Note: Descriptions are shown in the official language in which they were submitted.
CA 022~71~8 1998-ll-27
WO 97/45203 PCT/US97/08871
METHOD AND APPARATUS FOR CONTROLLING
FROTH FLOTATION MACHINES
Back~round of the Invention:
1. Field of the Invention
This invention relates generally to froth flotation m~r~ines for the separation of
particles from a liquid slurry or pulp. More particularly, this invention relates to methods and
apparatus for ~lltom~tic~lly monitoring, operating, and controlling continuous feed flotation
machines using "intelligent" computer control systems and remote sensing devices,
particularly laser spectroscopy type sensing devices.
2. Brief Description of the Prior Art
Flotation machines are used in many industrial applications for separation of
particulate materials from suspensions in a liquid, usually water. The particles to be removed
from the suspension are treated with reagents to render them hydrophobic or water repellant,
and a gas, usually air, is admitted to the suspension in the form of small bubbles. The
hydrophobic particles come into contact with the bubbles and adhere to them, rising with
them to the surface of the liquid to form a froth. The froth containing the-floated particles is
then removed as the concentrate or product, while any hydrophilic particles are left behind in
the liquid phase and pass out as the tailings. Flotation machines find particular utility in the
metals recovery industry, providing superior recovery of metals or metallic minerals from a
solid/liquid mixture known as a "pulp," "slurry," or "gangue." The flotation process can also
be applied to the removal of oil droplets or emulsified oil particles, as well as to fibrous or
vegetable matter such as paper fibers, bacterial cells, and the like.
In most applications, reagents known as collectors selectively render one or more of
the species of suspended particles hydrophobic, thereby ~c~i~ting the process of collision with
~ and collection by the air bubbles. It is also usual to add frothing agents to assist in the
formation of a stable froth on the surface of the liquid. The process of adding various
reagents to the system is known as conditioning.
Flotation machines have been developed in a number of different, known
configurations. In some conventional embodiments, the flotation machine includes a
CA 022~71~8 1998-ll-27
WO 97/45203 PCT/US97/08871
receptacle, a cell or tank with subsl~llially vertical walls, and an inner rotating member
known as a rotor-disperser. The rotor provides agitation to ~ 1 suspension of the pulps,
and may also draw external air into the tank through a standpipe. The disperser breaks the
air into minute bubbles and disperses it uniformly through the pulp, while also providing
S meçh~nical mixing of the air and pulp. Such cells may also include a false bottom and a draft
tube to provide a ~ ed flow path, ensuring maximum slurry recirculation and
air/slurry mixing. In other conventional embodiments, the rotor functions only for agitation,
and aeration is provided by an external means, usually a blower or compressor.
Alternatively, air may be dissolved into a liquid, which is then injected into the pulp or slurry.
The air used may be atmospheric air, or an inert gas such as nitrogen or argon. Other
proposed flotation cells include a tank and a means for generating ultrasound, which acts to
agitate the pulp and thereby achieve solids separation, as is described in U.S. Patent No.
5,059,309 to Jordan. Alternatively, separation of particles may be achieved by a combination
of air, and magnetic and/or electrical fields as described in U.S. Patent No. 5,224,604 to
Duczmal and Schneider.
Another known configuration of a flotation m~chine is the column. In a column, the
conditioned suspension is introduced toward the top of the receptacle, a tall vertical column,
and air bubbles are formed in the bottom of the column by blowing pressurized air through a
diffuser. A layer of froth bearing the floatable particles forms above the liquid and overflows
from the top of the column. The position of the froth-liquid interface is m~in~ained at a
desired level by controlling, for example, the flow of liquid from the bottom of the column.
Optionally, wash water is introduced near the top of the froth layer to create a downflow of
liquid which tends to reduce the ~ railllllent of undesired gangue particles in the froth
overflow. In these types of flotation columns, the liquid flows downward while the bubbles
rise vertically upward. Since the rise of bubbles is related strongly to their size, the bubbles
must be above a certain critical diameter to rise through the liquid and into the froth layer.
Various other alternative embodiments of the colurrm flotation m~chine have been described,
for example in U.S. Patent No. 4,938,865 to Jameson, which introduces an air/slurry mixture
into the column where separation takes place.
.
CA 022~71~8 1998-11-27
WO 97/45203 PCT/US97/08871
~ F.fficjent and effective operation of flotation m~chines requires ll~o~ g and
controlling a multitude of process and operational parameters. By '~process parameters" is
meant such parameters as slurry levels and bubble size, as described above. Other process
parameters include but are not limited to the density of the pulp in the chamber of the
flot~ti~n m~rhine, bubble concçntration and distribution, product and tailings removal rates,
reagent addition and consunnption rates, air flow, solids concçntration~ froth mass and
volume, froth level, pulp level, feed rate, and the like. "Operational parameters" is meant to
include various flotation machine operating parameters such as rotor speed and position,
draft tube position, crowder position, power consumption, and the like. These classifications
and examples are for convenience and example only.
Flotation m~chines present challenging problems with respect to the design and
inst~ n of sensors associated with the flotation machines, the ac~uisition of various
me~ure...ents, the ability to communicate data and power into and out of the flotation
m~rhine, as well as the ability to provide control devices within the m~c.hine and actuate
those control devices in response to a command from a central control computer. A special
challenge has been to improve efficient control of the various above-~liccu~sed parameters.
Each of these parameters must be adjusted to optimize both the economic operation of the
plant, as well the operating conditions, i.e., efficient throughput and desired levels of
purification.
Flotation machines are normally controlled by simple, feedback or feed forward
control loops. Various devices have been described which may be used to monitor important
parameters in flotation m~rhine operation. The most common of these describe a flotation
m~chine separation control system comprising a controller (e.g., a microprocessor) which
commnnicates with one or more sensors and in response to information received from the
sensors, ~ch~t~s a control apparatus (e.g., a valve) to adjust one or more control parameters.
For example, a control system addressing the level of pulp is described in U.S. Patent No.
4,343,654 to Lambert, wherein a computer communicates with a level sensor, and based on
these signals and preprogrammed instructions from the computer's memory, sends control
signals to a venting element which regulates air venting and thereby adjusts the degree of
frothing.
CA 022~71~8 1998-11-27
WO 97/45203 PCT/US97/08871
U.S. Patent No. 5,011,595 to Meenan uses feed forward and feeclb~c~ control
methods to detect solids c~ ncentrations and adjust the rate of chemical addition to feed
streams. The control system includes optoelectric detectors responsive to di~rellt solids
concen~rations and slurry parameters. The detectors forward signals regarding the solids
c~ llc~ ions and slurry p~a,l,e~ to a process controller. In response, the process
controller adjusts the rate of addition of chemicals to the feed stream of the froth m~rhine to
control the separation of solids from impurities. The controller calculates a feed forward
output from the signals and the controller output adjusts the addition of ~lilrelent chemicals or
additives to the processing machine The controller also calculates a feedbac~ output after
receiving a signal from a third detector which monitors the extent of separation and recovery
of solids from the processing machine. Related patents include U.S. Patent No. 4,797,550 to
Nelson and Oblad and U.S. Patent No.4,797,559 to Oblad, et al., which disclose a method
and apparatus for determining the reflectivity of the tailings from a coal flotation machine
using a laser and bifurcated fiber optic cell to deliver signals. Another control system
directed to controlling the quantity offroth is disclosed in U.S. Patent No. 5,062,964 to
Ortner and Pfalzer. A probe measures the level or amount of foam, and transmits a signal to
a controller or regulating apparatus, which in turn controls the amount of air introduced into
the system by way of a valve. Control of multiple parameters is achieved in U.S. Patent No.
3,551897 to Cooper, which describes measurement of various operating conditions, then
c~lcul~~ion of a pluralit~ of coefficients that together with certain equations describe the
process at a particular point in time, and then utilization of the same equations to adjust
certain parameters in order to optimize operating conditions and maximize profitability.
For the most part, the control systems of the prior art address only one or two
process or operational parameters, for example controlling the level of liquid or froth. In
U.S. Patent No. 5,073,253 to Bishop and Gray, a float supported by the froth works in
tandem with an ultrasonic level detector to provide a measurement of froth level. U.S. Patent
No. 4,938,865 to Jarneson utilizes a controller to operate a valve that introduces air into the
top of a column. US. Patent No. 4,552,651 to Sandbrook and Scandrol discloses devices to
measure pulp density and pulp level. The two signals are then combined to a single signal
which is utilized to control liquid level in the machine by adjusting the rate of withdrawal of
CA 022~71~8 1998-11-27
- First Replacement Page 5-
tailings. U.S. Patent No. 5~ 19~,4~3 to Duczmal and Schneider use a control device to maintain the
desired licluid level and to optimize the collection of froth. U.S. Patent No 1,797,550 and the paper
entitled "Fuzzy Model Based Control for a Mineral Flotation Plant" by A. Cipnano and M. Ramos
represent the closest prior art to the present invention.
Other simple controllers have been used to measure and adjust the density of feed or pulp.
For example, U.S Patent ~o. 5,36~,166 to Chumak, et al. discloses a control device to me~sure the
level and density of pulp and to control the flow rate of water and t'rothin~J agent. P~ differential
densitometer tor continuously measuring total undissolved solids in a li~luid in IJ'.S. Patient ~o
5,417,102 to Prevost.
The rate of addition of chemical reagents for conditioning is one of the more important
process parameters to control. This rate affects both the quality of the product (i.e., the amount of
mineral extracted and the punty), as well as the cost of the overall process. There have been a
number of disclosures addressing this parameter, in addition to U.S. Patent ~o. 5,011,595
discussed above. U S. Patent No. 4,~10,371 to Fonesca discloses a system to control the coal
content of coal tailings, including detecting the coal content of the tailings from the flotation cell and
controlling the supply of additives to the machine to optimize slurry coal recovery. .~ method and
apparatus for sensing variations in solids content in at least one output stream and then adjusting the
amount of flotation reagent is described in U. S. Patent No. 4,73 l,176 to Macdonald. A particle
size analyzer is used in U.S. Patent ~o. 4,559,134 to compare a size analysis of solid particles in a
separated stream with a size analysis of solid particles in a feed stream. The rate of addition of
collector reagent is adjusted in response by a controller.
An important operating parameter is the power consumption of the flotation machille.
Maximizing efficiency of the machine by proper rotor, disperser, and draft tube placement could
represent significant savings in the cost of operahon. However, it is not believed that any of the
pnor art has addressed a control system specifically designed to minimize power consumption
while maximizing efficiency and purity of product.
Furthermore, it is not believed that any of the aforementioned prior art provides a
comprehensive, computerized, "intelligent" control system for operating, controlling, and
monitoring various flotation machines. The ability to provide precise, real time control and
monitoring of such flotation machines constitutes an on-going, critical industrial need.
~1~4G'lDF3 SH~T
_
CA 022~71~8 1998-11-27
wo 97/4s203 PCT/US97108871
SummarY of the Invention:
The above~ ccucsed and other problems and deficiencies of the prior art are
overcome or alleviated by the several methods and apparatus of the present invention for
providing computerized, "intelligent" systems for operating, controlling, monitoring and
S r1iagnocirlg various parameters and processes of flotation m~chinçs By "intçlligent" is meant
use of computerized, control methods including but not limited to neural networks, genetic
algorithms, fuzzy logic, expert systems, statistical analysis, signal processirlg, pattern
recognition, categorical analysis, or a combination thereof. Thus in preferred embodiments
(but not necçss~rily all embodiments), this invention comprises at least one of these control
methods and other metho~lc more advanced than convçntion~l, stabilizing control. An
intelligent flotation machine of the type herein disclosed has the capability of sensing
information about itself, predicting its own future state, adapting and ~h~ngin~ over time as
process and operational conditions chan~e, knowing about its own performance, and
rh~nginsJ its mode of operation to improve performance. Specifically, the control system of
the present invention regularly receives instrument rea~lin~c7 digitized video images, or other
data indicating the state of the flotation machine; analyzes these readings in terms of one or
more self-gt:,lelaL~d, continuously updated, internal models; and makes changes in operating
variables as suggested by the internal models
In accordance with the present invention, a computer control system actuates at
least one of a plurality of control devices based in part on input from one or more monitoring
sensors so as to provide real time continnous operational control.
It will be appreciated that it is difficult to sense and communicate certain
parameters in real time within flotation machines. Thus, in accordance with an important
feature of the present invention, a variety of technologies including ultrasonic absorption and
reflection, laser-heated cavity spe.,~ sco,oy, laser-induced breakdown spectroscopy (LIBS),
laser-induced mass spectroscopy (LIMS), X-ray fluorescence, neutron activation
spectroscopy, I)r~;s:ju,e measurement, microwave or millimeter wave radar reflectance or
absorption, and other optical and acoustic methods may be utilized.
In a preferred embodiment, the sensor or sensors comprise a means for determining
input stream composition and particle size. Ideally, such a sensor would provide the data
CA 022S71S8 1998-11-27
Wo 97/45203 PCT/US97/08871
regar.ling the input stream without the necessity of removing samples from the process flow
in order to be analyzed at a sep~l~ location. Thus, a preferred sensor includes sensors
lltili7ing laser spectroscopy (e.g., laser-induced breakdown and laser-induced mass
spe-;Lloscol"~). Such sensors may osçill~te in an arcwise path, or move linearly along the
process flow or radius of the tank to provide a profile of the process stream without the
necessity of removing individuals samples from the process stream. Alternatively, multiple,
spaced sensors may be used to obtain a complete process stream profile.
A particularly preferred embodiment of the present invention employs an im~ging
system CO~lp~ lg video cameras producing images which are converted to data usable by
the process models of the present invention. This embodiment will further comprise an
advanced control system employing both pattern analysis by neural networks, as well as
statistics and color vector analysis. Mapping of high-dimensional input vectors to lower-
riin~en.~ion~l maps in a topological order-preserving manner by these advanced control
systems can be used to track to performance of a flotation process on a continuous basis,
which is highly advantageous when monitoring banks of flotation cells.
The computer controller used in the system of the present invention is preferably a
personal computer or workct~ion, with a associated display device (CRT screen) and
input/output device (keyboard or touch-sensitive screen). The controller may be located at
the froth flotation machine or at a remote location such as a central control room in a plant.
Importantly, the controller may control one or a plurality of flotation m~ ines at a single or
plurality of sites.
The above-described computerized control and monitoring system for froth
flotation equipment provides a comprehensive scheme for monitoring and controlling a
variety of input and output parameters as well as a plurality of operational parameters
resulting in greater efficiency, optimization of operation, and increased safety.
The above-discussed and other features and advantages of the present invention will be
appreciated and understood by those skilled in the art from the following detailed description
and drawings.
CA 022~71~8 1998-11-27
WO 97/45203 PCT/US97108871
BriefDescription ofthe Drawin~s:
Referring now to the drawings wherein like elements are numbered alike in the
several FIGURES:
FIGURES lA - D are schem~tic sectional views of flotation cells and flotation
columns with which the monitoring and control system of the present invention is used;
FIGURE -2 is a sch~m~tic view of the monitoring and control system for a flotation
m~-~hine in accordance with the present invention;
FIGURE 3 is a schen1atic view of a preferred monitoring and control system
employing a LIBS or LIMS sensor system in accordance with the present invention;FIGURE 4 is a schenn~ic view of a monitoring and control system employing a
LIBS or LIMS sensor for monitoring the composition and particle size of a dry or dewatered
process flow according to the present invention;
FIGURE 5 is a schematic view of a monitoring and control system employing a
LIBS or LIMS sensor for monitoring the composition and particle size of a dry or dewatered
process flow according to the present invention; and
FIGURES 6A-B are schematic views of a monitoring and control system
employing a LIBS or LIMS sensor for monitoring a wet or moist process'stream in
accordance with the present invention.
Description of the Preferred Embodiment:
This invention relates to methods and apparatus for automatically controlling,
operating, and monitoring flotation maf hint?c using "intelligent" computer controlled systems
and remote sensing devices. By "intelligent" is meant the use of computerized control
methods including but not limited to neural networks, genetic algorithms, fuzzy logic, expert
systems, statistical analysis, signal proc~ssin~ pattem recognition, categorical analysis, or a
con,b,l,al,on thereof to analyze input in terms of one or more self-generated, continuously
updated, internal models, and to make changes in operating variables as suggested by the
models. It is to be understood that the term flotation machine is used in its most general
sense, being inclusive of traditional flotation cells, or flotation columns, wherein flotation may
be accomplished by a variety of means, including air, ultrasound, magnetic or electrical
CA 022~7158 1998-11-27
Wo 97/45203 PCT/US97/08871
fields, or a combination thereof. It is further to be understood that flotation m~r.hine in the
context of the present invention may refer to a single cell or column, or to a bank of cells or
columns.
Referring to FIGURES lA-D, simplified examples of flotation m~c.hinec
S co.. t~-n~ ted by the present invention are shown. In FIGURE lA, a co,-""on flotation cell is
shown at l O. Flotation cell l O includes an impeller 12 mounted centrally for rotation about a
vertical axis ~ija~ent, but spaced from the bottom of receptacle 14, and having a pulp feed
trough 16 from which a feed tube 18 extends downwardly to a position just outside a
stabilizer 20. Con-lition~ng agents are received through the lower end of pipe 22. Pulp is
l O eventually discharged through outlet 24.
Figure IB depicts a flotation cell wherein an injection device 30 is mounted to
expel a two-phase gas-liquid mixture into the cell. FIGURE 1 C shows this cell mounted for
operation in a bank of cells.
FIGURE lD shows a typical flotation column 32 contemplated for use with the
present invention. Flotation column 32 comprises a column 33 provided with a source of
aeration 34 and wa~sh water 35. Aeration (bubble generation) may be achieved either directly
through internal ~palge,~ or after external cont~r,ting of gas with water or slurry. Wa~sh
water is usually added from an array of perforated pipes located just below the froth overflow
lip. Feed 36 is introduced about one-third of the way down form the top of column 33, and
descen~l~ against the rising bubbles from aeration source 34. Froth concentrate 37 overflows
the top colurnn lip, while tailings 38 exit from the bottom of the flotation column.
In accordance with the present invention, flotation m~rhines of the type discussed
above and in the prior art are provided with one or more sensors for the sensing of one or
more parameters related to the processes and operation of the flotation m~nhine In addition,
a computerized control system which may be located at the flotation machine, near the
flotation m~chine, or at a remote location from the flotation machine is provided for
illlela~;lion with the sensor or sensors in the flotation machine. This computer control system
includes a control computer and one or more control devices which are act--~ted in response
to a command signal from the control computer. Importantly, the response of the control
system will preferably be based both on sensor input and on a series of expert rules,
CA 022~71~8 1998-11-27
WO 97/45203 PCT/US97/08871
-10-
determined initially in advance and cc ntin~ ly updated based upon the control system's own
analysis of its pe,~l",ance. The controller will generate and c~ntin~lQusly update its own
"process model," using the data inputs described and one or all of several advanced analysis
techniques, including neural networks, genetic algo,ill--"s, fuzzy logic, expert systems,
S st~ti.~ti~l analysis, or a co"-binalion of these. The control system will have the ability to
independently select the best analysis technique for the current data set. The computer
control system will actuate one or a plurality of control devices based on input from one or
more monitoring sensors so as to provide real time, contin--ous, operational control. In
addition, the control system may include a monitoring system for data logging, preventative
m~intçn~r~ce, or failure and wear prediction. The control system may additionally include
~i~gnostiçs relating to the condition of the equipment.
Referring now to FIGURE 2, a schematic is shown depicting examples of the
monitoring sensors, control devices, and components and features of the control system of
this invention. FIGU~E 2 more particularly shows a flotation machine 40 having associated
therewith one or more process sensors 42 and/or one or more equipment sensors 44,
incl~ optional video cameras (or im~ine devices) 46. In addition, the flotation m~hine
is ~c.~oci~ted with one or more operational control devices 48. The sensors 42, 44
comm-mic~te through an appropriate communications system, i.e., an analog and/or digital
data acquisition interface 50 with the central control computer 52. One or more control
devices 48 comm~nic~te through an appropriate communications system, i.e., an analog
and/or digital control output interface 54 with the central controller 52. Alternatively, the
sensors 42, 44 and the control devices 48 communicate through a single, appropriate control
computer 52. As previously mentioned, the control computer 52 may be located on the
flotation ma~.hine, near the flotation machine, or at a remote location such as a control room.
Computer 52 has associated therewith a display 56 for displaying data and other parameters,
a keyboard 58 or other means for inputting control signals, data and the like, a memory or
recorder 60, and a modem 62 for inputting and outputting data to the control computer 52
from at least one remote location.
Still referring to FIGURE 2, the control computer 52 receives a variety of inputs
which have been catego- i~ed generally in terms of (1 ) information stored in memory when
CA 02257l58 l998-ll-27
WO 97/45203 PCT/US97/08871
the flotation machine is manufactured; (2) information programmed at the site where the
flot~tion m~rhine is to be used; (3) process parameters sensed by the process sensors 42; and
~ (4) equipment (operational) parameters sensed by the equipment sensors 44. The outputs
from the control col-lpuler may be generally categorized as (1) data stored in memory 60
S ~oci~ted with the control computer 52; (2) operational control of the flotation m~hine; and
(3) real time information provided to the operator at the monitor 56 associated with the
control computer 52. The various inputs and outputs are summarized in the following
Table.
.
TABLE
INPUTS OUTPUTS '~
1. INFORMATION ORIGINALLY STORED IN MEMORY 1. DATA STORED IN MEMORY w
OPERATIONS MAB~TENANCE INFORMATION OPERATIONS DATA
TRAINING INFORMATION PREVENTATIVE MAINTENANCE INFORMATION
e PROCESS MODEI S (OPTIONAL) FAILURE AND WEAR PREDlCTloN
PROCESS CONTROLS, GUIDELINES (OPTIONAL)
2. INFORMATION PROGRAMMED AT SITE 2. CONTROL OF OPERATIONS
OPERATING RANGES VOLUME OR MASS FLOW RATES
OUTPUT PARAMETERS DESIRED AIR FLOW RATEES
SITE SPECIFIC (E.G, ENVIRONMENTAL) DATA PULP LEVELS
PROCESS MODELS (OPTIONAL) FROTH LEVELS
PROCESS CONTROLS, CUIDELINES (OPTIONAL) pH ~
PARTICLE SIZE, CONCENTRATION, DISTRIBUTION o
3. PROCESS PARAMETERS SENSEn FLOTATION REAGENT ADDITION RATE
VOLUME AND MASS FLOWS PRESSURES
PULP LEVELS FLOW PATTERNS
FROTH LEVELS AGITATION SPEED v
FROTH MOBILITY POSITION t ORlENTAllON OF AGITATOR
FROTH VISCOSITY POSITION / ORIENTATION OF Dl~r~k~K _ ~o
FROTH COLOR POSITION I ORIENTATION OF DRAFT TUBE ~ '~
~ PARTICLE SIZE DISTRIBUTION POWER DRAW
SOLIDS CONCENTRATION
BUBBLE SIZEDISTRIBUTION 3. READOUTAT MONITOR ~
CHEMICAL COMPOSITION DIAGNOSTICS OF CONDITION OF EQUIPMENT ~1
3 0 BUBBLE DISTRIBUTION IN PULP ORDER SPARE PARTS
DENSITY MODEM/FAX FOR SPARE PARTS
j REGIONAL PRESSURES READOUT OF OPERATING PARAMETERS
- AIR FLOW RATE SCADA OR DCS
LIQUID FLOW RATE
3 5 FROTH FLOW RATE
REAGENT ADDITION RATES
VIBRATION
4. EQUIPMENT PARAMETERS SENSED
POSITION OF AGITATION MECHANISM ~-
POSITION OF CROWDER
POSITION OF DISPENSER
ROTATIONAL SPEED OF AGITATION MECHANISM
ELECTRICAL POWER DRAW OF AGITATOR
ELECTRICAL POWER DRAW OF COMPRESSOR x
CA 022~71~8 1998-11-27
wo 97/45203 PCT/US97/0887
-13-
~ Information Stored in Memory
Exarnples of information originally stored in memory include information relating to
the operation and ~ n~-ce of the flotation m~rhine and operator training inforrnation, all
of which will be readily available to an operator on display screen 56 associated with control
computer 52.
Information Programmed at Site
Examples of information programmed at the site where the flotation machine is tobe used include the operating ranges, equipment parameters, and desired feed parameters,
along with other site-specific data and environmental factors. Input into the control computer
also includes various process models, process controls, and guidelines. Tihese models and
goals may be either stored in memory or programmed at the site as appropriate.
Process and Equipment Parameters
A further important feature of the present invention is the large number of process
and equipment sensors 42, 44 which sense a variety of aspects relating to the flotation
m~hine, its operations, and its feed, tailings, and float streams. Particularly important are
sensors relating to rate of chemical addition, power consumption, aeration rate, and froth
layer thickness. Other process parameters which may be sensed include, but are not limited
to the bubble loading, volume or mass flow rates into the feed, concentrate, froth, or tailings
strearns; the air flow rates into the feed, concentrate~ froth, or tailings strearns; the density of
the feed, concPntrate~ froth, or tailings streams; the chemical or mineralogical composition of
the feed, concen1Tate, froth, or tailings streams; the pulp or froth levels; the particle size,
c~-ncentration, and distribution of solids in the feed, concentrate, froth, or tailings streams; the
bubble size, color, and distribution in the feed, concentrate, froth, or tailings streams; the pH
of the feed, concentrate, froth, or tailings streams; the rate of addition of flotation reagents,
including frothing agents, collecting agents, promoting agents, depressing agents, and the
like; the regional pressures and flow patterns within the flotation machine; acoustic ~micsion~
from the flotation m~hine; or digitized video images of the froth surface or other key parts of
the process, analyzed to determine the key characteristics of the subject being imaged.
.
CA 022~71~8 1998-11-27
WO 97/45203 PCT/US97/08871
-14-
Equipment parameters which may be sensed include but are not limited to ~ita~ionspeed, induced or forced air flow rate, position and orientation of a froth crowding device,
position and orientation of a draft tube, position and orientation of an agitator, position and
ori~nt~tion of a disperser, power draw of agitator motor; and power draw of other devices
such as a compressor.
It will be appreciated that it is often difficult to sense and communicate certain
parameters in real time within flotation machines. Thus, a variety of technologies including
ultrasonic absorption and reflection, laser-heated cavity spectroscopy, laser-induced
breakdown spectroscopy (LIBS), laser-induced mass spectroscopy (LIMS), X-ray
nuorescence spectroscopy, neutron activation spectroscopy, pressure measurement,microwave or millimeter wave radar reflectance or absorption, and other optical and acoustic
methods may be utilized in the present invention. A suitable microwave sensor for sensing
moisture and other constituents in the solid and liquid phase influent and effluent streams is
described in U.S. Patent No. 5,455,516, all of the conte,lls of which are incorporated herein
by reference. An example of a suitable apparatus for sensing using LIBS is disclosed in U.S.
Patent No. 5,379,103, all of the co~,~e~ of which are incorporated herein by reference. An
example of a suitable apparatus for sensing LIMS is the LASMA Laser Mass Analyzer
available from Advanced Power Technologies, Inc. of Washington, D.C. A preferredembodiment employing a laser spectroscopy-based sensor is described in detail hereinafter
with reference to FIGURES 3-6.
In a suitable acoustic sensor, one or more microphones, single-axis accelerometers
or multi-axis accelerometers are positioned on or near the flotation machine. Acoustic
emiCsi~nc çm~n~ting from the m~rhine, including sub-sonic, sonic, and ultrasonic waves, are
detected either directly by accelerometers, or by microphones as they are ~ "il~d through
the air. Acoustic emissions are converted to electromagnetic signals and digitized for
p,uc9ssi~ Processing may include, but is not limited to, Fourier transformation, fast
Fourier ~ fulnlalion and wavelet transformation. The signal is known to characterize
changes in the process taking place in the flotation machine. However, a stochastic model
relating the acoustic emission signal to the process and the machine's performance is
typically too complex to be useful. Therefore, the transformed emissions signal is preferably
_
CA 022~71~8 1998-11-27
Wo 97/45203 PCT/US97/08871
used as an input to an advanced control system, as described hereinafter where it may be
used in a neural network or other heuristic mo-leling system to control the performance of the
~ rn~hinP and the flotation process.
Suitable techniques for communication among the sensors, control computer, and
other components include hard-wired electrical systems, optical systems, RF systems,
acoustic systems, video systems, and ultrasonic systems.
Data Stored in Memory
Referring more particularly to the data stored in memory, it will be appreciated that
the computerized monitoring and control system of this invention may utilize theaforem~ntioned sensors to monitor various parameters with respect to time and thereby
provide a detailed historical record of the flotation machine operation 66. This record may be
used by the control computer to model flotation machine operation, adjust models for
flotation machine operation, or generally learn how the flotation machine behaves in response
to changes in various inputs. At any time, such operating data may be retrieved from the
memory of a computer local to the flotation machine or remotely. The data may be displayed
in real time while the flotation m~chine is operating using monitor 56, or as a historical
record of some prior operating sequence. This record may also be used to provide a data log,
provide trending and preventative m~inten~nce information, predict failure, and predict
m~rlline wear 68. Pre-formatted reports may present the retrieved data to show information
such as operating hours, alarms generated, number of starts, number of trips, electrical
power used, maximum and rninimunn values for measured variables, total feed processed,
and the like. Using the operating data, the flotation equipment m~nllf~ctllrer may
recornmPnd measures to avoid down time and to optimize run time. Also, mai~ltell~ce
procedures may be suggested based on the operating log of elapsed run time and unusual
operating conditions. The operating data log thus helps to trouble shoot various operating
c~ n~itiom of the flotation equipment. This enhances the flotation equipment m~nllfacturer's
ability to solve the customer's operational problems and to keep equipment on line.
Optionally, these data 66,68 may then be used to provide alarms or emergency notification
70 when certain critical levels are reached.
CA 022~71~8 1998-ll-27
WO 97/45203 PCT/US97/08871
-16-
Control of Operations
Controller 52 preferably co,...n-~.icates through standard col.-"....~;c~tion cards
used with personal computers or workct~tion~ As such, Ethemet, RS-232, and modemcapabilities exist for the operator's use. The present invention therefore allows a given plant
S to collect flotation machine operating data through a plant-wide Ethernet or other network.
Additionally, the present invention may c~mml~nicate with other process devices not supplied
by the m~nu~ . In this way the operator uses the control and monitoring system of this
invention to gather inforrnation on a larger portion of the process.
Using a cnnnected plant network, the operator may monitor the flotation m~chine's
real time performance and historical log. Suitable software for this activity includes operator
screens for data display, and message displays for operating assistance, and may also include
an on-line operation and m~int~n~nce manual. The operator may also control and optimize
the performance of the flotation machine through the plant network. The operating
parameters as described below may also become part of an overall Supervisory Control and
Data Acquisition (SCADA) system or Distributed Control System (DCS). As is well
known, in a SCADA system or DCS, microprocessor devices convert plant measurement
and status inputs into computer data for logging and transmission to higher level processors.
The SCADA system or DCS therefore connects to many controllers and field devices to
gather information and make global decisions. Supervisory, expert controllers make strategic
decicion.s for the operation of a process unit or plant and send out set points to dedicated
controllers which will make the changes to actuators and ultimately the process as a whole.
Continl-ing to refer to FIGURE 2, a further important feature of this invention is
that in response to the one or more parameters sensed by the sensors 42, 44, the operation of
the flotation machine and thereby its ultimate efficiency can be adjusted, changed, and
preferably optimized using one or more advanced computerized control methods. Control of
the machine includes control of mechanical state and operation, and control of operating
ranges to op~ e safe as well as efficient operation. Such advanced, computerized control
methods include but are not limited to neural networks, genetic algorithms, fuzzy logic,
expert systems, statistical analysis, signal processing, pattern recognition, categorical
analysis, or a combination thereof.
CA 022~71~8 1998-ll-27
~'VO 97/45203 PCT/US97/08871
Thus, in a preferred embodiment, this invention comprises at least one of these
control methodc and other methods more advanced than conventinn~l, stabilizing control
methods, for example, the simple fee~lbarl~ or feed forward control loops of the prior art.
The response of the system is based on a series of expert rules, delel,l..l-ed initially in
S advance and continn~lly updated based upon the control system's own analysis of its
pelrollll~'ce. The control system will generate and continuously update its own "process
model" using the sensor inputs described and the above-mentioned analysis techniques. The
control system may have the ability to independently select the best analysis technique for the
current data set.
While controller 52 may operate using any one or more of a plurality of advancedcomputerized control methods, it is also contemplated that these methods may be combined
with one or more of the prior art methods, including feed forward or feedback control loops.
Feed forward is where process and machine measurements (or calculated, inferred, modeled
variables normally considered ahead of the machine in the process) are used in the controller
l S 52 to effectively control the operation of the flotation m~chine. Feed forward schemes
inherently acknowledge that the conditions and state of the feed material to the flotation
m~c.hine change over time and that by sensing or calculating these changes before they enter
the flotation machine, control schemes can be more effective than otherwise might be
possible. Feedback is where measurements and calculated values that indicate process
performance and m~.hine state are used by controller 52 and the control scheme cont~ined
therein to stabilize the performance and to optimize performance as feed conditions changes
and m~ehine pe~ro~ ance changes in reference to set points and optimization objectives.
Process and m~chine models are embedded in controller 52, as are methods to
evaluate the models to determine the present and future optimum operating conditions for the
m~chine Optimum conditions are specified by flexible, objective functions that are entered
into the controller 52 by the operators or plant control system that is dealing with plant-wide
control and op~ ion The models contained therein are adaptive in that their form or
mathematical representation, as well as the parameters associated with any given model, can
change as required. These models include, but are not limited to first principles and
phenom~nologic~l models, as well as all classes of empirical models that include neural
,,
CA 022~71~8 1998-ll-27
WO 97/45203 PCT/US97/08871
-18-
network ~ep~ ;onc and other state space approaches. Optimi7~ion is accomplished by
conlbilli.lg the cont~ined knowledge of the process and m~rhine through these models with
expert system rules about the same. These rules embody operational facts and heuristic
knowledge about the flotation maçhine and the process streams being processed. The rule
S system can embody both crisp and fuzzy represent~tir)nc and combine all feed forward,
feedkac~, and model repres.ontationc of the m~chine and process to mai~lt~in stable, safe, and
also optimal operation, including the machine and the process. Determin~tion of the
op~ lulll operating states includes evaluating the model repres~."~lion of the machine and
process. This is done by combination of the expert system rules and models in conjunction
with the objective functions. Genetic algorithms and other optimization methods are used to
evaluate the models to determine the best possible operating conditions at any point in time.
These methods are combined in such a way that the combined control approach changes and
learns over time and adapts to improve performance with regard to the machine and the
process performance.
A detailed description of a suitable system employing an internal process model as
described herein for use in connection with the present invention is disclosed in U.S.
Application Serial No. 60t037,355, filed February 21, 1997, assigned to the assignee hereof,
all of the co~ which are incorporated herein by reference.
As discucsed above, the adaptive control system of this invention uses one or a
combination of internal andtor external m~chine andtor process variables to characterize or
control the performance of the flotation machine, in terms of the desired process outputs.
Pler~.~bly, the control system continually updates its knowledge of the process, so that its
control performance improves over time.
One of the important calculated values included in this process is the economic
pelro--.. al~ce of the flotation ma~hine. Economic performance includes base machine
operating costs, including power usage and chemical additive usage, the normalized
performance cost dealing with throughput rates and the quality of the products produced,
both in absolute terms and terms normalized for feed conditions, and the economic value of
the products produced.
CA 022~7158 1998-11-27
W 097/45203 PCTrUS97/08871
-19-
Still referring to FIGURE 2, in response to the one or more parameters sensed bythe sensors 42 and 44, the advanced control system of the microprocessor may actuate one or
more process and/or equipment control devices 48 to control operations. The operational
outputs from the central controller 52 may be processed though a control output interface S4.
S In some cases, the control devices will be ~ctu~ted if certain sensed parameters are outside
the normal or pres~lected flotation machine operating range. This operating range may be
programmed into the control system either prior to or during operation. Examples of
operational parameters which may be adjusted include but are not limited to volume or mass
flow rates into the feed, concentrate, froth, or tailings streams; the induced or forced air flow
rates into the feed, c~ n~entrate, froth, or tailings streams; the pulp or froth levels; the particle
size, concentration, and distribution of solids in the feed, concentrate, froth, or tailings
streams; the bubble size, bubble volume and bubble distribution in the feed, concentrate,
froth, or tailings streams; the pH of the feed, concentrate, froth, or tailings streams; the rate
of addition of flotation reagents, including frothing agents, collecting agents, promoting
agents, depressing agents and the like; the regional pressures and flow patterns within the
flot~tic n m~chine; the agitation speed in the machine; the position and orientation of a froth
crowding device; the position and orientation of a draft tube; the pOsition and orientation of
an agitator; the position and orientation of a disperser; the power draw of agitator motor; and
the power draw of other devices, such a compressor. The foregoing operational controls and
examples of actual control devices which will provide such operational control will be
described in more detail below.
Readout at Monitor
Referring still to FIGURE 2, other outputs include the real time status of various
parameters at the flotation m~chine. Thus, the operator may use the computerized control
and nlo~ u,ing system of the present invention to diagnose the present condition of the
e~ui~",~"l, order spare parts (a modem/fax 66 may be included for spare parts ordering), or
obtain a read-out as part of a SCADA system or DCS as described above.
A particularly preferred embodiment of the present invention employs an im~ing
system COIll~,lialllg video cameras or the like 46 producing images which are converted to
CA 022~71~8 1998-11-27
WO 9714s203 PCT/US97/08871
-20-
data usable by the process models of the present invention. Flotation plant operators
cu-lw~lly use visual observation of the color and consistency of flotation froths to estim~te the
perforrnance of a circuit. Automation of the visual analysis of froth compositions would
provide greatly enh~rlced process control. A description of a video sensor system for use in
mineral plOCÇ~ g operations is described in by J.M. Oestreich, et al., Minerals Fngineering,
Vol. 8, Nos. 1 -2, pp. 31 -3 9, 1 995, incorporated herein by reference. The color sensor system
described therein comprises a color video camera, a light source, a video-capture board, a
computer, and a computer program that compares measured color vector angles to apreviously stored calibration curve. Several cameras may be c~nnected to a single color
sensor computer or a single camera may simultaneously observe several locations using a
network of fiber-optic cables.
This preferred embodiment of the present invention may further comprise an
advanced control system employing both pattern analysis by neural networks, as well as
statistics and color vector analysis. As described by Oestreich, et al., above, gray level
dependence matrix methods are used to extract statistical features form digitized images of
froths. These statistical features constitute a compact set of the essential data contained in
the original image, which can then be related to the metallurgical parameters of the flotation
process by means of neural nets. Either supervised neural nets, such as learning vector
~u~ .on systems, unsupervised nets, such as self-organized mappings, or a self-
org~ni~inf~ neural net which can map high-dimensional input vectors to lower-dimensional
maps in a topological, order-preserving manner are used. Topological maps have the
advantage that they can be used to track the pe,~"na"ce of flotation processes on a
continllous basis, as opposed to the discrete classification by other classification paradigms.
For example, when considering a process system consisting of a bank of flotation cells, the
process could be monitored by means of a characteristic profile on a two-dimensional feature
map, which would enable the early detection of deviation from optimal conditions by
intelligent automation systems through comparison of the actual profile of the system with an
ideal or optimal profile.
In addition to color, both viscosity and mobility of froths may be recorded and
analyzed by visual means. Thus, in a further embodiment of this invention, a series of
CA 022~71~8 1998-ll-27
WO 97/45203 PCT/US97/08871
mod~ s are used to monitor di~rent features with a high degree of accuracy. Thus a
m~hin~ vision system based on the i~ ")rel~lion of visual features of froth structure has a
modular structure, in which one module will distinguish between froths based on differences
in morphology, a next module will base the distinction on froth mobility, another will extract
cl~omalic in~ol.l,aliol-, another average bubble size, and so on.
Referring now to FIGURE 3, a preferred embodiment of the present invention is
shown wherein the intelligent control system shown generally at 110 includes one or more
laser-induced breakdown spectroscopy sensors (LlBS sensors) and/or laser-induced mass
spectroscopy sensors (LIMS sensors). Ln3S and LIMS sensors are particularly useful in the
deterrnination of elementa5 composition in siJM~ that is, without the need for removal of a
sample for analysis at a separate location. This represents a significant advance over the
prior art, for example analysis of composition by X-ray analyzers. X-ray analyzers have in
fact been used to determine concentration of certain elements in flotation flow streams, but
require removal of a sample and analysis at a separate location. Each analysis generally
requires at least fifteen to twenty minutes. Furtherrnore, such analyses must necec.c~rily be
discrete measurements, and thus cannot provide on-going (that is, continual), real-time,
composition detern in~tio~C~
In contrast, the control system 1 10 according to the present invention allows fast,
discrete or continuous, real-time analysis. The general configuration of the intelligent control
system according to the present invention 110 comprises the control computer 112 described
in detail above, receiving data from an LIBS sensor 114. An LI13S-type sensor suitable for
use with the present invention is described in aforementioned U.S. Patent No. 5,379,103 to
Zigler. Such sensors are capable of measuring the percent concentration of one or more
elementc in a mixture. Controller 112 actuates at least one control device 116 in response to
the data received from the I,I:BS sensor 114 and an internal process model as described in
detail above. Control device 116 affects an operational parameter of the processing system
118 COI~IA~ g a multi-component mixture 120.
LIBS sensors are particularly suited for del~.,nilling elemental composition in
~centi~lly dry or dewatered solids or froths. Thus, while the particular embodiments
described herein are directed to a froth flotation m~clline7 other processing systems using
CA 022~71~8 1998-11-27
WO 97/45203 PCT/US97/08871
-22-
LIBS sensors in association with a controller to monitor composition are within the scope of
the present invention. Such processing systems are those which have sarnple streams which
do not need to be dry or dewatered, including, but not limited to, thickeners, filters,
centrifuges, analysis of the molten metal or slag streams of smelting furnaces, chemical
process solutions, and the like.
A particularly plef~,led embodiment using LIBS sensors in conjunction with the
intelligent control system according to the present invention is shown generally at 210 in
FIGURE 4. This embodiment exemplifies analysis and control of samples which do not
require dewatering. Crushed ore 212 for separation by at least one froth flotation m~rlline
214 is moved along belt 216 to the grinding apparatus 218. After grinding, the ore is
conditioned or stored in conditioning tank or feed box 222. Reagents may be added to the
grinding a,~ us 218 and/or to the conditioning tank/feed box 222 via reagent addition
system 224. The ground, conditioned material is then subjected to froth flotation in at least
one flotation mQr.lline 214. LIBS sensor 226 analyzes the composition of one or more
constituents of the crushed ore 212, and communicates these data to the intelligent controller
228. Preferably, the controller uses these data as input to a computer program which uses
neural network and pattern analysis to characterize the sample and estim~te its composition
in terms of chemical compounds or minerals contained. In response to this analysis and an
internal process model, the controller may then send signals to the grinding apparatus 218, to
the reagent addition system 224 to make adjustments to the rate of reagent addition, or it may
actuate at least one control device 230 affecting the operational parameters of the froth
flotation machine. Such operational parameters include, but are not limited to, the impeller
speed, the aeration rate, the froth wash, flow rate, the various levels of each phase, the feed
rate, and the like.
Referring to FIGURE 5, a second preferred embodiment using LIBS sensors in
conjunction with the intelligent control system of the present invention is shown generally at
231. In this embodiment, the LIBS sensor performs an analysis of a few key elements in dry
or dewatered samples, for example the concentration of copper, molybdenum, iron, silica,
and m~ne~ium in copper flotation concentrates. Thus, LIBS sensor 232 is positioned close
to the froth overflow 234 of froth flotation machine 218. The LIBS sensor therefore
CA 022~71~8 1998-11-27
WO 97/45203 PCT/US97108871
-23 -
incorporates r~ .ed optics to allow operation on or near process streams, providing
tolerance for vibration, dust, and moisture. The sensor 232 may also comprise one more
me~ nicm~ for movement of the device, by translation, rotation, or random dithering, so that
succes~ive analyses are taken from Ji~,ellt parts of the sample stream. Data from sensor
232 are co.~,n.,.. icated to control computer 228, which may actuate one or more control
devices as described above.
The above-described embodiments are directed to analysis of ~s~çnti~lly dry or
dewatered samples. Such embodiments are particularly useful in that the analysis is fast, and
provides real-time data with respect to a process flow. In a third preferred embodiment of
the present ihvention using LIBS sensors, wet samples are dewatered and analyzed to
provide data for the intelligent control system. This embodiment still provides fast, real-time
analysis. Referring to FIGURE 6A-B, LIBS sensor 312 senses a sample stream 314 from
conditioner tank/feed box 316. LIBS sensor 318 senses a sample stream 320 from tailings
322 from froth flotation machine 324. Data from each sensor are communicated to the
control computer 326, which in turn affects the operational parameters of the froth flotation
system by actuating various control devices as described above. Because the sample streams
320, 322 contain moisture, each must be dewatered prior to analysis by the LIBS sensor.
Accordingly, each sample stream is first passed through the system 340 as shown in
FIGURE 6B.
The system 340 comprises the sample feed 320, 322 (usually in the form of a
slurry), a slurry head tank 342, and a continuous, vacuum belt-press filter 344. Preferred
vacuum belt-press filters are available from Eimco, Salt Lake City, Utah. Both slurry head
tank 342 and vacuum belt-press filter 344 must be appropriately sized to provide the
required level of dewatering prior to activation for analysis. Slurry 320, 322 enters the head
tank 342, and is discharged evenly onto the filter belt 346, where it is dewatered prior to
analysis and then discharged. LIBS sensors 321, 318 analyze the dewatered sample 348.
This system is particularly advantageous in that it allows analysis of materials of unknown
moisture content, without requiring that samples be completely dried. The materials may be
analyzed without the necessity of transport to a remote site.
--
CA 022~71~8 1998-11-27
WO 97/45203 PCT/US97/08871
-24-
ln still another preferred embodiment, other dewatering devices may also be usedas applup-,ale. For example, the device 350 shown in FIG. 6C functions by passing a
sampling medium 352, for example, a moving belt or a rotating disk, through the slurry 354
being analyzed, to capture by adhesion a thin layer 3 56 of the solids in the slurry. Said thin
layer 356 may then be dried by a moving ah~lleanl 358, prior to L~BS analysis, and removed
from the sampling medium 352 by a water spray or a scraper 360 after analysis.
While the present invention has been described in conjunction with froth flotation
machines, it will be appreciated that many of the sensing, monitoring and control techniques
and instrum~nS~ion may be used in connection with any processing system for a
multicomponent mixture.
While preferred embodiments have been shown and described, various
modifications and substitutions may be made thereto without departing from the spirit and
scope of the invention. Accordingly, it is to be understood that the present invention has been
described by way of illustrations and not limitation.
What is claimed is:
. .
