the load, which would imply an inefficient use of energy, the mill shell has an inner liner provided with projections (lifters) and depressions (plaques). The profile of this liner highly influences the movement of the load and the path of falling particles. In addition, in conjunction with the viscosity of the pulp in the mill, the liner profiles can have a substantial effect on the actual critical speed.
Generally, the models of power consumption for mills do not include the effect of lifting bars design, although some designs have proven to give a greater cataract effect than others at the same fraction of critical speed and filling level. Therefore, they should give maximum power at different values of filling and speed.
Considering the behavior of power consumption in rotary mills, it can be concluded that this equipment should be operated at filling values and critical speed dose to maximum power consumption, as this determines the maximum production capacity of the mill.
The filling level in a mill corresponds to the useful internal volume of the equipment being occupied by the load, consisting of balls, mineral and water. Keeping the proper level of charge in the mill is one of the most important elements for efficient grinding. During the operation, you must ensure that the liners are protected from the direct impact of the balls. This is achieved by maintaining an optimum level of loading, supplemented by a control of speed that allows the point of impact being produced at the foot of the load, benefiting the grinding action, as both a low load and overload will harm the process of grinding.
The parameters most commonly used to describe the filling level are those corresponding to the load level and the level of grinding media (balls). Both parameters are calculated as the ratio of the volume used by the load bed or ball bed, and the available volume of the mill, taking into account the porosity of the bed.
Anglo American Chile at El Soldado mine describes a development of 2014 by Nelson Iglesias, Francisco Vicuna and Miguel Becerra, for estimating the level of total load filling (Jc) using a C-Model of Morrell (Morrell. 1993) to estimate the power of a SAG mill, which is calibrated by the laser scanning technology called FARO, where the measuring of load and balls of the filling levels is carried out by stopping the grinding being performed. The C-Model is essentially a balance of friction between concentric layers within the rising part of the mill load.
The patent of invention U.S. Pat. No. 6,874,364 B1, dated 5 Apr. 2005, entitled “System for monitoring mechanical waves from a moving machine”, Cambell et al, discloses a system for monitoring the mechanical waves of a machine, which during its operation moves the material therein; the system includes at least one sensor on the machine in a location away from the central axis of the machine, the sensors detect acoustic waves and include a transmitter for transmitting signals representing the mechanical waves detected to a receiver at a remote location from the sensor(s), a data processor connected to the receiver for receiving signals from the signal receiver representing the mechanical waves and processing the signals to produce output signals for display on a screen, wherein the output signals for display represent one or more parameters indicating mechanical waves emitted from the machine during a predetermined time.
The patent of invention U.S. Pat. No. 6,874,366 82 dated May 4, 2005 entitled “System to determine and analyze the dynamic internal load in revolving mills, for mineral grinding”, Pontt et al, describes in a dynamic way a direct system and method and online measurement of various parameters related to the volume occupied by the internal dynamic loading of rotating mills when being in operation, specifically an online measurement of the total filling of volume dynamic load, the dynamic volume load of balls, dynamic volume mineral filling and the bulk density of the internal load of the mill. The present invention comprises a number of wireless acoustic sensors connected to the outer body of the mill, a receiver and/or conditioning unit situated near the mill, a processing unit and a communication unit.
The patent of invention U.S. Pat. No. 5,698,797 patent, dated 16 Dec. 1997, entitled “Device for monitoring a ball grinder”, Fontanille et al, describes a device for controlling a ball mill having a cylindrical casing containing a mass of balls, which, when the mill is rotating at its nominal speed, takes up a position between two generatrices (1b, 1b′) separated by an angle between a minimum alpha min angle and a maximum alpha max angle and a mass of coal, which, when the mill rotates at its nominal speed, occupies a position between two generatrices (1c, 1c′) separated by a beta angle. An emitter of electromagnetic waves is arranged inside the mill and at least one receiver is arranged outside the mill. The receiver is associated with an electronic circuit to determine at least one parameter of the mill selected from the amount of balls, the amount of coal and the cylinder wear.
No other evidence has been found referring to the operation control of rotary mills in the mining industry, more specifically to an electronic system and method for the online estimate of the balls filling level and the load level in a rotary mill, by sensing the vibrations that occur inside the mill.
In a first object of the invention, a system for online estimate of the filling level of balls and load level in a rotary mill is proposed, comprising a set of vibration sensors unit, associated with a magnet, which transmits signals of data via an antenna to a receiver that receives data signals from the antenna, where the receiver is connected by optical fiber to a signal processor which in turn communicates to a control server via a UTP cable, wherein the server also obtains operation data from the mill to determine the filling of balls and load; in addition, a second object of the invention is proposed, as well as a method for the online estimate of the filling level of balls and load level in a rotary mill comprising as follows: obtaining the operational data of a mill and the data from a signal processor; where data of the mill operation correspond to input grain size distribution of fresh ore, tonnage of fresh ore, water supply; rotation speed of the mill, recharge of balls, average power and pressure on the mill bearings, among the most relevant; and wherein the data from signal processor correspond to the average and variance of the mill vibrations, which were obtained by a first processor; calculate the instantaneous power p(t) consumed by the mill and determine the foot and shoulder of the mill load; assess models of foot, shoulder, power, pressure, wear and tear of liners and wear of ball; iterate the values of balls filling and load filling in the mentioned models; and compare the prediction from models with values measured throughout the system, and the values obtained from the mill operation, to achieve a minimum error in the set of variables.
The present invention relates to a system and method for online estimate of the filling level of balls (62) and loading level (63) in a rotary mill.
As shown in
The set of vibration sensor unit (10) processes the information received from the plurality of accelerometers (14) in the first processor (13); the accelerometers (14) have different ranges of operation, so that when there is saturation in any of them, the information is valid in at least one of them, in the understanding that the operating ranges cover the entire spectrum of vibrations produced by the mill (60) during operation.
The first processor (13) collects the information from the accelerometers (14) for a number of turns preset and processes this information covering the entire circumference of the mantle of the mill (60) and sends all values during the interval of time available for each sensor of vibrations (10, 10′); the values obtained allow determining the shoulder (66) and foot (65) of the volume of load in the control server (50).
The data obtained by the accelerometers (14) that are processed by the first processor (13), together with those of sound obtained by microphones (11), and the synchronizing signal for a predetermined amount of turns allows the first processor (13) determining the direction of rotation, the quadrant where the foot (65) should be located and the quadrant where the shoulder (66) should be located. With sound intensity it is verified that the quadrants are the right ones. Then, the average of amplitude and variance of vibrations are calculated for the entire circumference of the mill (60) and for each turn. Then an average between turns is obtained, getting an average and a variance of a turn that represents them all. Subsequently, the line considered with the receiver (30) is awaited and in the transmission time defined before the results obtained are sent along with the possible quadrants for foot (65) and shoulder (66).
The operating method shown in
Step (72) allows the calculation of the instantaneous power p(t) consumed by the mill (60) and determining the foot (65) and shoulder (66) of the load (63) of the mill (60); wherein in order to determine the foot (65), first a first order filter is performed to remove noise, the variance maximum values are sought within the applicable radial quadrant and finally the radial location of most energy is determined, which corresponds to the foot (65) of the total load.
To determine the shoulder (66), first a first order filter is performed to remove noise, the maximum values of reverse variance are sought within the applicable radial quadrant and finally the radial location of most energy is determined, which will correspond to the shoulder (66) of the total load.
Step (73) allows assessing the models of foot (65), shoulder (66), power, pressure, and wear of liners and of balls (62). The models of foot (65) and shoulder (66) correspond to models defined by parametric equations, for example, those described by Morrell, where an angle of foot (65) and shoulder (66) is obtained as a function of loading (63) and balls (62) filling. The model of power available in the state of art and also described by Morrell allows obtaining the power consumed by the mill (60) depending on the filling of loading (63) and balls (62). The model of pressure on the bearings of mill (60) corresponds to linearization of the statistical behavior during the years of operation of the mill (60) where a pressure that depends on the filling of load (63) and balls (62) is obtained. The model of liner wear corresponds to linearization of the statistical behavior during the years of operation of the mill (60) where the actual lining wear during operation is measured and average wear is obtained. The model of balls (62) wear corresponds to linearization of the statistical behavior of consumption of balls (62) over the years considering the sizes of input and output sizes of the balls (62), thereby calculating the wear average.
Step (74) allows iterating the filling values of balls (62) and load (63) within the aforementioned models, and comparing the prediction of models with the values measured by the system and the values obtained from the operation of the mill (60) until achieving a minimum error in the set of variables.
Finally, the step (75) allows obtaining the optimal values for the filling of balls (62) and filling of load (63) for which the overall error is minimal.
Number | Date | Country | Kind |
---|---|---|---|
1220-2015 | May 2015 | CL | national |
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/CL2016/000021 | 5/6/2016 | WO | 00 |
Publishing Document | Publishing Date | Country | Kind |
---|---|---|---|
WO2016/176788 | 11/10/2016 | WO | A |
Number | Name | Date | Kind |
---|---|---|---|
5698797 | Fontanille et al. | Dec 1997 | A |
6874364 | Campbell | Apr 2005 | B1 |
6874366 | Magne Ortega et al. | Apr 2005 | B2 |
20040255680 | Ortega | Dec 2004 | A1 |
20080097723 | Pontt Olivares | Apr 2008 | A1 |
20140150524 | Magne et al. | Jun 2014 | A1 |
Number | Date | Country |
---|---|---|
2378570 | Jan 2001 | CA |
Entry |
---|
Araya Lara, Hugo Boris; International Search Report and Written Opinion of the International Searching Authority, issued in International Application No. PCT/CL2016/000021; dated Sep. 6, 2016; 14 pages. |
Spencer, S.J.; Acoustic Emissions Monitoring of SAG Mill Performance; Proceedings of the Second International Conference on Intelligent Processing and Manufacturing of Materials; Jul. 10-15, 1999; 8 pages. |
Huang, P. et al.; A Study on the Technique of Measuring the Fill Level Based on the Vibration Signal of the Ball Mill Shell; 2009 WRI World Congress on Computer Science and Information Engineering; Mar. 31-Apr. 2, 2009; pp. 202-206. |
Sagmilling.com; Morrell C-model, developed as part of PhD thesis of Stephen Morrell, University of Queensland retrieved from https://wiki.sagmilling.com/index.php/Morrell_C-Model on Apr. 3, 2018; page last modified on Jun. 18, 2016; 2 pages. |
Number | Date | Country | |
---|---|---|---|
20180126384 A1 | May 2018 | US |