V5.1 Full Manual Feb 2003 English
V5.1 Full Manual Feb 2003 English
V5.1 Full Manual Feb 2003 English
mineral
processing
simulator
Version 5.1
November 2001
Revised February 2003
Isles Road
Indooroopilly Qld
AUSTRALIA 4068
Telephone 07 3365 5842
Facsimile 07 3365 5900
Email JKTech@jktech.com.au
Internet www.jktech.com.au
JKSimMet is a powerful tool for analysis and simulation of
mineral processing plant data. As the program developers do
not control data collection, analysis or interpretation, it is the
sole responsibility of the JKSimMet user to verify that input
data are accurate, and that both process unit operation conditions
and stream outputs are reasonable.
CONTENTS
Page No
ACKNOWLEDGMENTS iv
1. OVERVIEW
1.1 About JKSimMet 1-2
1.2 Equipment Requirements 1-4
1.3 Cautionary Tales 1-5
1.4 Program Structure 1-6
1.5 JKSimMet Support 1-7
2. INSTALLING JKSimMet
2.1 Contents of the Package 2-2
2.2 Computer Hardware/Software 2-3
2.3 JKSimMet V5 Installation 2-4
2.4 Compatability Between V4 and V5 2-5
2.5 What Is New in Version 5.0 2-6
2.6 What Is New in Version 5.1 2-8
3. LEARNING JKSimMet
3.1 How JKSimMet Works 3-2
3.2 The Mouse 3-6
3.3 The JKSimMet Display 3-7
3.4 JKSimMet Startup 3-8
3.5 Working with an Existing Project 3-9
3.5.1 Selection of a Flowsheet 3-10
3.5.2 Simulation 3-11
3.5.3 Displaying the Simulation Results 3-16
3.5.4 Printing the Simulation Results 3-20
3.5.5 Summarising the Results - Overview 3-22
3.5.6 Summarising the Results - Report 3-23
3.5.7 Exporting Data from JKSimMet 3-25
3.58 Finishing a JKSimMet Session 3.25
3.6 Making Changes to an Existing Flowsheet 3-27
3.6.1 Selecting the Flowsheet to Use 3-27
3.6.2 Altering Operating Conditions 3-29
3.6.3 Saving the Session 3-33
3.6.4 Graphing Your Results 3-35
3.7 Creating a New Project 3-38
3.7.1 Starting a New Project 3-38
3.7.2 Define Flowsheet Name 3-40
3.7.3 Drawing a New Flowsheet 3-41
3.7.4 Create Connecting Streams 3-44
3.7.5 Adding a Circuit Feed Stream 3-47
4 USING JKSimMet
4.1 JKSimMet Description 4-2
4.1.1 JKSimMet Simulation Technique 4-3
4.1.2 JKSimMet Capabilities 4-3
4.1.3 JKSimMet Constraints 4-4
4.1.4 JKSimMet Expandability 4-5
4.2 Definition of Terms used in JKSimMet 4-6
4.3 The JKSimMet Cursor 4-7
4.4 The JKSimMet Menus and Toolbars 4-8
4.4.1 The Main JKSimMet Menu 4-9
4.4.2 The Functions Toolbar 4-9
4.4.3 The JKSimMet Tools Toolbar 4-11
4.5 JKSimMet Windows 4-12
4.5.1 The Session Window 4-12
4.5.2 The Project View Window 4-14
4.5.3 Equipment Data Windows 4-15
4.5.4 Port Data Windows 4-16
4.6 Building and Manipulating a Flowsheet 4-17
4.6.1 Loading a Project 4-17
4.6.2 Defining the Project Name 4-18
4.6.3 Defining the Flowsheet Name 4-19
4.6.4 Building the Flowsheet–Equipment Units 4-20
4.6.5 Building the Flowsheet–Connecting Ports 4-23
4.6.6 Flowsheet Related Problems 4-25
4.7 Editing the Flowsheet Data 4-26
4.7.1 The Equipment Data Window 4-26
4.7.2 Editing the Equipment Data 4-29
4.7.3 The Port Data Window 4-32
5 MODEL FITTING
5.1 Introduction to Model Fitting 5-2
5.2 Data Collection 5-3
5.3 Background 5-7
5.4 How the Model Fitting Program Works 5-8
5.5 A Simple Example 5-10
5.6 Learning Fitting 5-13
5.6.1 Preparation for Model Fitting 5-13
5.6.2 Start Model Fitting 5-14
5.6.3 Selecting Data 5-15
5.6.4 Setting up the Parameters 5-19
5.6.5 Master/Slave Fitting 5-21
5.6.6 Fit the Model Parameters 5-22
5.7 Checking the Fit 5-24
5.8 Presentation of Model Fitting Results 5-25
5.9 Problems Related to Model Fitting
and Possible Solutions 5-27
5.10 References 5-30
6 MASS BALANCING
6.1 Introduction to Mass Balancing 6-2
6.2 Data Collection 6-3
6.3 Background 6-4
6.4 How the Mass Balancing Program Works 6-7
6.5 A Simple Example 6-9
APPENDICES
A1 Introduction A-2
A2 Hydrocyclone (Model 200, 201) A-7
A3 Single Deck Screen (Model 230) A-21
A4 Efficiency Curve Models (210, 610, 211, 611, 203)
(General Classifier Models) A-31
A5 Efficiency Curve Variable d50c (Model 251) A-37
A6 Crusher (Model 400) A-41
A7 Rod Mill (Model 410) A-59
A8 Perfect Mixing Ball Mill (Model 420) A-69
A9 Autogenous Mill Model (Model 430)
and Semi –Autogenous Mill Model (Model 431) A-81
A10 Size Converter Model (Model 490) A-101
A11 Variable Rates SAG Model (Model 435) A-103
A12 High Pressure Grinding Rolls (Model 402) A-123
A13 Simple Degradation (Model 480) A-141
A14 Splitters (Models 810, 811, 812, 870) A-145
B ERROR MESSAGES
C JK BREAKAGE TESTING
ACKNOWLEDGEMENTS
More than twenty-five years of development has gone into the
simulation models used in JKSimMet. This represents a huge
contribution by the students and staff of the Julius Kruttschnitt
Mineral Research Centre (JKMRC). There is not sufficient space
available to acknowledge all the contributors separately, and only a
few outstanding contributions are mentioned.
The founding Director of the JKMRC, Professor Alban Lynch, and
his co-worker, Dr T C Rao, developed the first practical models of
grinding and classification, and successfully applied them at Mount
Isa Mines. Professor Lynch and his successors Dr Don McKee and
Dr Tim Napier-Munn have presided over subsequent
developments.
Dr Bill Whiten is responsible for the generalized model structure,
many of the models, and the general purpose data-fitting routines.
The simulator structure has gone through several software
generations and hardware implementations. The original engine
was programmed by Dr Alex Kavetsky, who has also contributed a
great number of the models. The major contributors to the DOS
simulator are principally Mr David Wiseman, and also Dr Fred
Hess and Dr Thomas Kleine.
The original documentation was developed by the Centre for
Information Technology Research at the University of Queensland.
The testing and debugging of JKSimMet has mostly been done by
JKTech, headed by Dr Rob Morrison and assisted by Mr Chris
Bailey, Mr Dennis Noreen and Mr Philip Baguley.
Major thanks are due to the many sponsors who have contributed to
the AMIRA projects which have resulted in the development of
JKSimMet.
Special thanks are also due to the organizations listed below which
purchased pre-release copies and have helped by testing the
software in an industrial environment:
• ZC Mines Limited • Renison Limited
• Bougainville Copper Limited • Billiton Research B.V.
• Western Australian School of Mines (Kalgoorlie)
Version 4
Six years of JKSimMet marketing have lead to the licensing of more
than 150 sites world wide. Meanwhile, application and model
development continue at the JKMRC. The development of
Version 4 and revision of the manual has been the result of
contributions from the all the JKTech team, in particular Michal
Andrusiewicz and Phil Baguley.
Version 5.0
Version 5 is a complete rebuilding “from the ground up” of the
JKSimMet interface to bring it into Windows 95/98.
The major conceptual changes to the interface are due to Stephen
Treloar-Bradford. Detailed implementation has been by Phillip
Baguley and Phil Beak. The DLL engines were programmed by
Phillip Baguley and Michal Andrusiewicz.
The Help files were developed by Andrew Schroder.
Cathy Evans has developed the V5 documentation.
Ricardo Pascual developed the V4 to V5 conversion program.
Rob Morrison provided overall project leadership.
Version 5 provides a platform for future mineral process modelling
at the JKMRC.
Overall, the development of JKSimMet V5 represents a major
achievement for the development team and a major investment in
the future for JKTech
Special thanks are also due to the V5 beta testers in industry.
Version 5.1
Version 5.1 is an upgrade of Version 5.0 which operates in the
Windows 2000 environment.
Several extra models have been added to the extensive model library
and many operability improvements have been made. A series of
bug fixes is also included.
The evolution of V5.1 has been accomplished as a joint project
between the JKTech software group and the JKTech consulting
group.
Most of the changes have come from suggestions by many of the
current users, assembled and tested by the JKTech consultants and
coded by the JKTech software group.
(Blank Page)
CHAPTER 1
OVERVIEW
1. OVERVIEW
Limitations Provided that the data used in the process models are relevant to
the ore being studied, JKSimMet can be used to generate detailed
design information. Until experience is gained in detailed design
studies using JKSimMet, it is recommended that design tasks be
carried out in consultation with JKTech.
Backup Work As you input each section of data (say a flowsheet or a data set) you
should save your work to the hard disk. Usually you will want to
overwrite your earlier version. If you do this regularly, then when,
not if, there is a power failure or other mishap, your work up to the
last save will be waiting for you on the hard disk; it will not have
been lost forever.
Backup Work You should use the backup facilities within JKSimMet to backup
Files simulator work sessions to a server or other archival storage such
as a Zip Disk.
• Main Program
• Supporting DLL’s
• Program Database
• Project Databases
• user manual
• model documentation
• context sensitive Help files.
Installation and JKTech can provide assistance to install the system and can also
Training provide on-site training to match particular user needs.
Email Help JKSimMet project files can be sent electronically to JKTech via the
Internet for assistance. Send files to jktech@jktech.com.au
Updates Updates and bug fixes will be supplied for one year from date of
installation/supply and are available under a maintenance contract
thereafter.
CHAPTER 2
INSTALLING JKSimMet
2. INSTALLING JKSimMet
Notes for Note 1: Your computer must be using NT4 with service pack 5 or
Windows NT or later or 2000.
2000 installation Note 2: As JKSimMet V5 requires several device drivers, you must
have full administrator privileges to install or uninstall JKSimMet.
If you choose not If you choose not to use the default path (/Program Files/JKSimMet
to use the default V5.1) for installation, you will need to modify the UnZip path for
path for the JKSimMet V5.1.exe file.
installation
Modify this
line to your
install path
In addition, the file structure has changed so that .JKSM5 files are
considerably smaller and no longer grow with use. Compaction is
no longer required. This change in file structure has resulted in
much faster loading and saving.
Many of the user settings which were “forgotten” on file save and
load are now “remembered”. For example, the graph colours, the
lock status and % passing size are now stored with the file, as are
the default settings for data and error displays in port windows.
Almost all of the reported bugs have been fixed and as many as
possible of the feature improvements requested by users have been
implemented.
CHAPTER 3
LEARNING JKSimMet
3. LEARNING JKSimMet
Learning Learning JKSimMet is designed primarily as a tutorial exercise. It
JKSimMet is anticipated that the first time user of JKSimMet might spend two
to three hours working through this chapter step by step. In this
way the user will gain sufficient confidence and knowledge to
begin using the system in earnest.
Given the nature and design of JKSimMet, the user will very
quickly be able to learn the basic operating techniques. It is
assumed that the user already understands the techniques of mineral
processing simulation and also has some appreciation of the
standard features of the MS Windows 95/98/ME/NT/2000
interface.
Once a model has been built the engineer can alter the design and
change the parameters as he sees fit until he arrives at a satisfactory
design or an optimum operating condition for an existing plant.
Specifying Once the flowsheet has been drawn the engineer must provide data
Flowsheet Data for each process unit and also provide raw data in the form of flows
and size distributions for the streams in the circuit. This is done by
stepping through the process units and the streams one-by-one,
adding circuit data and building up an annotated description of the
modelled processing circuit on the screen. The unit data for the
process equipment may come from previous experience, from a
design database or they may be derived from plant data. The
stream data can be entered in one of three size distribution formats,
depending on the preferences of the user. The engineer can review
or correct the data at any time after entering the data.
Flowsheet Once the flowsheet has been specified and the required unit and
Simulation stream data have been entered, the simulation can be run. The
results of the simulation are stored and can be displayed on the
screen or printed as required. The following options are available
for examining the results:
After analysis of the results, you can alter the flowsheet, adjust the
equipment parameters or port data and repeat the simulation
process until you obtain a satisfactory result.
Flowsheet A new capability in V5 is that a subset of the flowsheet may be
Selection selected for simulation, mass balancing or model fitting.
The Mouse The standard two-button mouse is used as the pointing device in
JKSimMet. In this manual we refer to "left-click" and "right-click"
which simply means to press the left- or right-hand button on the
mouse.
The manual assumes that you are familiar with common mouse
techniques such as double-clicking and click-and-drag.
In data windows, the position of the active data cell (i.e. the cell
where anything that you type will appear) is indicated by a thick
grey border
Cursor The mouse can be used to move the cursor when working with
Movement JKSimMet. In the equipment and stream data windows the cursor
control keys (also known as the arrow keys) may also be used to
move the cursor from one data cell to the next.
Appearance As with all MS Windows programs, the preferences which the user
sets for the Windows desktop will provide colours and fonts for
many of the tools and menus within JKSimMet.
Keyboard Access Most of the functionality of JKSimMet can also be accessed from
the keyboard using standard MS Windows conventions.
Users may have as many windows open on the screen as they wish
at any one time. An XGA video card and a large monitor are
recommended for this strategy.
Many of the windows are divided into several distinct areas which
are accessed by selectable tabs. Each area is used to convey
specific types of information.
Note that most windows may be minimised for convenience.
However, some non-critical changes (eg. an equipment name
change) may require that a window be closed before other windows
are updated.
The files which define the flowsheet, process units and streams that
make up the demonstration circuit are already on your computer.
They were installed onto the hard drive during the JKSimMet
installation procedures. They can be recalled by following a few
simple steps outlined below.
Open
Project
Step 3 Move the cursor to the red book of the project which
you want to load, in this case the Learner Flowsheets
project, and left-click, hold and drag it across to the
JKSimMet desktop to load the project. Note that when
you click on a project name, its file name, complete
with directory location appears in a strip at the bottom
of the Project View window.
Step 4 Left-click on the main window to make it the active
window.
Loading an Step 1 Left-click on the text box at the bottom right of the
existing JKSimMet flowsheet window to view a drop-down list
Flowsheet of the flowsheets which have been created in the
Learner project.
Step 2 Move the cursor down the list to highlight the name of
the flowsheet which you want to use (in this case
Example Ball Mill – Cyclone simulation) and left-click
on this to bring the chosen flowsheet into view on the
main screen.
Changing the Step 3 If the flowsheet you want to work with is not
size of the completely visible in the window you can change the
flowsheet height and width of the window by placing the cursor
window on the bottom, right corner of the flowsheet window
and left-clicking and dragging the window edge until it
is the required size.
Drop-down list
of flowsheets
in this project
3.5.2 Simulation
The Example Ball Mill – Cyclone simulation flowsheet already
contains all the stream data and parameters required to simulate this
circuit. We will use the JKSimMet simulation capabilities to
predict the product stream size distributions and capacity of the
simulated circuit, but first we will find out how to look at the
equipment unit data and port data.
Key to the The demonstration circuit consists of a ball mill and a nest of four
Demonstration hydrocyclones. These equipment units are connected by streams.
Circuit The streams enter and leave the equipment units through feed and
product ports.
Note that there are also two specialised units in this circuit, these
being the Feed and the Water Feeder. The Feed unit allows new
feed material to be introduced to a circuit as dry solids or a slurry.
The Water Feeder allows the addition of water to the circuit.
Examining The data windows for each equipment unit and its associated
equipment and streams can be accessed by placing the cursor over the unit and
port data right-clicking the mouse button to view the pop-up menu.
Left-clicking on the word Equipment on the pop-up menu brings
the equipment data window into view.
Left clicking on the name of a stream port in the pop-up menu (in
the cyclone example these choices are combiner, underflow or
overflow) brings the data window of that stream port into view.
Selectable tabs Note that the port and equipment data windows use selectable tabs
to provide access to the several types of data which are available
within each window. To view the available data left-click on each
tab in turn.
Arranging If there are several windows open on the JKSimMet desktop the
windows on the user has several options to organise the windows to suit their
desktop needs.
The options available can be seen by left-clicking on the icon at
the top, left-hand corner of the window which you want to move,
close etc. Clicking on this icon brings into view a drop-down
menu which allows the user to move, minimise or close the data
window by selecting the appropriate command.
Move a window Select the word Move on the drop-down menu and then move the
mouse or use the keyboard arrow keys to move the window as
required. To stop moving the window left-click with the mouse or
press the Enter key on the keyboard.
Alternatively a window can be moved by simply left-clicking and
dragging on the window title bar to move the window to where
you want it.
Minimise/Restore To minimise a window select the word Minimise on the drop-
a window down menu or left-click on the minimise button at the top, right-
hand side of the title bar. This shrinks the window to a small title
bar at the bottom of the JKSimMet desktop area. To return the
window to its previous size and position left-click on the Restore
button at the top, right-hand side of the title bar.
Close a window Select the word Close on the drop-down menu or left-click the
Close button at the top, right-hand corner of the window or hold
the Ctrl key down and press the F4 key.
Resize a window The width and height of the flowsheet window can be adjusted by
selecting the word Size on the drop-down menu. The size of the
(Flowsheet window can then be adjusted by using the arrow keys or by left-
window only) clicking and dragging a corner of the window
Data Display The data windows contain all of the information that JKSimMet
uses to perform the simulation and also show the results of the
simulation. The port data windows list the raw and calculated
values for mass flows of water and solid and the size distribution
values while the equipment data windows show the model
parameters used for simulation together with any data that result
from the simulation (e.g. cyclone operating pressure).
Step 3 The port data window has three areas for the user to
examine. The major part of the window is the area
where the data are listed. Two selectable tabs allow the
user to view the mass flow data for water and solids
Concept: The JKSimMet user can view a variety of data in the stream data
Data Formats window by selecting the required format from the Format, Data
Type and Error drop-down sub-menus. The size distribution data
can be displayed in one of three formats - % retained at size,
cumulative % retained at size or cumulative % passing size. The
Data menu gives the user the option of displaying GSIM data
types (experimental and calculated data only) or SDs data types
(experimental data, calculated data, SDs and errors) or all data,
which as the name implies, displays all of the data types including
balanced and fitted data as separate columns.
Concept: Exp Data which the user has entered which are
Data Types (Experimental the results from sampling, sizing and assay.
Stream Data)
Changing The user can change the width of the columns in which the data
column width are presented in both the port data and equipment data windows.
To change the width of a column move the cursor to the right-
hand edge of the cell at the top of the column whose width you
wish to change. When it is positioned over the border line, the
cursor will change from the usual arrowhead to a vertical line
with arrows on each side of it; left-click and drag with this cursor
to change the column width as required and release the mouse
button when the column width is to your satisfaction.
Reports The simulator can print reports in several formats, these are:
Quick Text To print the contents of a port or equipment data window simply
Printing open the required data window and click on the Print Preview icon
on the JKSimMet toolbar. A Print Preview window will display the
data as they will be printed; clicking on the Print icon in this preview
window prints the page(s) immediately. Alternatively, the data
window contents can be printed immediately by clicking on the Print
icon in the data window.
Flowsheet Print The current flowsheet can be printed by selecting File on the
JKSimMet main menu, selecting the Print Flowsheet option and then
selecting the required option for colour or monochrome printing
from the sub-menu. Selecting Print to Clipboard sends a copy of the
screen image to the Windows clipboard from which it can be pasted
into other suitable applications such as MS Word or MS Paint.
Graph Print Quick or generic graphs can be printed via the Print icon on their
window or copied to the clipboard.
Report Printing Step 1 Click on the report icon on the tool bar.
Step 2 Click on the Print icon on the Report window to print the
default report.
For details on the Report feature refer to section 3.5.6.
Ending the Step 1 To quit from JKSimMet move the cursor to the File
JKSimMet menu on the menu bar at the top of the screen and left-
Session click to view the drop-down File menu.
Step 2 Move the cursor down the File menu to select Exit.
Step 3 A pop-up window will ask you whether you want to
leave the session. Left-click on the Yes button.
Step 4 Another pop-up window will ask you whether you want
to save the last changes to the file. In this case left-click
on the No button so that the Learner project file on the
hard drive remains unchanged for the next user.
Step 3 Open the equipment data window for the Feed using
the procedure outlined in section 3.3.6. Alternatively,
lock the flowsheet using the Lock icon on the toolbar
and double click on each piece of equipment when you
wish to view the data window for it or its ports.
The Feed is a special equipment unit which represents
the flow of new material into a circuit. The Feed
equipment data window allows us to examine the feed
stream data, both mass flow and size distribution data.
It layout is the same as that found in the port data
windows which contain stream data.
One of the powerful tools which JKSimMet provides for the user is
the ability to adjust the data for the components of the test circuit.
While it is difficult and costly to experiment with real equipment,
the JKSimMet simulator allows the engineer to experiment with a
wide range of changes and to view the predicted results of these
changes.
Understanding the power of this adjustment method is important
and this section proceeds by:
• showing you how to make changes and re-simulate
• providing exercises for you to practise
• familiarising you with some short-cuts and additional useful
techniques.
The general technique is to decide the changes you want to make,
select the component whose parameters you want to change, make
changes to the parameters, re-simulate and observe the results.
You then have the choice of making further changes, undoing the
changes and trying some other ideas or accepting the changes and
saving the file on disk as a permanent record.
The parameters are characteristics of the equipment models and
their ports which can be altered. In a real plant we can alter most
equipment parameters (with varying degrees of difficulty and with
varying degrees of expense!). A few stream parameters, such as the
mass flowrate and feed size distribution can also be varied.
Simulation allows us to vary any of the parameters which affect the
process performance such as ball mill size and ore hardness with
great ease.
Concept: Note that in the data windows some of the data values are
Changing Data displayed in blue characters and some in black.
Fields
Blue text on a white background indicates that
Blue
the user can change the displayed data. To
change the data, highlight the old value by
double-clicking on it, type in the new value
and press Enter to register the change.
Concept: Water JKSimMet allows for water addition to the feed port of an
Addition to equipment unit by connecting a Water Feeder unit. The water
Equipment addition can be specified in tonnes per hour or as the amount
Units required to achieve a given feed density or simply as that
determined from the densities of the combined feed streams (i.e. no
water added). The water addition control method is specified in
the Water Feeder equipment window using the drop-down list
labelled Model.
Cyclone Step 1 Bring the cyclone equipment window into view and
Variation change the number of parallel cyclones back to 4.
Exercises
Step 2 Change the vortex finder diameter from 0.365m to
0.390m and run the simulation. Note the cyclone
pressure drop (by looking in the Performance Data
table of the cyclone equipment window).
Step 3 Reset the vortex finder diameter to 0.365m.
Ball Mill Step 1 Bring the ball mill product port data window into view
Variation and note the product 80% passing size.
Exercises
Step 2 Bring the ball mill equipment data window into view
and then select the work index for the simulated mill and
increase the value of the index by 2.0.
Step 3 Run the simulation again and observe the increase in
ball mill product size (which is displayed in the ball mill
product port data window).
Step 4 Left click on the Simulate window to make it the active
window. Left-click on the Control tab and then left-
click on the Start Condition drop-down list and select
Experimental Data.
Step 5 Now view the circuit feed stream data by right-clicking
on the Feed unit and selecting the Equipment option.
Change the value of TPH solids for the feed, run the
circuit simulation and observe the mill product stream
80% passing size. Repeat these steps until the original
product size is achieved.
Step 2 Move the cursor to the Model drop-down list and left-
click. Move the cursor to highlight the Water Feeder –
Required % solids option and left-click to select this
option.
Once you have made changes to the test circuit data, you should
remember that the changes have only been made to the copy held in
the computer's memory. To record the changes for posterity, you
must also make sure that the files on the hard drive have been
updated. This is done by saving the test to the hard drive.
It is good practice to save your work at regular intervals while you
are making changes. This will protect your work against power
failure, computer malfunction or mistakes that you will inevitably
make from time to time.
Saving the To save the session as it is now, perform the following steps:
Session
Step 1 Left-click on the File menu on the main JKSimMet
menu bar.
Step 2 Move the cursor to highlight the Save As option and
left-click to open the Save As window.
Step 3 Type the new filename in the File name box and select
the directory in which you want to save the file. If
required, you can create a new folder for storing
JKSimMet files by clicking on the Create New Folder
button.
Step 4 Once the filename and its directory have been entered,
click on the Save button to save the file.
About this From the Example Ball Mill – Cyclone flowsheet, you will create a
Exercise graph of size distribution data by:
Step 2 Move the cursor to select the word Graph from the
drop-down menu. JKSimMet will open a window
which graphs the stream data for the feed and products
streams of the selected unit, as shown below.
Step 5 The final feature of this basic graph is that the user can
select any port attached to the unit for its data to be
plotted individually. Left-click on the Show Single
Port button at the top left corner of the graph window to
view only one data set on the graph and then select the
required port from the drop-down list of port names
(Single Port Selection list). Note that this drop-down
list of port names is inactive until the single data set
option has been selected.
Concept: Once you have set up the graph to your satisfaction you can print
Printing a the graph window. Whether the graph is printed in colour depends
Graph on whether or not the printer connected to your computer can print
in colour.
Printing a Step 1 Use the Printer Setup option in the File menu to set the
Graph orientation of the paper to landscape or portrait as
required.
Step 2 To print the graph click on the Print icon on the Quick
Graph window. This will print the graph immediately.
Step 3 Quit from JKSimMet by selecting Exit from the File
menu on the main JKSimMet menu. When asked
whether you want to save any changes to your file
respond with no in order to keep the original example
for future users.
The first step in every new project is to build the flowsheet. Then
some of the data required for the equipment and streams of the new
circuit will be entered by you, the user, and some will be copied
from an existing project. The techniques available to you for
examining the data such as graphing and printing, will also be
described in this section.
All of the projects that you create will be saved on the hard drive
and will be quite separate from the Learner Flowsheets project.
Step 1 Start JKSimMet and left-click on the Open Project icon in
the toolbar. This will bring the Project View window on
to the screen with the Saved tab active.
Step 2 Left-click on the New tab to make this the active tab.
Step 3 Click on the red Default Project icon and drag it across on
to the JKSimMet flowsheet window. This will load the
Default Project which is a blank project, for you to work
on.
Each flowsheet must be given a name so that the user can select
the required flowsheet for display. In this case we are only creating
a simple flowsheet with a single circuit, but it is still advisable to
name the flowsheet.
Define Flowsheet Step 1 Left-click on the JKSimMet flowsheet window to make
Name this the active window and then right-click on any blank
area of this window.
Step 2 On the pop-up menu which appears, move the cursor to
highlight the word Flowsheet and then move the cursor to
highlight Properties on the sub-menu which appears and
left-click to select this option. This will bring into view
the Flowsheet Definition window.
You now have a blank flowsheet window on the screen in front of you and you can
begin to draw in a new circuit. You must position the process equipment before
connecting the ports of the new circuit.
Create New Step 1 To create a new equipment unit on the flowsheet first
Equipment Units left-click on the Project View window to make it the
on the Flowsheet active window and then left-click on the tab labelled
New to make it the active tab. Note that once a project
has been loaded there is only one item in the New list
in the Project view window – Default Equipment .
Deleting a Unit An equipment unit can be deleted by placing the cursor over the
unit on the flowsheet and right-clicking to view the drop-down
menu. Select the Delete option from the menu and the unit (and
any attached streams) will be deleted from the flowsheet.
Concept: Re-using equipment data that have been created for a previous
Re-using data project is a convenient short-cut. It allows the engineer to quickly
construct similar flowsheets based on the same components.
Use Existing Step 1 To use equipment units which have been used in a
Equipment Units previous project first left-click on the Project View
on the Flowsheet window to make it the active window and then left-
click on the tab labelled Saved to make it the active tab.
This brings into view a list of the projects which have
been saved in the current directory on the hard disk. If
the project which you want to access is in another
directory, click on the Browse Directories button and
select the required directory in the Select Directory
window.
Cyclone product
(overflow) port with
no stream connected.
Cyclone combiner port
with one stream connected.
Cyclone product
(underflow) port with
no stream connected.
Step 3 Move the cursor to the product port of the unit you
wish to join.
In the example flowsheet that you are building the units should be
connected as shown in the picture below. Repeat steps 1 to 4 above
to connect all of the units as shown. Note that your streams may
follow slightly different paths, depending on the order in which you
make the connections and the relative positions of the units on the
flowsheet.
Errors in JKSimMet will not allow you to draw connecting streams which
Connecting could not exist in a real plant. For example, JKSimMet does not
Streams allow you to draw connecting streams from one combiner port to
another combiner port or from one product port to another when
you draw in the circuit diagram.
Deleting If the user makes a mistake when drawing connecting streams the
Connecting stream can be deleted as follows:
Streams
Step 1 To delete a stream place the cursor over the equipment
unit from which the stream emanates as a product and
right-click to view the pop-up menu.
Step 3 Move the cursor along the list in the Delete sub-menu to
select the port name whose stream you wish to delete.
Note that if you choose the combiner port from the
Delete sub-menu all streams connected to that port will
be deleted.
Concept: Each equipment unit has a three-stream combiner at its feed port
Unit Feed (hence the name combiner used to denote the feed port). If there is
Ports more than one stream entering the unit, the combiner port data
window displays the data for the combined feed streams.
Concept: Each equipment unit has one, two or three product streams,
Unit Product depending on what type of unit it is. On the flowsheet, each
Streams product stream which leaves a unit is denoted by a product port
(shown as a short grey line which resembles a short length of pipe
with a flange at the end). Only one product stream flows from each
product port.
Adding a New Step 1 Left-click on the Saved tab in the Project View window
Feed to a to view the Saved Equipment list. Double left click on
Flowsheet the Learner Flowsheets Template.
Step 2 Left-click on the plus sign of the Example Ball Mill-
Cyclone Simulation circuit.
Step 3 Left-click and drag the Feed icon on to the flowsheet
and place it near to the feed end of the Rod Mill.
Step 4 Left-click on the flowsheet to make this the active
window.
Step 5 Draw in a connecting stream between the Feed unit
product port and the Rod Mill unit combiner port.
Adding Water Step 1 Left-click on the New tab in the Project View window
to a Flowsheet to view the Default Equipment list. If only the Default
Equipment icon is visible, double-click on the closed
book icon to view the list.
The circuit flowsheet is now complete and should now look like
the flowsheet shown below. At this stage it is advisable to Lock the
flowsheet by clicking on the Lock button on the JKSimMet toolbar.
If required, the user can add various information such as stream
name labels or equipment unit information blocks to the flowsheet.
The techniques for annotating the flowsheet in this way will be
discussed in the next section.
Concept: Each of the units and ports on the flowsheet can be annotated with
Information an Information Block which can display data for that that item on
Blocks the flowsheet. For equipment units the information block displays
two items of data while for ports up to four items of data can be
displayed. The user can select which data items are displayed in
the information block. The information block can be placed in any
position the user chooses on the flowsheet screen.
Firstly you will place information blocks on the flowsheet for some
key ports.
Adding a Port Step 1 Make the flowsheet window the active window.
Information Block
To the Flowsheet Step 2 Left-click on the Information Block Configuration
button on the main JKSimMet toolbar. This opens the
Configure/Assign Information Blocks and Labels
window as shown below.
The user can also add an information block for each equipment unit
on the flowsheet. This allows the user to observe the effect of any
changes to the circuit on unit parameters such as cyclone operating
pressure. For the example which you are working on here, an
information block for the cyclones would be useful.
The final option for annotating the flowsheet is to add one or more
Labels. This allows the user to type in their own text in a text box
which can be formatted in a range of colours and styles.
Step 7 When you are happy with the format of your text box as
shown in the Preview, click on the Add Label button to
place the label on the flowsheet. Click and drag the
label to the required position. Note that once a label
has been placed on the flowsheet its text and format
cannot be edited.
Step 8 If you want to delete a label simply double-click on it to
remove it permanently from the flowsheet.
ROD MILL
Equipment Step 1 Place the cursor over the Rod Mill icon on the
Unit Data flowsheet and right-click to view the drop-down menu.
Entry
Step 2 Move the cursor down the drop-down menu to select
the Equipment option.
Step 3 The unit data window for the Rod Mill appears on the
screen, ready for you to enter the data listed above.
Note that there are already data in the window. These
data are typical values for a rod mill which have been
selected as default data for this unit. You will replace
these data with the values listed previously.
Note: Notice that there are more data elements than
will fit in the unit data window. As discussed
previously you can view the various groups of data by
left-clicking on the appropriate selectable tab.
Step 4 First change the name of the Rod Mill to Rod Mills
No. 21,22,23 by clicking and dragging across the
existing name to highlight it, typing in the new name
and then pressing Enter. Note that the new name
appears in the title bar of the rod mill data window as
soon as you press Enter.
Step 5 Now change the number of parallel units to three by
double-clicking on the number in the cell labelled
Parallel and typing the number 3.
Step 6 Enter the data for the simulated and original mill in the
data cells with blue text which are visible under the tab
named Scaling. Left-click on a data cell to make it the
active cell and then use the arrow keys to move the
active cell as required.
Note: It is possible to overwrite any of the values that
appear in blue.
Quick Graph The Quick Graph feature provides a quick and easy method to
check for errors or discontinuities in sizing data by plotting a graph
of cumulative percent passing or cumulative percent retained vs.
size.
Tool window The JKSimMet tool window (Mass Balance, Model Fit or Simulate
window as appropriate) provides a summary of stream data showing
the results which have been calculated by running the balance, fit or
simulate tool.
Flowsheet print The flowsheet (streams and units) can be printed via the print icon.
Selecting Print Flowsheet from the file menu allows the flowsheet
to be printed or sent to the clipboard from where it can be pasted
into MSPaint for editing or to any clipboard aware application.
Single Rod Mill Reduce the number of rod mills from three to one. Add the finer
Exercise feed cyclone parameters and scale the new feed rate, ball mill size
and cyclones to suit one rod mill.
Step 1 Make the Rod Mill equipment data window the active
window.
Change Number Step 2 Change the number of rod mills to 1 and press Enter to
of Rod Mills register your change
Step 3 Position the cursor over the Water Feeder icon which is
connected to the rod mill feed port and right-click to
activate the drop-down menu.
Step 4 Move the cursor to select the Equipment option to bring
the Water Feeder unit data window into view.
Set Rod Mill Step 5 Left-click on the Model drop-down list and move the
Feed Density cursor to select the Water Feeder – Required % Solids
option
Step 6 In the Operating Conditions area of the data window
overtype the 'Required % solids' field with the new
value of 75.
Step 7 Press Enter to register your changes.
Cyclone Feed Note that the cyclone feed and product are now MUCH finer than
Size Exercise before. This causes a problem with the simulation of the existing
circuit because the cyclone model is NOT valid for large variations
in feed size. A second set of cyclone model parameters is given
below for you to try out. As an exercise, enter the data into the
correct windows and run the simulation again. Examine the circuit
data to see how the different cyclone parameters affect the circuit
performance.
Control tab The Control tab allows the user to set the parameters for the
simulation. For the most part, the default values for the
parameters should be appropriate. However, for flowsheets with
very large flows, the convergence limit can be reduced to
increase the “accuracy” of what goes in equalling what comes
out.
Select tab This tool is a more general version of the select list used in the
Mbal module of Version 4. The standard operating condition
will be to select all equipment and streams. However, it is often
useful to work with a subset of the flowsheet. To do this, the user
defines a new select list as follows:
Step 2 Select only the equipment and streams which are part
of the circuit of interest by placing a tick in the box
next to the name of each in the select list. Ensure that
all other select boxes are empty.
For example, if you are working with a rod mill - ball mill circuit
and wish to simulate and fit the rod mill only, do the following:
Using Subsets When working with a subset of a flowsheet which does not
of a Flowsheet contain a Feed unit, you must select the stream (or streams)
which is (or are) are the feed to your chosen sub-circuit.
Renaming In the Select list the stream names appear as Stream 1, Stream 2
Streams in the etc. rather than the descriptive names which are visible in the
Select list port data windows. If you wish to give a stream a more
meaningful name in the Select list, right click on its name and
select Rename from the pop-up menu which appears. Type the
new stream name into the text box and click on OK to confirm
the change. Note that these names are only used in the Select list.
Run Simulation This is the working tab which allows the user to start and stop a
simulation. It also provides a summary of port data for those
equipment ports which have been selected for inclusion in the
simulation.
The Run Simulation tab also displays the Convergence value and
the number of iterations which the simulation algorithm has gone
through. These values are updated while the simulation is
proceeding.
Configure the To configure the summary table on the Run Simulation tab, click
Run Simulation on each data column header in turn and select the required data
data summary from the drop-down list of port data which appears. If you want to
change the % Passing Size X or the X% Passing Size values, open
the Flowsheet Properties window (using the View option on the
JKSimMet main menu) and type in the desired value. You will
need to close and then reopen the Simulate tab to apply the new
% Passing Size values.
If you want to view the updated summary data values after each
iteration as the simulation is proceeding, ensure that the Simulation
Updates box on the Control tab has a tick in it. This feature allows
the user to check the data for unrealistic values (e.g. cyclone
underflow percent solids of 90%) and to stop the simulation if
necessary. If the Simulation Updates box is not ticked, the
summary data are not displayed until the simulation is complete
Exporting the The Copy to Clipboard button which is between the Start and Stop
simulation data buttons, copies the simulation data summary to the clipboard. This
summary feature can be used to easily compare several alternative
simulations by copying the data summary and then pasting it into a
clipboard compatible spreadsheet such as MS Excel for
comparison.
The data can either be plotted simply using the JKSimMet Quick
Graph feature to choose the appearance of the graph as in
section 3.6.4 or the full graphing facilities can be used to configure
the plot to the user’s requirements. This gives you the ability to
prepare sophisticated graphs suitable for publication and
presentation.
In this section you will follow a prepared example that will guide
you through the creation of a graph. The example is set up for
simulated sizing curves for all streams in the Example Ball Mill
Cyclone flowsheet.
The steps you will follow are:
• definition of the overall format of the graph including labelling
of axes, tick marks and so on,
• definition of the data sets to be plotted and of the method for
drawing curves,
• assembly of a graph from the definitions of data and format,
• note that while annotation of the graph is not available, an
automated legend facility has been added.
• production of the final graphs on the screen and printer.
Graph The Graph Definition window allows the user to define the format
Definition of the graph and to select which data sets are plotted on the graph,
using three selectable tabs to access the data fields.
Types of Data Up to fifteen curves can be drawn on a single graph. The types
that can be of data that can be used to produce curves are:
Graphed
• Graphs of sizings of all raw and calculated data for the streams
on a flowsheet.
• Efficiency curves for the raw and calculated data for all the
classification devices on a flowsheet.
• Selected functions used in the mathematical models for all of
the equipment units in a project.
Before starting the graphing example, we suggest that you save the
Learner project under a new name, for example Graph Demo. This
will avoid corrupting the Learner Flowsheets file for future
JKSimMet learners.
Defining the For the tutorial example you will set up the graph format by
graph format following the instructions below:
Step 4 Repeat Steps 2 and 3 of this section for the X-axis label
and for the graph title.
Axes and Data Within the Axes and Data Interpretation area of the Format tab do
Interpretation the following:
Range Step 5 In the columns marked Min and Max type in the
minimum and maximum data values required for the
graph axes (i.e. the range) for both the X-axis and
Y-axis. The values are .01 and 100 for the X-axis and 0
and 100 for the Y-axis.
Scale Factor Step 6 Set the scale factor as required; 1 is the usual value and
this is used for our example.
Plotting Scale Step 7 Double-click in the Plot Style cell to view the drop-
down list and select the required axis format from the
list. In this case select Logarithmic for the X-axis and
Linear for the Y-axis.
Grid On Step 8 If gridlines are required for the X-axis or Y-axis place a
tick in the appropriate box in the Grid On column.
Number Step 9 Move the highlight to the number format column and
Formats double-click to view the available options on the
drop-down list. Select the required format from the list.
In the tutorial example use Decimal for both the X-axis
and Y-axis.
The Format tab should now look like the example below.
Having defined the graph format you are now ready to define which
data are plotted on the graph.
• the data set to be graphed (this is done by choosing the item type
and then selecting one item from a list of those available),
• the range of data to be graphed, that is the minimum and
maximum values that are to appear,
Invoking Data Step 1 Select the Port Data tab at the top of the Graph
Definition Definition window.
Item Selection Step 3 Position the cursor in column 1 of the row labelled Port
and double-click (or left-click and then press Enter) to
view the drop-down list of port names. Move the
highlight to the port data which you want to plot and
double-click to select it.
Data Type Step 6 Move the highlight to the Data row and double-click to
view the drop-down list of data options.
Graphical Note that you can have either a point or a line to represent a data
Representation set. It is not necessary to have both. When a single data type has
been chosen for plotting (e.g. Exp or Fit) both the line and point
marker represent this data. However, when the paired data types
have been chosen for plotting (e.g. Exp & Sim, Exp & Fit or Exp &
Bal) the point markers represent the experimental data and the line
represents the second item of the data pair (Fit, Sim or Bal as
appropriate). This feature is useful for comparing the calculated
data with the experimental data.
Line Type Step 7 Position the highlight over the appropriate cell in the
Line row, and double-click to view the list of available
line types.
Symbol at Points Step 8 Move to the Point row and double click to view the list
of symbols which can be used to represent the data
points. This defines the symbol that is displayed to
mark the coordinate points on each curve within the
graph.
Spline Step 10 The user can choose to use spline interpolation for the
Interpolation curve which is drawn for each data set. To use spline
interpolation left click on the spline box to place a tick
in it.
Graph Over Step 11 Move the highlight to Min and Max rows and set the
Range minimum and maximum plotting range values (on the
x-axis) for each curve as required.
Step 3 Use the Display X Axis Grid and Display X Axis Grid
buttons on the graph window to add or remove
gridlines. Similarly the legend can be added or removed
by clicking on the Display Legend button on the graph
window. Note that the position of the legend cannot be
changed.
Step 3 Return to the Graph Definition window by clicking on
the Edit Graph Definition button on the graph window.
Select the Format tab and change the label settings.
Step 4 Click on the View/Refresh Graph button to view the
adjusted graph.
This procedure can be repeated until you are satisfied with the
resulting graph.
Create a New Step 1 Left-click on the Overview Config button on the main
Overview JKSimMet toolbar to bring an overview window into
view. As you can see, the overview window opens with
the default setting which displays four columns of data
for all of the streams in the current flowsheet.
Deleting Step 4 The user can remove streams from the overview list by
Streams from simply placing the cursor anywhere in the appropriate
the Overview row and then clicking on the Delete Row icon to
delete the row. In this example, remove the last stream
from the list (Cyc Feed Water Add).
Change column Step 5 If a column is too narrow for you to read the text in it,
and window size place the cursor over the right border line in the title
cell at the top of the column and click and drag the
column border to the required width. If the Overview
window is too small to view all of the data, click and
drag the bottom, right corner of the window to change
the window size as required.
Adding Streams Step 6 If you want to add a stream to the list (for example if
to the Overview you delete a stream by mistake) click on the Insert Row
icon to add a new row to the bottom of the overview
list.
Step 7 In this new row, place the cursor on the cell in the
Equipment column and press Enter to view a drop-
down list of the equipment units in the flowsheet.
Select the equipment unit to which the required stream
is connected and press Enter to place your selection in
the cell.
Step 8 Move the cursor to the Port column and press Enter to
view a list of ports associated with the equipment unit.
Select the name of the port by which the required
stream enters or leaves the equipment unit.
Selecting Stream You will now select the stream data to be displayed in the
Data for Display overview.
Step 9 Place the highlight in the title cell at the top of the
column whose data you wish to change. Press Enter to
view a drop-down list of the types of stream data which
are available for display. Select the type of data you
want to include in the overview and press Enter to
place this choice in the table. In this example, select
TPH Solids for the first column.
Note that if you are using Mass Balance and have entered
component data you may select Components from the list of data to
be displayed in the overview. If you choose Components as the
type of data to be displayed, you must then select the component
you want displayed from your list of components. This is done by
selecting the required component from a drop-down list which
becomes available in the cell below the title cell in the Components
column. This second row of the data selection cells is blank if any
other type of data is selected for display.
Select Data Type Step 10 To select the data type place the highlight in the title
cell in the third cell down from the top of the column
whose data you wish to change. Press Enter to view a
drop-down list of the types of data which can be
displayed, including Experimental data, the various
forms of calculated data and data SDs. Select the type
of data you want to include in the overview and press
Enter to place this choice in the table. In this example,
select Sim (simulated data) for the first column.
Clearing a To clear display data from a column, place the highlight on the top
Column cell in the column to be cleared, press ENTER to view the drop-
down list and select None.
calculations. Click on Yes to make the stream the basis for the
recovery calculations. Note that this stream is now listed in bold
text in the overview table to denote that all recoveries are
calculated with respect to this stream.
Exporting You may transfer an overview to the clipboard using the Copy to
overview data Clipboard and Copy Grid to Clipboard icons on the overview
window. The Copy to Clipboard icon copies only the data cells
selected by the user to the clipboard while the Copy Grid to
Clipboard copies the title cells and all of the data cells to the
clipboard.
Create a New Step 1 Left-click on the Report button on the main JKSimMet
Report toolbar to bring the Report window into view. As you
can see there is a default set of selections made for the
report.
Selecting Data Step 4 Select whether the report will print port data only or
for the Report equipment data only or both port and equipment data
by selecting the appropriate choice on the Print What
drop-down list. In this case choose the option Both.
Step 5 Select the equipment and port items whose data you
want to be printed in this report by clicking on the box
next to the name of each to place a tick in the box. If
you place a tick in the wrong box simply click on it
again to delete the tick. Note that each equipment item
and each port can be selected individually. For our
example, select the Cyclone and Cyclone Feed Water
Add equipment data and the Primary Mill product and
Cyclone combiner, overflow and underflow port data
for inclusion in the report.
Selecting Data Step 6 Next select the type of data to be listed in the report by
Types for the placing a tick in the box next to the name of the
Report required data types in the Data types to print area of the
Report window. In this case only tick the simulated
data (Sim) box.
Selecting Port Step 8 If you have included port data in your selected items, as
data is the case here, you can choose to print the Totals data
and/or the size distribution data for the ports by placing
a tick in the appropriate boxes in the Port data to print
area of the Report window. Note that if Component
data have been entered, these can also be selected for
inclusion in the report here.
The Report window should now look like the picture shown below.
Step 11 Use the Next Page and Previous Page buttons on the
Print Preview window toolbar to view all of the pages
in the report and check that they show the required
data.
Printing the Step 12 To print the report simply click on the Print button on
report the Print Preview window toolbar. Alternatively the
report can be printed directly from the Report window
by clicking on the Print button on that window’s
toolbar.
Exporting data A useful feature of the Report Print Preview window is the ability
using Report to export data in report form from the simulator in a variety of
formats. Four buttons on the Print Preview window toolbar
provide the following data export features:
These data export options allow the user to transfer data to other
applications for preparation of presentations and reports.
3.12 Summary
By working through this section on Learning JKSimMet, you have
learned to:
• run a supplied demonstration simulation
• display and/or print the results of simulations
• change some of the simulation data
• re-simulate
• build your own flowsheet, import some of its data from a
previous circuit and input new data.
You have also learnt how to plot graphs from the simulation
results.
In this way, you have learnt all the basic techniques necessary to
use JKSimMet. Additional advanced techniques for model-fitting
and for the maintenance of your system are covered in subsequent
sections.
CHAPTER 4
JKSimMet----
REFERENCE
4. USING JKSimMet
Contents of This chapter covers all the basic operational features of JKSimMet.
this Section While Chapter 3 is a tutorial, Chapter 4 is structured as a reference
section. Section 4.1 (JKSimMet Description) contains an overview
of JKSimMet. Section 4.2 contains some important definitions of
key terms. Section 4.4 (Menus and Toolbars) describes the
operating structure and its conventions while Section 4.5 describes
the various types of windows used to display information in
JKSimMet.
The cursor which indicates your position on the screen takes several forms in
JKSimMet, depending on what operation the user is performing.
The arrowhead This shape is the usual form of the cursor for pointing in all
cursor JKSimMet data windows, graph windows etc.
The Arrowhead The arrowhead cursor with crosshairs appears when the cursor is
with crosshair positioned over an equipment unit on the flowsheet. The change in
cursor shape of the cursor indicates that the user can either move the
equipment unit by left-clicking and dragging it on the flowsheet or
can access the drop-down menu for that unit by clicking the right-
hand mouse button. Note that equipment cannot be moved if the
flowsheet is locked (see 4.4)
The spanner The arrowhead cursor changes to the spanner in hand cursor when
cursor it is positioned over a feed or product port to which a stream can be
connected. The orientation of the spanner and the word next to it
changes to guide the user during the stream connection process.
Note that if a port already has its maximum number of streams
connected, the cursor will not change to the spanner when it is
positioned over the port connection point.
The File submenu New opens the Project View window so that the user can load a
new project.
Open opens the Project View window so that the user can load a
project.
Close shuts the current Project. Opening a second (or a new)
project also closes the current project after offering an option to
Save the current project.
Save saves the flowsheet and all the data associated with the units
and streams as a data file. These data files are managed
automatically by JKSimMet.
Save As allow the user to save a copy of the current project under a
new name and/or in a new directory. The default file name
extension is .jksm5.
Print displays a Print Preview window of the active window,
allowing the user to print the active window if required.
Printer Setup allows the user to select a printer and specify the
number of copies printed etc. in the standard way.
Print Preview allows the selected window to be displayed on
screen as it would appear in the printed version. An option to copy
the printed format to the clipboard is also offered by most print
preview screens.
Print Flowsheet provides options for the user to print the
flowsheet to file or the clipboard in colour or monochrome.
Exit closes JKSimMet. The user is prompted to save the current
project if the project has not been saved recently.
As its name suggests, the Toggle Tool Bar button toggles the
JKSimMet Tool toolbar on and off (i.e. makes it visible or not).
The Run button allows the user to run Simulate or Model Fit or
Mass Balance – whichever is currently active. If none is active, the
button has no effect.
The Lock the Flowsheet button does just that, locking the
flowsheet and preventing items on the flowsheet from being
accidentally moved while trying to access data on the flowsheet.
This is particularly useful when large, detailed flowsheets are being
used as it minimises the time spent waiting for the screen to be
redrawn if the user accidentally moves one item of equipment.
Users can still change data while the flowsheet is locked. While the
flowsheet is locked, double clicking on a piece of equipment will
open its window.
The Flowsheet Size drop-down list allows the user to set the size of
the flowsheet at 1x1 or 2x2 panels, depending on the users
requirements.
The Run Model Fit button brings the Model Fit window into view.
This window allows the user to select equipment and ports from
the flowsheet to be included in the model fitting procedure. For
detailed information about the model fitting tool see Chapter 5.
The Run Simulation button brings the Simulate window into view.
This window allows the user to select equipment and ports to be
included in the simulation procedure. The simulation tool is
discussed in more detail in section 3.8.
The user has two options at this point – to create a new project or to
load an existing (i.e. previously saved) project. To do either, the
user must first bring the Project View window into view by
clicking on the Open Project icon on the toolbar.
To create a new project the user drags the Default Project on the
New tab of the Project View window onto the session window.
This loads a blank project in which the user can create one or more
flowsheets, using the equipment from the Default Equipment file
on the New tab of the Project View window or existing equipment
from project listed on the Saved tab. The procedure for creating a
flowsheet will be discussed in more detail in the following sections.
Alternatively, the user can load a previously saved project from the
list which is visible on the Saved tab of the Project View window.
This will also be covered in more detail in the following sections.
Once a blank project or an existing project has been loaded in the
session window the JKSimMet Toolbar is available, giving access
to the mass balancing, model fit and simulate tools. Note that the
toolbars can be moved and resized as required by the user.
Buttons to
print data
Model type
selected from
drop-down list
Buttons to
copy and
paste data
Buttons to
copy and
paste data.
Selectable tabs to
access the mass flow,
sizing and assay data
associated with the
port.
The layout of the port data window is the same for all ports. Note
that the name of the port shown in the window title bar is defined
by JKSimMet, based on the name of the unit to which the port is
attached and the appropriate name for the port according to its
location on the equipment unit (eg. feed, underflow or overflow for
cyclones or feed and product for ball mills). In the example above,
the data window belongs to the Underflow port of the Primary
Cyclones, so its name is Primary Cyclones Underflow.
The data contained in the port data window are discussed in detail
in Section 4.7.4 on entering data.
The user can also enter more detailed information about the project
in the Project Definition window which is accessed by clicking on
the Project Properties button. The default port selection may also
be set from this window.
When all the required changes have been made, the Flowsheet
Definition window is closed by clicking on the close button in its
title bar.
Adding equipment You may double click on an existing project to reveal its
or flowsheets from flowsheets and then double click on each flowsheet to reveal its
another project equipment. The equipment can be dragged and dropped into a new
flowsheet. You may also drag a complete flowsheet from a saved
project into the current project.
Editing an Once an equipment unit icon has been placed on the flowsheet the
equipment unit user can edit or manipulate it in several ways.
on the flowsheet
Lock To lock a unit in place on the flowsheet click on the Lock button on
the main JKSimMet Functions toolbar. This prevents accidental
movement of the equipment unit when the user is working on other
items on the flowsheet. Locking the flowsheet is particularly useful
when working with large, complex flowsheets since accidentally
moving a unit requires the flowsheet to be redrawn, a process
which can take several seconds.
Equipment Most of the options for editing an equipment unit which are
Properties available to the user are presented in a pop-up menu which appears
when the cursor is placed over the equipment unit and the right
mouse button is clicked. Note that the options listed in the menu
act on the equipment unit to which the pop-up menu is attached
(and on its associated ports). The format of the pop-up menu is the
same for all equipment units, except that the names of the ports
change according to the type of unit (e.g. a ball mill would have
only Combiner and Product ports listed, while the cyclone in the
example below has Combiner, Underflow and Overflow).
Feed and A unit can have only one feed port to which up to three input slurry
Product Ports streams and one water addition can be connected. This feed port is
called a combiner to highlight the fact that its data represent the
combined streams if two or three streams are connected to this port.
Note that there are some specialised equipment units which do not
have a feed port. These are the Feed unit which is a source of new
feed (dry solids or solids and water) to a circuit and the Water
Feeder which is a source of water additions to a circuit.
The flow of material between the equipment units is created on the flowsheet by
connecting the feed and product ports of the appropriate equipment units.
Connecting Ports To connect a product port to a feed port the user places the cursor
over the product port of the equipment unit first. When the cursor
changes to a hand grasping a spanner with the word JOIN next to it,
left click to start the connection process. The word next to the
cursor will change to FEED (or PRODUCT if you are connecting feed
to product) in black text to tell you to what type of port you need
to connect. Position the cursor over the port to which you want to
connect the stream and when the cursor changes so that the spanner
changes orientation and the word FEED (or PRODUCT) is now in
white text, left click to make the connection. A connecting stream
will be drawn on the flowsheet as soon as both ends are connected
to the correct ports.
Note that JKSimMet will not allow the user to connect a feed port to another feed port.
Similarly, it will not allow a product port to be connected to another product port.
Data Entry There is a convention that the fields in the data window which are
Conventions available for data entry have a white background. This helps the
user to see at a glance which fields can be edited. The text fields
and those which have a drop-down list for selection of a field entry
have black text on a white background. Those fields which require
a number to be entered are displayed as blue text on a white
background. These conventions apply to both the equipment and
port data windows.
Note that the exact appearance and colours of each window will
also depend on how your MS Windows desktop is set up.
Equipment data The basic layout of the equipment unit data window is the same for
window layout all types of equipment; the common interface features and those
which vary between equipment types are discussed below.
Name box
Copy and
Model drop- paste buttons
down list
Number of
parallel units
Selectable
Equipment tabs
data area
Changing the Each piece of equipment can be given a name chosen by the user.
Equipment name To rename an equipment unit left-click on the text in the Name box
to highlight it. Then type in the required name for the equipment
and press Enter to register the change.
Defining the The equipment icon on the flowsheet can represent one unit of
Number of equipment or several units operating in parallel. The user defines
Parallel Units the number of units operating in parallel by left-clicking on the
number in the Parallel Units box to highlight it, typing the required
number in the box and pressing Enter.
Selecting the Each equipment type has one or more process models associated
Equipment Model with it which JKSimMet uses in the Model Fit and Simulation
procedures. The user can select which model is used to represent
the equipment in the process by left-clicking on the drop-down list
labelled Model and highlighting the required model to select it.
Note that the contents of the data area of the equipment data
window will change according to which model is selected.
Accessing the When a model type has been selected, the contents of the data area
Equipment data of the equipment data window will change to display the
appropriate data for that type of model. There is often too much
information to be displayed in the available space so JKSimMet
uses selectable tabs in the data area to provide access to groups of
data. To view each group of data the user clicks on the selectable
tab to bring it into view.
If you want to view all of the equipment data at the same time a
useful technique is to use the Print Preview window to display the
entire contents of the equipment unit data window. To view the
Print Preview window simply click on the Print Preview button at
the top, right corner of the equipment data window. The window
can be resized and the Zoom set to 100% to make the text easier to
read. When you have finished looking at the data close the Print
Preview window.
Editing the There are two methods to enter numerical data for an equipment
Equipment data unit. One option is to type the data into the appropriate data fields
by typing data in the data area of the window. To do this, use the cursor to select
the cell (denoted by the grey border around the cell), type the new
value and press Enter to accept this value. If you make a mistake in
data entry you can revert to the previous value by pressing the Esc
key but this will only work if the Enter key has not been pressed.
Note that if you enter a value which is outside the normal range for
any data item a warning message will be displayed to tell you that
the value is outside the normal range and asking whether the user
wants to use this value or have it clipped to the maximum value of
the normal range.
values for the item and lists other information about the item which
is not relevant here.
Entering The second method available for entering data in an equipment data
Equipment data window is to copy and paste the data. The values can be copied
using copy and from the data window of another unit of the same type or from an
paste Excel spreadsheet. The data are pasted into the appropriate cells in
the data window by selecting those cells and then clicking on the
Paste Clipboard to Selected Cells button.
Out of range data Note that, as before, if you paste a value which is outside the
normal range for any data item a warning message will be
displayed. If in your view, the value is reasonable, answer Yes to
the warning message and your value will be used.
However, when you use an out of range parameter, you should
check your simulation results for reasonableness even more
carefully than usual.
Port data The layout of the port data window is the same for all ports. The
window layout only feature which varies is the number of columns in the data area
of the window.
Printing
buttons
Selectable
tabs
Port data
area
Format list Clicking on the format box brings into view a drop-
down list of the sizing data formats which are
available. These are % Retained, Cumulative %
Retained and Cumulative % Passing. The user can
select the required format by highlighting its name
on the drop-down menu and left-clicking.
Data list The Data drop-down list allows the user to select
which data types are displayed in the data area of the
port data window. The number of columns in the
data area varies depending on the data types
selected.
Data transfer The four buttons in this area of the equipment unit
data window allow the user to transfer data to and
from JKSimMet and other programs by copying and
pasting data to and from the clipboard. The buttons
perform slightly different functions as follows:
Data area The lower section of the port data window contains
the area where data such as mass flows and size
distribution data are displayed.
Selectable These three tabs provide access to the groups of data
tabs which describe the flow of material through the port.
The tabs are labelled Totals, Size Distribution and
Components.
Default Port You may set the default size format, data format and error format
Format from the Project Properties window.
Selecting the If the data which describe the material flowing through the port
Format for includes sizing data you can select the format for this data to be
sizing data displayed on the Size Distribution tab using the Format drop-down
list. Click on the Format box to view the list of options, move the
cursor to highlight the required format and click on it to select it.
The available options are % Retained, Cumulative % Retained and
Cumulative % Passing size
Selecting the The user can select the types of data which are displayed in the data
Data types area of the port data window. To do this click on the Data box,
for display move the cursor to highlight the required data group and click on it
to select it. The available choices are:
GSIM Displays two columns of data – Exp (experimental) and
one other which is either Sim (simulated) or Fit (model
fitted) or Bal (mass balanced), depending on which
JKSimMet analysis mode is currently active.
SD’s Displays the two columns as for GSIM, together with a
column for experimental data standard deviations (SDs)
and another for calculated Error.
All Data Displays all of the data types which are available in
JKSimMet – Exp, SD, Sim, Fit, Bal, and Error.
Selecting the The Error column in the data area of the port data window can
Error type display the absolute (Abs), percentage (Pct) or weighted (Wtd)
error for the simulated (Sim), fitted (Fit) or Balanced (Bal) data as
required. The user selects the required error type from the Error
drop-down list.
Accessing the With mass flow, sizing and component data there is too much
Port data information to be displayed in the available space in the data area
so JKSimMet uses selectable tabs to provide access to groups of
data. In the case of port data windows the data is grouped as
Totals, Size Distribution and Components. To view each group of
data the user clicks on the selectable tab to bring its data into view.
If you want to view all of the port data at one time a useful
technique is to use the Print Preview window to display the entire
contents of the port data window. To view the Print Preview
window simply click on the Print Preview at the top, right area of
the port data window. The window can be resized and the Zoom
set to 100% to make the text easier to read. When you have
finished examining the data close the Print Preview window.
Editing the There are two methods to enter numerical data for a port. One
Port data by option is to type the data into the appropriate data fields in the data
typing data area of the window. To do this, use the cursor to select the cell
(denoted by the grey border around the cell), type the new value
and press Enter to accept this value. If you make a mistake in data
entry you can revert to the previous value by pressing the Esc key
but this will only work if the Enter key has not been pressed.
Entering The second method available for entering data in a port data
Port data using window is to copy and paste the data. The values can be copied
copy and paste from the data window of another port or from an Excel
spreadsheet. The data are pasted into the appropriate cells in the
data window by selecting those cells and then clicking on the Paste
Clipboard to Selected Cells button.
The Totals The Totals tab contains the mass flow data for solids and water
data tab through the port. If experimental values are available for these data
they are entered in the appropriate data cells. Note that if a solids
mass flow and solids SG have been entered, only one of the other
data items is required and the remainder will be calculated by
JKSimMet. For example if you enter the TPH solids and the
% Solids, JKSimMet will calculate the TPH Water, Pulp Density
and Vol. Flow. The user must ensure that the Solids SG is correct.
The size for the % Passing x mm and the percentage for the x %
passes size data items can be set in one of two ways. If you want
the values to be applied only to this port, double-click on the label
of the item you want to change. This brings into view an Enter New
Value window in which the required value is entered.
If you want to change the size for the % Passing x mm and/or the
percentage for the x % passes size data items values for all ports,
right click on a blank area of the flowsheet to view the pop-up
menu and select Flowsheet and then Properties to bring the
Flowsheet Definition window into view. The size for the % Passing
x mm and the percentage for the x % passes size data items can be
entered in the appropriate boxes in this window. (See section 4.6.3
for more information on the Flowsheet Definition window).
The Size The Size Distribution tab contains the sizing data for the solids
Distribution flowing through the port. If experimental data are available for
data tab these data they are entered in the appropriate data cells. The first
step is to enter the sizes to which the data relate. The size values
are in millimetres. Note that if the required size distribution is
defined for the first equipment unit placed on a flowsheet, any
further unit placed on the flowsheet will automatically use these
size data in the port data windows.
There are several options for entering the size data. The first option
is to use the √2 button on the port data window to place a √2 size
series in the Size column. The first step is to zero the size data by
typing a zero in the Top Size box and pressing Enter. As the
warning message will tell you, doing this will delete all size and
sizing data. Then type the new top size in the Top Size box and
click on the √2 button. JKSimMet places a √2 size series of 30
values from the user-defined top size down to zero. To truncate the
size list simply type a zero where required in the column. The user
can also edit individual values in the list as required.
The Components If the user has defined a list of components to be used in mass
tab balancing in the Mass Balance window then the Components tab in
each port data window will be configured to accept data. In this
case the user can enter component data such as assay data for solids
flowing through the port. If experimental values are available for
these data they are entered in the appropriate data cells. Note that
the component data are only used in mass balancing in JKSimMet.
The Set SD’s Estimates of the accuracy of the experimental data can be provided
button by entering data standard deviations (SDs) for the data. To do this,
make the appropriate selectable tab active and selects SD’s in the
Data drop-down list so that the SD values can be seen in the data
area. . There are two methods for the user to enter SD values for
the data. The first method is simply to type in the required SD
values in each data cell in turn. The second method uses the Set
SD’s button to apply a selected SD model to all of the data on that
tab. In most cases users will use the two methods to enter SDs by
applying an SD model to all of the data and then fine tuning some
SDs by typing new values in.
Opening the A Feed data window can be opened in two ways. The first method
Feeder data is to place the cursor over the Feed icon on the flowsheet and right-
window click to bring the drop-down menu into view. Then move the
cursor to select the word Equipment from the menu and left-click.
The second method can be used to open the Feed data window
when the flowsheet is locked. In this case, double-clicking on the
feeder icon opens its data window.
The Feed data window has one feature in common with a standard
equipment data window – the Name box where the user can define
a name for the feeder. The remaining parts of the Feed data window
are the same as a standard port data window with Totals, Size
Distribution and Components tabs to access various groups of data.
Note that the Feed has only one port, a product port. As it is the
source of new material to be added to the circuit it does not have a
feed port.
Opening Water A water feeder data window can be opened in two ways. The first
Feeder window method is to place the cursor over the water feeder icon on the
flowsheet and right-click to bring the drop-down menu into view.
Then move the cursor to select the word Equipment from the menu
and left-click. The second method can be used to open the water
feeder data window when the flowsheet is locked. In this case,
double-clicking on the water feeder icon opens its data window.
Water Feeder The layout of the water feeder data window is shown in the picture
window layout of the window below. As can be seen, some of the elements found
on the equipment unit data windows appear here, such as the Name
box and Model drop-down list. However, the layouts of the
Operating Conditions tabs are unique to the water feeder data
window.
Data area The lower section of the water feeder data window
contains the area where data about the water
addition are entered. The contents of the section
varies between the three water addition models
which are available.
Selecting the The water feeder has three water addition models associated with it
Water Feeder which JKSimMet uses in the Mass Balance, Model Fit and
Model Simulation procedures. The user can select which model is used to
represent the water addition to the circuit by left-clicking on the
drop-down list labelled Model and highlighting the required model
to select it. Note that the contents of the data area of the equipment
data window will change according to which model is selected.
The water addition models are described below.
Feed Streams When this model is selected the water feeder does not add any
model water to the equipment unit to which it is attached. The water
content of the feed to the equipment to which the water feeder is
connected, is controlled by the water contents of its feed streams
only. There are no data to be entered by the user on the Operating
Conditions tab for this model.
Required % Solids When this model is selected JKSimMet adjusts the water addition
model to obtain the required percent solids in the equipment feed. The
user must enter the value of the required percent solids on the
Operating Conditions tab.
This model is useful in simulations, where the user can select the
required percent solids for a feed port, e.g. a cyclone feed, and
JKSimMet adjusts the cyclone feed water to cope with any changes
in cyclone feed mass flows which are caused by other changes to
the flowsheet.
Water Addition In this model the water addition is fixed at the value set by the user.
model The user must enter the value of the required water addition in the
New Water Addition box on the Operating Conditions tab.
This model is useful in mass balancing and model fitting where the
user often has measured water addition data to be incorporated in
the flowsheet. If the flowsheet data are being mass balanced the
user can also enter standard deviation (SD) values for the water
addition.
After a mass balance, the “calc” result needs to be copied manually
to “Exp” for use in simulation and model fitting.
Port data
information block
Equipment data
information block
Access the Ports Left click on the Configure/Assign Information Blocks and Labels
Information Block button on the JKSimMet toolbar to bring the Configure/Assign
Configuration tab Information Blocks and Labels window into view. Click on the tab
labelled Port to make this the active tab.
Drop-down list of
data types to be
displayed.
Selecting the port The first step in adding a port information block is to decide
data for display whether to display one type of data (e.g. Experimental data only or
Simulated data only) or to view two data types (e.g. Experimental
and Simulated) together.
Note that ore feeder and water feed data can be accessed via the
equipment table – (Section 4.8.2)
One data type The default setting is for one type of data to be displayed. This
means that the user can view the chosen data (e.g. experimental)
for up to four different data items (e.g. TPH solids, % solids, Vol.
Flowrate and % Passing size). The user can select the data type to
be displayed from the drop-down list that appears in the lower part
of the Configuration area. The available choices are Exp, SD, Sim,
Fit, Bal, Calc Bal SD and Error.
Select required
data type from
drop-down list
Add an Once you have configured the contents of the information block to
information your satisfaction, click on the Apply button. This creates an
block legend information block legend which shows the data type selected in the
information block title bar and lists the names of the data items in
the appropriate boxes.
Placing an To place the information block for a port on the flowsheet select
information block the name of the port from the list at the left side of the window and
on the flowsheet then click on the Add New Block button. The information block
will appear on the flowsheet behind the legend information block
and may be dragged to the required position. Note that the name of
the port to which the data apply is displayed in the title bar at the
top of the information block.
Deleting an To delete the information block from the flowsheet simply click on
information the Close button at the top, right corner of the information block.
block Note that when an information block is deleted it does not reappear
immediately in the list in the Configure/Assign Information Block
and Labels window. To make the deleted port appear in the list
close and then reopen the Configure/Assign Information Block and
Labels window.
Displaying Dual To view two data types for two data items in the information block
Data Types in the user must click on the box labelled Allow Dual Data Types to
information place a tick in the box. The Configuration area of the
blocks Configure/Assign Information Block and Labels window will
change to display two drop-down lists from which to choose the
data type, as shown below.
The data type and data item selection procedures are the same as
discussed previously but in the dual data type case the user must
choose two data types (one from each drop-down list) and can only
select two data items to display.
When dual data type information blocks are being used, the title bar
of the legend information block shows the names of the two data
types selected for display as shown below.
Access the Left click on the Configure/Assign Information Blocks and Labels
Equipment button on the JKSimMet toolbar to bring the Configure/Assign
Information Block Information Blocks and Labels window into view. Click on the tab
Configuration tab labelled Equipment to make this the active tab. Note that the
interfaces of the Port and Equipment tabs are similar; both have a
list of the things for which an information block can be displayed
on the left of the window and an information block configuration
area at the right side of the window.
The Equipment tab of the Configure/Assign Information Blocks
and Labels window
When an equipment
unit is selected, the
data items available
for display are listed
here
Selecting the The first step in adding an equipment unit information block is to
equipment unit select the required equipment unit from the list at the left of the
for display Configure/Assign Information Blocks and Labels window. Left-
click on the name of the unit to select it.
Selecting the Once a unit has been selected, a list of the data items for that
equipment data equipment unit type is displayed in the Configuration area of the
for display window. The user can select one to two items from this list for
display in the information block. To select a data item simply click
on its name and the data item will appear in the information block
above the list. To change the selected data items the user must use
the Clear button to remove all selected items from the information
block and to select the required ones from the list again.
Add an equipment Once you have configured the contents of the information block to
unit information your satisfaction, click on the Apply button to confirm the
block selection. Click on the Add New Block button to place the
equipment unit information block on the flowsheet. This creates an
information block which shows the names of the data items and
their values and also shows the name of the equipment unit in the
information block title bar.
Deleting an To delete the information block from the flowsheet simply click on
information the Close button at the top, right corner of the information block.
block
Water Feeder The Water Feeder equipment unit is a specialised form of the
equipment unit equipment unit. The data items which can be displayed in its
information information block include information about the experimental and
blocks calculated water additions and the SD and weighted error of the
water addition.
Access the Label Left click on the Configure/Assign Information Blocks and Labels
Configuration tab button on the JKSimMet toolbar to bring the Configure/Assign
Information Blocks and Labels window into view. Click on the
Labels tab to make this the active tab.
Select format
from these
options Preview of label
is displayed here
Enter the label To enter the text which you want to display in the flowsheet label
text double click the default text in the Label Text box to highlight it
and then type in the required text. As you type, the text appears in
the Preview area of the window. The position of the text in the box
can be adjusted by using the Enter key to add blank lines and the
space bar to add extra spaces as required.
Format the label The user can select the alignment of the text from the choices in the
Text Alignment area of the window.
The Label Properties area allows the user to wrap text in the label
by clicking in the Word Wrap On box to place a tick in the box.
Similarly the user can place a border around the label by ticking the
Label Border On box.
The user can use the Autosize function to set the height and width
of the label box automatically. Alternatively, if the Autosize box is
not selected the user can type the required dimensions of the text
box into the Height and Width boxes which are situated above the
Preview area of the tab.
Select the label The user can change the default background colour of the label by
background colour double clicking on the Label Background Colour panel of the
window. This brings up the colour palette from which the user can
select an existing colour or create a custom colour for the label.
Add the label to Once you have configured the label to your satisfaction, click on
the flowsheet the Add Label button to place the label on the flowsheet. The label
can be dragged to the required position on the flowsheet. Note that
once the label has been placed on the flowsheet its format and
contents cannot be edited or changed in any way
Labels Labels for the graph as a whole and for the X and Y axes are
specified by typing in the text for the labels and using the Font and
Font Size drop-down lists to format them as required. To type text
into a label, double-click on the existing text to highlight it and
then type the replacement text. The text can then be set to the
required typeface and size by selecting the required items from the
Font and Font Size drop-down lists.
Axes and Data The axes and data interpretation section of the Format tab defines
Interpretation the ranges and scales of the axes.
Select other
existing user-
defined formats
from this list.
The default
graph data set is
displayed here
The Graph Definition window with the Port Data tab selected.
Port Data
The various data cells in the Data Selection area of the Port Data
tab are discussed below. Each column in the Data Selection area is
used to configure the data presentation for one port. Note that the
user can select an individual port more than once in the Data
Selection area. For example, if the user wanted to present the
experimental data for a port with green dots and the simulated data
with a blue line, it would be necessary to configure this format in
two separate columns.
Port Selection The first row of the Data Selection area is labelled Port and it is in
this row that the user defines the name of the port whose data are to
be plotted on the graph. Double click on the Port cell to view a list
of the port names on the current flowsheet. Move the highlight to
select the required port and press Enter to make the selection.
Note that once a port name has been selected, JKSimMet places a
standard set of choices in the formatting cells in that column. The
user can edit these if required.
To remove a port from the selection to be graphed and clear all the
other selections in its column, double-click on the Port cell and
select None from the drop-down list of port names.
Format The Format defines which type of plot is presented for the port.
Double-clicking on the Format cell brings into view the drop-down
list of available graph plotting formats.
Data Move the highlight to the Data row and double-click to view the
drop-down list of data types which can be selected for graphing.
If the single data type options are selected (Exp, Sim, Fit or Bal),
both the line and point markers for these data represent the chosen
data type. However, when the paired data types are selected for
plotting (e.g. Exp & Sim), the data point markers represent the
experimental data and the line represents the second item of the
data pair (Fit, Sim or Bal as appropriate). This feature is useful for
comparing the calculated data with the experimental data.
Line The Line option allows the user to choose the style of line which
will be used to represent the data. The choices are accessed by
double-clicking on the Line row cell and selecting the required line
style from the drop-down list. Note that the user can choose to
have no line plotted by selecting the option None on the list.
Point JKSimMet places a marker at every data point on the graph. The
user can select the style of marker to be used for the port from a list
of point marker styles. The choices are accessed by double-clicking
on the Point row cell and selecting the required marker style from
the drop-down list. Note that the user can choose to have no point
marker plotted by selecting the option None on the list.
Colour The user can choose what colour is used to display the line and
point markers on the graph. To view the list of available colours,
double-click on the Colour cell. Move the highlight the required
colour and left-click to select it.
Spline The user can choose to use spline interpolation for the curve which
is drawn between the data markers for each port. To use spline
interpolation left-click on the spline box to place a tick in it.
X Min and X Max The user must define the minimum and maximum plotting range
values (along the X axis) for the data. These values are typed into
the appropriate cells
Equipment Data
Many of the formatting cells on the Equipment Data tab perform
the same function as in the Port Data tab. Only those cells which
perform different functions are discussed below.
The Graph Definition window with the Equipment Data tab selected.
Equipment The first row of the Data Selection area on the Equipment Data tab
is labelled Equipment and it is in this row that the user selects the
item of equipment whose data are to be plotted on the graph.
Double click on the Equipment cell to view a list of the equipment
names on the current flowsheet. Move the highlight to select the
required item and left-click to make the selection.
Note that once an item of equipment has been selected, JKSimMet
places a standard set of choices for that particular equipment type
in the formatting cells in that column. The user can edit these if
required.
To remove an equipment item from the selection to be graphed and
clear all the other selections in its column, double-click on the
Equipment cell and select None from the drop-down list of
equipment names.
Function The Function cell defines which type of data function is presented
for the equipment. Double-clicking on the Function cell brings into
view the drop-down list of available functions. The list of functions
will change according to the type of equipment which has been
selected.
Move the highlight to the required function and left-click to select
that function.
The remaining formatting cells perform the same function as those
on the Port Data tab and have been discussed in the previous pages.
The Graph window has several features which allow the user to
make changes to the appearance of the graph without returning to
the Graph Definition window and to print or copy the graph. These
are accessed via the buttons on the Graph window toolbar.
The Display X Axis Grid and the Display Y Axis Grid buttons
allow the user to add and remove gridlines from the graph.
The Display Legend button adds or removes the legend. Note that
if the legend overlaps the plot area of the graph this can be
overcome by making the Graph window wider.
The Edit button makes the Graph Definition window the active
window, allowing the user to edit the format or data definitions.
The Refresh button redraws the graph. This allows the user to
update the graph after changing data or formats.
The Copy to Clipboard button copies the graph to the clipboard
from where it can be pasted into word processing documents,
presentations etc.
The Print Graph button immediately prints the graph to the
currently selected printer. The printed graph will have the same
appearance (overall size, relative dimensions etc.) as it does in the
graph window. The size of the graph can be changed by adjusting
the Graph window as required. Note that a message window may
appear while JKSimMet spools the graph to the printer.
Selecting the Graph option from the menu brings the Quick Graph
window into view. The name of the equipment unit to which the
data relate is shown in the title bar of the Quick Graph window.
Note that, by default, the graph plots the size distribution data for
all of the ports connected to the equipment unit as a cumulative
weight percent passing size format. These settings can be changed
using the buttons on the Quick Graph window toolbar.
The Show All Ports button displays on the graph the size
distribution data for all of the ports connected to the equipment
unit.
The Display X Axis Grid and Display Y Axis Grid buttons allow
the user to add and remove gridlines on the graph. If the gridlines
are switched on, clicking on the button again removes the gridlines
from the graph.
The Sizing Format drop-down list allows the user to select the
format for plotting the size distribution data. The options are
% Passing (cumulative weight % passing), % Weight (weight %
retained) and % Retained (cumulative weight % retained).
The Single Port Selection drop-down list allows the user to select
which of the ports attached to the equipment unit has its sizing data
displayed on the graph when the Show Single Port button is
selected. The list of port names changes to reflect the type of
equipment unit selected. Note that this list is only accessible when
the Single Port button is selected; when the Show All Ports option
is selected this drop-down list is inactive (greyed out).
The Data Type drop-down list allows the user to select the type of
data to be plotted on the Quick Graph. The choices are
Experimental, Calculated, Absolute Error and Exp and Cal (both
experimental and calculated data plotted on the graph). Note that
the type of data plotted as Calculated (mass balanced, fitted or
simulated) depends on which JKSimMet tool is selected at the
time.
Viewing the port Quick Graph provides a shortcut for users to quickly access the
data window data window for any of the ports whose data are plotted on the
Quick Graph. To do this the user simply clicks on the line whose
data window he wishes to examine. This brings that port data
window into view.
Identifying lines While Quick Graph does not provide a legend, the user can find out
on the graph which port a line represents on the Show All Ports graph by
pointing at the line with the cursor. When this is done a pop-up
label displays the name of the port to which the data relate.
Identifying data On the Show Single Port graph the user can find out what the X
points on the and Y values are at any data point by pointing at the data marker
graph with the cursor. When this is done, a pop-up label displays the X
and Y values at that data marker.
The Name box is a text box where the user can type a name for the
current overview.
The New Overview button adds a new overview to the select list.
As the name implies the Insert column and Delete Column buttons
add and delete a data column in the overview table.
As their names imply, the Insert Row and Delete Row buttons add
and delete port data rows in the Overview table.
The Recovery box allows the user to set the overview table in
Recovery Mode where recovery data are presented in place of the
actual mass flow data.
The Print button prints the overview table on the currently selected
printer.
The next step in configuring the overview is to decide which data are to be displayed in
the table and in which order. Before doing this it may be necessary to make the
Equipment and Port columns wider in order to read the names of these items. If the
window is too small to view all of the data the user can adjust the window to the
required size.
Make a data The width of each column in the overview table can be adjusted by
column wider placing the cursor over the right border of the title cell for the
column and clicking and dragging the border until the column is
the required width.
Resize the The Overview window can be resized by clicking and dragging any
Overview window side or corner of the window.
The user can arrange the order in which the port names appear in the table by selecting
the required port name in each row. Firstly the user must select the equipment unit to
which the port is attached and then the name of the port itself.
Select an To select an equipment unit for display in the list double click on
equipment unit the appropriate cell in the column labelled Equipment. This brings
for display into view a drop-down list of all the equipment units on the
flowsheet. Move the cursor to highlight the required equipment
unit name and press Enter to register the change. If a row of blank
cells is required to help make the table easier to read the user can
select None from the list of equipment names. All other cells in this
row will remain blank.
Select a port name Once an equipment unit has been selected in the Equipment
for display column the user can select the required port. To do this, double-
click on the appropriate cell in the column labelled Port to bring
into view the drop-down list of ports associated with the equipment
unit. Move the cursor to highlight the required port name and press
Enter to register the change.
The default overview table may contain more or less data rows than are required. Rows
can be deleted or added as required using the Delete Row or Insert Row buttons.
Delete a row from To remove a row of port data from the list in the overview simply
the overview table click anywhere in the row and click on the Delete Row button. A
JKSimMet dialogue window will ask you to confirm that you wish
to delete this row. Click on Yes to remove the row from the
overview table. More than one row can be deleted by highlighting
two or more adjacent rows and using the Delete Row button as
described above.
Add a row to the To add a row to overview table click anywhere in a row in the table
overview table and then click on the Insert Row button. Note that the new row is
always added immediately above the cursor position and by default
this row contains data from the first port of the first unit in the
equipment unit list. You can select the information to be displayed
in the new row by clicking on the relevant cells and selecting from
the drop down lists.
Once the list of port names has been defined the next stage in configuring the overview
is to define which data are displayed in the data columns. All of the data which appear
on the Totals tab of the port data window (e.g. TPH solids, % solids) are available for
display in the overview table. If component data have been entered for a port these can
also be selected for display. As well as defining what data items are displayed the user
must also define what data type (e.g. experimental, fitted etc.) is displayed in each
column. The user can configure as many data columns as required to display the port
data.
Select a data item Each column displays the values of a selected data item. To define
for display the data item place the cursor in the title cell at the top of the data
column and double click. This brings into view the drop-down list
of all available data items.
Move the highlight to select the required item and press Enter to
confirm the selection. Note that an item can be selected in more
than one column. This allows the overview to display, for example,
one column with experimental data, one with data SDs and one
with fitted data. Selecting the option None from the drop-down list
results in all other cells in the column being blank (a feature which
can help to make large tables easier to read).
Note that the available size markers are set from the Flowsheet
Properties window.
Select a If the Component data item has been selected at the head of a
component name column the user must select the name of the component in the
for display second row of the title section for the column. To select the
component name double click on the second row cell in the column
to view a list of components available for display. (The list of
names will vary according to the component names which the user
has defined). Move the highlight to select the required component
and press Enter to make the selection.
Note that the second row cell remains blank if Component has not
been selected as the data item.
Select a data type Each data item has several data types associated with it and the user
for display can choose which of these is displayed in each column. Double
clicking on the third row cell in the column brings into view the
drop-down list of data types.
Add a data Clicking on the Insert Column button adds a new data column to
column to the the overview table. Each new column is added to the left of the
overview cursor position. The newly added column is configured with
experimental data for TPH solids and so must be configured to the
users requirements.
Remove a data A column can be deleted from the overview table by placing the
column from cursor anywhere in the column and then clicking on the Delete
the overview Column button. A JKSimMet dialogue window will you to confirm
that you want to delete this column. Click on Yes to delete the
column. More than one column can be deleted by selecting two or
more adjacent columns and using the Delete Column button as
described above.
The stream port with respect to which all of the recovery values are
calculated is indicated by its name being shown in bold text in the
table. This port is known as the recovery basis port. The default
recovery basis port is the circuit feed.
Change the The user can change the recovery basis port by placing the cursor
recovery over the name of the new recovery basis port in the overview table
reference stream and right clicking. A JKSimMet dialogue window will ask you to
confirm that the chosen port is to be the basis for the recovery
calculations. Click on Yes to confirm the change. The recovery
value sin the overview table will change to reflect the change in
recovery basis.
Printing the The flowsheet can be printed in colour or black and white to the
Flowsheet printer or copied to the clipboard. Select File from the main menu
followed by Print Flowsheet and select the desired option.
Note that the printed format allows the user to see all of the data
that are contained on all selectable tabs in the data window. The
data which are displayed on separate tabs in the window are printed
in consecutive areas of the printed data.
It is worth checking that the columns in the printed tables are wide
enough for the data values to fit. If the columns are too narrow,
close the print preview window, make the column in the data
window wider and then open the print preview window again.
Print Preview The Print Preview window which opens when the Print button is
window clicked shows how the printed form of the data will appear.
Print Preview window at 100% zoom factor with window resized to view all data
Once the print preview is satisfactory, click on the Print icon in the
Print Preview window to print the data. To remove the Print
Preview window from the screen close its window.
Printing Overview The user can configure one or more Overview tables to summarise
tables data selected by the user (see section 4.11 for details of the
Overview features). Printing the overview table follows the
standard procedure of making the overview window the active
window and then clicking on the Print button on the JKSimMet
toolbar. This brings the Print Preview window into view and
allows the user to check that the appearance of the printed
document is satisfactory. Make any adjustments required and then
click on the Print icon on the Print Preview window to print the
data.
Printing The Quick Graph feature allows user to create size distribution
Quick Graphs graphs for the feed and products of each equipment unit. These
graphs are printed by clicking on the print button on the toolbar of
the Fast Graph window. Note that the graph prints as a bitmap and
therefore text and graphics can appear with jagged edges. A less
jagged printout can be obtained by using the Copy to Clipboard
button, pasting the graph image into a word processing program
(e.g. MS Word) and then printing
To print the
graph as shown
click on this
Print button
Create a New To create a new report configuration click on the Create New
Report Report button in the Report window. This brings into view a
table which lists all of the ports and equipment items on the current
flowsheet. Note that in the default configuration none of the items
in the grid are currently selected.
Name a report To name the new report format double-click in the Name box to
highlight the default name of the report configuration and then type
in a new name for the report. Press Enter to confirm the name
change. The new name will now appear in the Report list and also
in the title bar of the report window.
Setting the sizing The user can choose the format which is used to present any size
data format distribution data in the report. To do this click on the Format drop-
down list and select the required sizing format from the list.
Selecting Data The user must select the equipment and port items whose data are
for the Report to be printed in the report. To do this the user can click on the box
next to the name of each item to place a tick in the box.
Alternatively, if all items listed in the window are to be included in
the report, click on the Select All Items button at the top of the
report window. To remove an item from the report simply click on
the items box again to delete the tick.
Using a Circuit A shortcut for selecting ports and equipment for inclusion in the
Select list report is to use the circuit select list option. If this is ticked the user
can choose from the drop-down list one of the Select lists which
were defined as part of the simulation, model-fitting or mass
balancing procedure. The items from the flowsheet which were
included in the select list are automatically ticked for inclusion in
the report. This feature is useful when working with large, complex
flowsheets.
The Print What The report window has a Print What drop-down list which allows
list users to print port data only, equipment data only or to print both.
This list allows users to (temporarily) not print port or equipment
data items without having to remove the ticks from all those items
in the list.
Selecting Data To select the type of data to be listed in the report (e.g. Exp, Sim
Types for the etc.) place a tick in the box next to the name of the required data
Report types in the Data types to print area of the Report window. The
user can select as many data types as required for inclusion in the
report.
Selecting Error The user can choose to include the data error in a report by placing
data for inclusion a tick in the Error box in the Error Type area of the Report window.
in a report The user must then select from the adjacent drop-down list the
particular error that is to be included in the report. The error data is
useful when working on fitting or mass-balancing data.
Selecting Port If port data have been selected for inclusion in the report the user
data can choose to print the Totals data and/or the size distribution data
for the ports by placing a tick in the appropriate boxes in the Port
data to print area of the Report window. Note that if Component
data have been entered, these can also be selected for inclusion in
the report here. If component data have not been entered, this
option is inactive (as shown here).
Previewing a Once you have configured the report to your satisfaction, click on
report printout the Print Preview button to view the report as it will be printed.
By default, the Print Preview window opens at Page 1 of the
printout with the Zoom setting at 25% of normal size. The user can
change the Zoom setting using Zoom drop-down list and if required
can resize the Print Preview window by dragging any edge or
corner.
The Next Page and Previous Page buttons on the Print Preview
window toolbar allow the user to view all of the pages in the report.
Printing the To print the report simply click on the Print button on the Print
report Preview window toolbar. Alternatively the report can be printed
directly from the Report window by clicking on the Print button on
that window’s toolbar.
Preparing a The Report window has a box marked Summary. When this box is
Summary report ticked, the Report feature uses a summary mode to present the port
and equipment data in the printed report in a different format to the
standard format. The user can choose to use whichever mode suits
their requirements.
In the case of the port data, the Summary mode prints all of the data
of a given type (e.g. Experimental) for all ports in one table. Each
data type selected is printed as a separate table, with all ports listed
in each table. This compares with the normal report mode which
prints the data for each stream on a separate page, with all data
types for each stream being listed on this one page for each stream.
Exporting data A useful feature of the Report Print Preview window is the ability
using Report to export data in report form from the simulator in a variety of
formats. Four buttons on the Print Preview window toolbar
provide the following data export features:
These data export options allow the user to transfer data to other
applications for preparation of presentations and reports.
Note that once any of these file types has been opened, further
saves will append data to it. That is, records of several simulations
in sequence can be accumulated for comparison.
CHAPTER 5
MODEL FITTING
5. MODEL FITTING
Overview For both plant designer and plant operator, model fitting is
primarily concerned with the collection of accurate experimental
data, at either pilot or full plant scale. The model fitting process
provides a powerful means of data examination or assessment as
well as the compression of thousands of data points into a few
parameters.
Sample Analysis Stream size distributions are crucial characteristics for many of the
JKSimMet models. Therefore:
• use a set of sieves that you can trust,
• use the same set of sieves for sizing all of the samples in each
test (sieves can have variations and holes!).
• use a 2 sieve series (size fractions can always be combined
later for convenience)
• sieve to the top of coarse sizes, ie. less than 5% on top screen
and as close to the bottom as possible.
Percent Solids The percent solids of a slurry as measured with a Marcy scale are
subject to error, due to solids density variations in the circuit. Such
variations are common in cyclone underflow streams. Therefore,
percent solids determined from wet and dry sample weights are
preferred.
Steady State JKSimMet is a steady state simulator. Hence, models can most
usefully be fitted to data which were taken at steady state. There
are two practical approaches to this problem:
For a detailed ‘How to Do It Guide’ see the Help Files and Chapter
5 of the Monograph Reference.
Hence, the results can be very useful either for existing plants or for
proposed designs.
Replicate For serious plant testing, where small differences may be worth
Sampling large sums of money, it is often worth carrying out at least one
multiple sample test. That is, instead of taking just one sample set,
take 5 to 10 replicate samples. Then process and analyse each
replicate separately. These 5 to 10 replicates will provide a mean
and standard deviation for every data point. This will provide
invaluable information about the accuracy (or lack thereof) of every
data point.
Concept: If the precision of each data point is measured (or can be estimated
Weighted Sum from experience), then each difference between experimental data
of Squares and simulation prediction can be normalized by dividing by its
precision. That is, a small difference between an accurate data
point and its simulation prediction will make the same contribution
to the weighted sum of squares as a large difference from an
inaccurate data point.
n
∑ (xi - x-)2
i=1,n
Standard Deviation =
(n -1)
Concept: The simulator takes all of the feed streams as input and uses the
Least Squares models and parameters to predict all of the circuit streams.
Fitting If some (or all) of these streams are measured (sampled and sized,
etc), the experimental measurements can be compared with the
simulator predictions. The sum of squares of the differences
between measured data and simulated results is taken as a measure
of goodness of the model fit. The best estimates of the parameters
are expected to be those which MINIMISE the sum of squares.
Hence, the model fitting program adjusts user selected model
parameters to find a best set of parameter estimates which make
the simulator output match the experimental measurements as
closely as possible.
5.3 Background
The JKSimMet models are provided with a set of default
parameters and, in most cases, a range of parameter values. (See
the Supplementary Parameters Manual supplied by JKTech).
For any real mineral processing operation, the best-fit parameters
will almost certainly be different from the default values provided
with the system.
There are several classes of parameters used as model inputs:
Schematically:
SIMULATION
Feed Description
Circuit
MODEL Simulator Configuration
PARAMETERS
Predicted Products
and Streams
MODEL FITTING
Feed Description
Circuit
Simulator Configuration
Adjusted
Model
Parameters
Predicted Products
and Streams
Iterate to
Minimum Sum
of Squares
Measured Products
and Stream Data
Model Fit
MODEL PARAMETER
ESTIMATES
Step 5 Left-click on each item in the list on the Select tab and
observe whether or not it has been selected for fitting.
If the item is selected for fitting the box labelled
Selected will contain a tick. Note also that as you click
on each item in the list it is highlighted in red on the
flowsheet. This feature is useful for identifying which
items you are selecting when the flowsheet is a complex
one.
Step 6 Left-click on the Parameters tab to examine the list of
unit parameters to be fitted. You will find the list
named Ball Mill Parameter List which shows the three
spline knots for the ball mill as the parameters to be
fitted. The initial values of the knots for the fitting are
1.0, 3.0 and 4.0.
Step 7 Select the tab Data to view a list of the port and
equipment data which can be selected for use in the
fitting. The list is named Ball Mill Data List. Note that
the Data list defines which data (and SDs) the models
use in the weighted sum of squares which is minimised
in fitting.
Step 8 The final step before running the model fitting is to set
the standard deviations (SDs) of the stream data which
will be used in the fitting (in this case, the Ball Mill
Product). Bring the Ball Mill Product port data window
into view and, from the drop-down list under Data
Type, select the SDs option. This allows you to view
the data SDs and the Error data along with the
measured and calculated data values. The SD values are
the estimates of the accuracy of the data while Err
(Error) data are differences between experimental
values and those calculated by the model-fitting.
Hint: Experimental, Calculated, SD and Error data can
also be examined on the same screen using the
overview facility which is available from Overview or
on the Data Tab of the Model Fit Window.
Step 9 Make the Model Fit window the active window and
click on the Run Fit tab. Click on the Start button to
start the model-fitting process. The model fitting
program will take these initial estimates provided by the
user and search for better ones, given the experimental
and calculated streams values. It searches until it finds
a minimum residual error (weighted sum of squares of
differences).
If the program finds what looks like a genuine
minimum, it will terminate by providing parameter
error estimates (SDs). In this example, the fit is quite
good. The Errors SDs value is less than 1. This means
that the data are slightly more accurate than the entered
error estimates suggest.
Hint: set the scroll bar slider on the Run Fit window to
allow you to see the fitted parameters being updated
after each iteration.
Step 10 The best fit values for the R/D* knot parameters are
listed in the Selected Model Parameters section, along
with the SDs of these values. (You may need to scroll
across to view these data).
Step 11 Look at the Ball Mill Product port data. Examine the
differences between experimental and fitted data by
selecting the Abs-Fit option from the Error Type drop-
down list and observing these values in the error
column. The differences are relatively small.
Step 12 As an exercise, try graphing the experimental and fitted
data.
This concludes the short guided tour of the model fitting sub-
system. The next step is to work through the tutorial and reference
section with a set of real data.
If you enter new data, or in any way alter the flowsheet which you
select, you should run a new simulation, even if one has been run
before.
Select the The next task is to define which port data the models must match.
Ports This means selecting the ports to be used during the fitting using
for Fitting the Data tab in the Model Fit window.
The maximum number of ports which can be fitted in one fit list is
ten .
Entering and Essentially, this involves entering your data and declaring your
Editing Port confidence about each item in the data set.
Data
Step 1 Bring into view the port data window of the stream
whose data you wish to edit. Note that you may find it
convenient to minimise these windows on screens other
than a very large computer display. A minimised
window will reopen at the tab and position at which you
closed it.
Step 2 Left-click on the Data section and select the SDs option
from the drop-down menu. This brings the SD and
Error columns into view in the port data window, along
with the experimental and calculated data.
Port Data The Totals area of the port data window contains data pertaining to
the total stream: solids, water and volumetric flowrates, percent
solids, pulp SG and solids SG values. Note that numeric characters
displayed in blue on a white background can be entered by the user.
Numeric fields with a grey background are calculated by
JKSimMet and cannot be edited by the user.
The Size Distribution tab area contains the list of the sizings from
Top Size to Size 30 with the value (%) for each size. Whether the
% experimental value refers to % Retained, Cumulative % Retained
or Cumulative % Passing depends upon the setting of the stream
format field shown at the top left of the screen. This setting can be
changed as required by selecting the required sizing format from
the drop-down list in the Format box.
Notice that there are often more size distribution data than will fit
in the port data window. If this is the case, scroll bars at the edge
of the window will allow you to view all of the data. The Page-Up,
Page-Down, Home or End keys or the cursor arrow keys can be
used to move around.
The user must set the SD values, so we shall deal with these data
fields now.
• Globally change all the SDs to one of the six other available
options by:
Concept: Note that a zero SD means the error, i.e. the difference between
Ignoring this experimental and calculated value, will be ignored in the
Data model fitting process. NB This is different from mass balance
where a zero FIXES the result at the measured value!
Error Display Error values can be expressed in one of several ways: Absolute
Error, Percentage Error, or Weighted Error. The user can further
choose whether the error displayed is related to the mass-balancing,
model-fitting or simulation mode of JKSimMet. The user can
determine which of these forms is displayed by the following
procedure:
Step 3 Position the cursor over the Error cell at the top of the
data window and left-click on the black inverted
triangle to make the drop-down list appear.
Error type
drop-down
list
Step 4 Select the error type required from the drop-down list
which is displayed.
Absolute Tells you the actual difference between the calculated and the
Error experimental values.
Percentage Tells you the percentage difference between the calculated and the
Error experimental values.
Weighted Tells you the number that the parameter-fitting program will use in
Error its weighted error sum of squares. You will probably find this the
most useful setting.
Equipment The Parameter tab data in the Model-Fit window is used to define
Parameter which equipment parameters are adjusted in the model fitting.
Selection Initial estimates of the values of these parameters are entered in the
Guessed Value column. These initial estimates of the parameter
values are necessary for the first iteration of the model fitting
process. In subsequent runs of model fitting for the same model,
the user can use the values output from the previous model fitting
run.
The task is to select the equipment (ie model) parameters you want
to adjust in the fitting process.
Parameter Step 2 Place the cursor in the first data cell of the equipment
Selection column and press Enter to bring into view the drop-
down list of equipment on the flowsheet
Cancel Entry To delete a row of data from the parameter list, select the
equipment name in the row you wish to delete and press Enter to
bring the drop-down list into view. Select the option None from
the list to clear the data from the row.
Repeat the above steps for each of the parameters you wish to
include in the model fitting. The maximum number of parameters
is 10. You may, however, define as many parameter lists as you
wish.
Stopping The maximum number of steps field can be set to stop the
Execution of the iterations of the fitting program when required. Alternatively, the
Program fitting program can be stopped at any time, simply by clicking on
the button marked Stop on the Run Fit tab.
Concept: The same approach can be used for each stream point. These
Stream Data values are reported for each fitted stream in the Data tab of the
SDs Model Fit window. A small stream data SD, i.e. where SD
approaches a value of one, indicates a good match between
experimental and calculated data for that stream.
• The summary values in the lower half of the Model Fit window
also indicate the success of the fit. Low values in the Residual,
Error Sum, and Errors SD fields indicate a good fit; large values,
a poor fit. Moreover, in the case of these fields, cross
comparisons between fittings can be made. If these values are
smaller in the most recent run of the fitting than they were in the
previous run, the fit is getting better. If they are getting larger,
you are going in the wrong direction.
• The engineer can also judge the relative success of the fitting by
looking at the stream data windows. Examine the Error column.
Weighted Error and Percentage Error versions of the difference
between calculated and experimental data are most useful.
These are displayed by selecting the appropriate item from the
Error drop-down list.
Printing Model Given that model fitting concerns the experimental (raw) data and
Fit Results the predicted (fitted) data for streams, our task is to print these two
types of data for the streams concerned.
Step 1 Bring into view the data window for the port whose
data you want to print.
Step 2 Left-click on the Print icon on the main JKSimMet
toolbar to view the Print Preview window for this item.
Step 3 If the Print Preview shows that the layout is to your
satisfaction, click on the Print icon at the top, right-
hand corner of the Print Preview window (you may
need to resize this window to see the Print icon)
Step 4 Repeat Steps 1 to 3 for all the other ports whose data
you want to print.
These steps also apply to any other window which has data that you
want to print, such as equipment data.
The best way to produce a printed copy of the error and SD
information on the Parameters and Data tabs in the Model Fit
window is simply to print this window.
You can now refine the format of the plot and print the plot, etc, as
outlined in the section 3.9 (Learning Graphing). Repeat the above
steps for each of the streams for which you wish to compare the
raw and fitted data. The quality of fit is represented by the
closeness of the points to the line, (the closer the better).
Overview This facility provides an excellent summary. Set the % passing
size properties from the Flowsheet Icon on the tool bar, e.g. P80
and % -75 µm.
Skill Model fitting is not a cut and dried procedure. The only way to
versus Practice acquire a useful skill level is to practice on a wide range of real
data. JKSimMet offers a user-friendly environment for what are
really very complex and powerful mathematical techniques.
Data Note that it is necessary to have as much feed and product data as
possible for each of the unit Models to be tuned. Simulation
requires only feed data, but fitting must have some product data as
well.
Not Enough Even when you have the necessary data to perform model fitting, it
Data is essential to ensure that there are enough readings to be useful for
fitting; in general terms, the more data the better.
SDs and The SD settings in the stream data window may be set so that they
Emphasis can cause such an over-emphasis on one parameter that the
potential of the fitting is compromised. Always try to make the
SDs as good an estimate as possible.
Scale Factors The Scale Factor in the Parameters section of the Model-Fit
window can also be a source of problems. If the scale factor is too
big the fitting may stop, because any adjustment in the parameter
produces such a large change that it steps over the minimum of the
sum of squares. On the other hand, however, if the scale factor is
too small, the fitting may stop because any adjustment produces a
change of so small a scale as to be judged insignificant, even
though you may not be close to a minimum point. So, be very
careful with scale factors. As a guide, perhaps a scale factor one-
tenth of the magnitude of the parameter estimate would be a
reasonable place to start.
Large Weighted Examine the weighted errors carefully. These often indicate
Errors suspicious data points. A typical example is a screen top size
which contains several times the predicted weight, because the
laboratory screen stack did not extend to a large enough top size.
Set the error to zero for this fraction to fix the problem.
Knot Positions Where spline functions are used, the knot values can usually be
fitted, but not the knot positions.
5.10 References
GY, P.M., 1982. Sampling of Particulate Materials: Theory and
Practice, 2nd Ed. Elsevier, Amsterdam.
LYMAN, G.J., 1986. Application of Gy's Sampling Theory to
Coal, International Journal of Mineral Processing, 17,
pp 1-22.
LYNCH, A.J., 1977. Mineral Crushing and Grinding Circuits,
(Elsevier, Amsterdam).
NAPIER-MUNN, T.J., MORRELL, S., MORRISON, R.D., &
KOJOVIC, T. 1996. Mineral Comminution Circuits – Their
Operation and Optimisation. JKMRC Monograph Series in
Mining and Mineral Processing 2. Series Editor T.J. Napier-
Munn, Julius Kruttschnitt Mineral Research Centre, University of
Queensland.
CHAPTER 6
MASS BALANCING
6. MASS BALANCING
Overview Even the most carefully collected plant survey data are subject to
many sources of variation. Some of these errors are due to:
• statistical effects
• sampling procedures or design
• assaying procedures
• sizing procedures
• fluctuations in plant flowrates.
6.3 Background
Mass Balancing can be thought of as a type of model fitting. The
models in this case are quite fundamental. Hence, they do not
impose the experience knowledge (which is built into other
mathematical process models) onto the data.
These mass balancing models are:
a b
Concept: The mass balancing module takes all selected streams and
Mass Balancing calculates the smallest set of data adjustments which will make the
data consistent.
If some (or all) of these streams are measured (sampled and sized,
etc), the experimental measurements can be compared with the
data. The sum of squares of the differences between measured data
and adjusted data is taken as a measure of goodness of fit of the
model.
Hence, the mass balancing program adjusts user selected flowrates
to find a best set of flowrates which make the balance output match
the experimental measurements as closely as possible.
Concept: If the precision of each data point is measured (or can be estimated
Weighted Sum from experience), then each difference between experimental data
of Squares and simulation prediction can be normalized by dividing by its
precision. That is, a small difference (or adjustment) between an
accurate data point and its simulation prediction will make the
same contribution to the weighted sum of squares as a large
difference from an inaccurate data point.
Concept: For accuracies of retained size analyses, the Whiten errors often
Whiten Standard provide a realistic estimate. These are calculated as relative
Deviation errors:
• For fractions greater than 10%, a standard deviation of 1.0% is
assumed.
• For fractions less than 1%, a standard deviation of 0.1% is
assumed.
• For fractions between 1% and 10%, a standard deviation of
0.1% plus one tenth of the fraction is assumed.
Schematically:
Mass Balancing
Product 1
Feed
Product 2
Feed (+ δ) Product 1 (+ δ)
Product 2 (+ δ)
(minimize adjustments)
(minimize adjustment)
Model Fitting
Product 1
Feed SIMULATION
Product 2
Parameter Adjustments
Step 3 Click on the Select tab to view the list of units and
streams. Note that the current list is called Select-1 and
that the user can set up more than one list of selected
items.
Click on each stream or unit name in turn to see which
experimental data have been selected for use in the
mass balance. The presence of a tick in the boxes
marked Selected, Water, Feed etc. indicates that the
item will be included in the mass balance. You will
find the unit named Cyclone has three ports selected.
You will find that all the units and water additions and
streams are selected initially. You may turn these on
and off with by left-clicking on the box to make the
tick appear or disappear as required.
Step 4 Click on the Run Balance tab and then select the GSIM
option on the check box on the top right hand corner of
the Run Balance tab window. Also ensure that Select-1
is the option chosen in the drop-down list for the Select
list.
Data Entry Step 5 Bring the Cyclone Overflow data window into view.
Step 6 From the Data drop-down list select the SDs option and
from the Error drop-down list select the Abs-Bal
option. You will see the stream data in a format which
is almost exactly like that in Model Fitting.
Stream Data The Totals tab section of the window contains the experimental
solids and water mass flow values and related data for the stream.
The Size Distribution tab contains the list of the sizings from Top
Size to Size 30 with the value (%) for each size fraction. Whether
the % experimental value refers to % Retained, % Passing, or
Cumulative % Passing depends upon the choice of stream format
selected in the Format drop-down list at the top left of the screen.
The Components tab section of the window contains any assay data
for the stream.
Scrolling Notice that the Size Distribution tab section contains too much data
to display all of it in the window at one time. You can scroll
through the data by using the Page-Up, Page-Down, Home or End
keys, or the cursor up and down control keys. You may also click
on the Print icon on the JKSimMet tool bar for a printout of the
window you are working on.
Data The user must set the data standard deviation (SD) values in the
Accuracy column labelled SD. This column must contain a value in each cell
Entry where corresponding experimental data are entered. Note that this
standard deviation is your estimate of data accuracy obtained from
repeat samples or from experience.
Concept: Note that in mass balancing mode a large SD means the error will
Ignoring be largely ignored. This is different from Model Fitting where a
Data in Mass zero SD switches an error off completely. In mass balancing mode
Balancing a zero value for the SD will make the mass balancer hold the
experimental value constant, ie. it will not be adjusted.
Concepts: If both the experimental value and SD are set to zero, the mass
Calculating balancer will treat this datum as unknown, and estimate a value, if
Missing Data there are sufficient other data provided. This is useful when flow
data cannot be obtained in the stream sample survey.
Now that we know about setting SDs, we can continue our tour.
Step 7 Look at the data window for each port. For this
example you should examine the Feeder called Cyclone
Feed and the Cyclone Underflow and Cyclone
Overflow port data windows.
Step 8 Bring the Mass Balance window into view and select
the Run Balance tab. Left-click on the button marked
Start to run the mass balance algorithm. The program
will execute and when it is completed the results will
be displayed.
Step 9 Bring into view the port data windows and examine the
raw and adjusted data in each stream.
Step 10 Compare the raw and adjusted data graphically by
selecting Graph from the Cyclone properties list (Right-
click on Cyclone Icon). Select experimental then
calculated for a quick comparison.
Note that for GSIM, percent solids and internal water flows are
always enabled.
Because each Select list has a name, you may set up several
different lists to examine different sections of a circuit. You can
select every stream and flow to balance or any single unit or
selection of streams.
Note: in V5.0 stream data are stored in equipment ports. To
balance a subset of the data, you need to choose both equipment
and ports.
The “water” check box, applies to Water Feeders. These should
only be used with GSIM or when “% solids as a component” is
checked on the component tab.
6.6.4 Component
The mass balancing module can analyse two types of data, namely
components or size distributions. The model-fitting and simulation
modes in JKSimMet only use size distributions. The size format is
called GSIM for Grinding Simulation. JKSimMet uses GSIM as
the default component type. If GSIM is selected via the GSIM
Mode check box on the Run Balance tab or the Components tab, no
further data entry is required.
To perform a mass balance using data other than size the user must
select a Component list and give it a name. The next step is to
define each of the components.
Step 1 Left-click on the Component tab to view the
component list. For the Copper Flotation example, the
list is called CuFe and the components are copper and
iron assays, %Cu and %Fe.
Step 2 Select the SD’s error display type from the Data drop-
down list.
Error Display The right-most column under the Components tab in the stream
data window is the Error column, and the values in it can be
expressed in one of three ways; Absolute Error, Percentage Error or
Weighted Error. The user can select which of these forms is
displayed by the following procedure:
Absolute Error Tells you the actual difference between the calculated and the
experimental values.
Percentage Tells you the percentage difference between the calculated and the
Error experimental values.
6.6.5 Water
If you have selected %Solids as a component on the current Mass
Balance window Component tab you can also mass balance any
water additions to the circuit.
In Version 5, mass balance water additions are made by selecting
the appropriate water feeder and including it on the Select list. For
this example, you will need to add two water feeders to the
flowsheet. Measured data and estimated SDs are entered in the
water addition data window. This removes the need for the water
addition list used in Version 4.
See Appendix A1.2 for a detailed description of the Water Feeder.
Some typical water addition and % solids data are tabled below.
Select % Solids as a component on the Component tab (see section
6.6.2 on Components) and also select water additions on the Select
tab (see section 6.6.4 on Selecting Data) to use this facility
Water Feeders:
Clnr Feed Sump 20 SD 1.0
Scav Conc Sump 15 SD 1.0
Remember:
Select the Water Feeders and tick % solids as a component in the
Mass Balance dialogue window.
Step 2 Ensure that the correct Select and Component lists are
selected in the cells above the main data area. For the
Copper Flotation example we will use the existing
Select list Mass Balance Select List 1 and the
Component list CuFe.
Step 3 Left-click on the Start button to start the mass
balancing program.
Concept: The same approach can be used for each stream point. These
Stream Data values are reported for each fitted stream in the Mass Balance
SDs window and the stream data windows.
Note also that such a result may mean you are trying to balance
around a splitter or a classifier which is not classifying.
We shall deal with these in turn. For mass balanced data, graph
plotting is limited to GSIM format.
6.8.1 Overview
The overview window gives you a powerful means of summarising
your data and checking it for adjustment problems. Each overview
data set defined by the user displays a list of data from all selected
streams. The user can select the types of data which are displayed
in the overview window.
The best way to use the overview feature is to compare
experimental and calculated values for each assay (or size fraction)
across the complete circuit. This will give a very useful picture of
the accuracy of the data and the mass balance. Note that the
overview window can be configured to show either data or
calculated Recovery information.
Recovery To examine recovery data for the components in the streams select
Selection Recovery by placing a tick in the Recovery box.
Note: The balanced recovery selection will calculate recoveries
based on mass balanced assays and flowrates.
By default, the recovery is calculated with respect to the circuit feed
stream. This reference stream is labelled in bold text in the
Overview table. To change the stream which is the reference for
You can now refine the format of the graph and print it etc., as
outlined in the section 3.8 (Learning Graphing). Repeat the above
steps for each of the streams for which you wish to compare the
raw and calculated data. The goodness of fit is represented by the
closeness of the points to the line; the closer the lines and points,
the better the fit.
Skill Mass Balancing is not a cut and dried procedure. The only way to
versus Practice acquire a useful skill level is to practise on a wide range of real
data. JKSimMet offers a user-friendly environment for what are
really very complex and powerful mathematical techniques.
Different Sizing Be very wary of changes in size measurement technique e.g. from
Techniques screens to Cyclosizer.
Different Assay Where assay techniques change between stream samples, as they
Techniques sometimes do for different assay ranges, there may be inherent
biases within the assay techniques. These will lead to biases within
the mass balance.
Data Note that it is necessary to have enough feed and product data to
achieve a useful mass balance.
Some Common There are a couple of simple traps which can appear in many
Mass Balancing guises. If you become aware of these now you may recognize them
Pitfalls more easily when you encounter them.
b
a
m
Feed
x Copper Au x Lead x Zinc x Tailings
Circuit Circuit Circuit
x x x
Cu Au Pb Zn
Conc Conc Conc
6.11 References
LYNCH, A.J., 1977. Mineral Crushing and Grinding Circuits,
(Elsevier, Amsterdam), Chapter 7.
LYMAN, G.J., 1986. Application of Gy's sampling theory to coal,
International Journal of Mineral Processing, Vol 17:1-22.
GY, P.M., 1982. Sampling of particulate materials: theory and
practice, 2nd Ed, (Elsevier, Amsterdam).
MORRISON, R.D., 1976. A two stage least squares technique for
the general material balance problem, JKMRC Internal
Report No 61 (unpublished).
APPENDIX A
Model Descriptions
A1 Introduction
This appendix of model descriptions contains:
These provide:
• A source for ore
• A source for water
Ore Feeder The ore feeder (called Feed) is a specialised piece of equipment
which has a single ‘product’. The Feed unit allows you to set up
the flowsheet ore SG and the default size distribution.
The size markers, i.e. Percent passing a particular size and size at
a particular percent passing can be set by double clicking on those
fields on the Totals tab. Note that while the flowsheet properties
dialogue allows you to set global properties for Data Information
blocks and tools such as simulation and model fit, these
properties for the feeder and ports may be set at different values
for each.
Water Feeder The water feeder replaces the ‘Unit Feed Density’ section of each
model in JKSimMet Version 4. The three models provided with
the Water Feeder are functionally identical to the three options for
‘Unit Feed Density’.
Option 2 – This option allows the user to set a maximum percent solids for
Required % the total feed to the connected equipment.
Solids
If the feed percent solids is higher than ‘Required % Solids’ the
water feeder adds additional water to achieve the required percent
solids.
Option 3 – Water Water Addition is the recommended mode for common use. The
Addition user specifies the required water addition in cubic metres per
hour.
This option has two more uses. The experimental water addition
may be used as a parameter in Model Fitting. That is, a model fit
may use water addition as a parameter when water flows were
unmeasured or the measurement is dubious.
Note that percent solids or water flow from the circuit should be
constrained by a small SD value to provide a constraint on total
water addition.
The User provides the ‘exp New Water Addition’ and an ‘sd’
estimate on this model. The other requirement is that the Water
Feeder and Water are selected on the Select Tab of the Mass
Balance tool.
Variable Rates The Variable Rates SAG model also has some differences –
SAG Model detailed in Appendix 11.
A1.2.4 Splitters
Splitters The range of splitter models has been increased. These are
discussed in Appendix 14.
Further, the effects of inlet diameter, cone angle and cylinder length
have been evaluated as
Here Kw1 and Kv1 are constants also depending on feed solids
characteristics. The current data indicate that Kw1 and Kv1 are
independent of cyclone diameter for geometrically similar cyclones
treating identical feed solids. Small quantities of viscosity
modifiers such as clay, can have a marked effect on these variables.
Efficiency Curve The efficiency curve used in this model is given below:
Relationship
Scaling Facilities for scaling the operation of the hydrocyclone are built
into the model.
100
Increasing Alpha
% of Feed to Overflow (corrected)
80
60
40
20
0
0.00 0.50 1.00 1.50 2.00 2.50
d/d50 (corrected)
180
160
Increasing Beta
120
100
80
60
40
20
0
0.00 0.50 1.00 1.50 2.00 2.50
d/d50 (corrected)
180
160
% of Feed to Overflow (corrected)
120
80
60
Efficy. curve at d50c
40
20
0
0.00 0.50 1.00 1.50 2.00 2.50
d/d50(corrected)
A2.4 Symbols
Symbol Meaning
Cyclone Roping If the cyclone feed density is less than 35% solids by volume, the
Constraint SPOC constraint (Laguitton 1985) is claimed to predict onset of
roping.
In tabular form:
at sg 2.7 at sg 4.0
Feed Underflow Feed Underflow Feed Underflow
Density Density Density Density Density Density
% by Volume % by Weight % by Weight
5 53 12.4 75.3 17.4 81.8
10 54 23.1 76.0 30.8 82.4
15 55 32.3 76.7 41.4 83.0
20 56 40.3 77.5 50.0 83.6
25 57 47.4 78.2 57.1 84.1
30 58 53.6 78.8 63.1 84.7
35 59 59.2 79.5 68.3 85.2
Increased Q d50c Rf Rv
Typical values of KQO are in the range 300-600. The scale factor
for fitting should be 100.
Classification Equations A2.4 to A2.6 define the cut size. KDO is typically a
Size (KDO) small number - say 0.001 to .00001. Therefore, a scale factor of
0.0001 is usually suitable.
Water Split % to The actual water split to overflow (Cal WS) is fitted rather than the
O/F (Cal WS) two parameters, KV1 and KW1, which are defined by a single
water split.
After fitting, the calculated values of KV1 and KW1 are displayed
on the cyclone equipment data window (Model Parameters tab).
Efficiency Curve The reduced efficiency curve is an "S" shaped function as shown in
α and β)
(α Figure A2.1.
If the efficiency curve is a poor fit at coarse sizes, try the alternative
fines modified or spline efficiency curve models.
Fit KD0, KQ0, α and Cal WS for each data set independently, and
determine the average values of KV1 and KW1 for each cyclone
data set from the fit. Use the average values of KV1 and KW1 in
each cyclone data set. Use Master/Slave to fit KD0, KQ0, α and (if
required) β, over all data sets.
A2.8 References
DE KOOK, S.K., 1956, Symposium on recent developments in the
use of hydrocyclones - a review J. Chem. Metal. Min. Soc.
S.Afr., Vol. 56:281-294.
KAVETSKY, A., 1979. Hydrocyclone modelling and scaling.
JKMRC report to AMIRA, November.
KELSALL, D.F., 1953. A further study of the hydraulic cyclone.
Chem. Eng., Sci., Vol. 2:254-273.
LAGUITTON, D. (Ed), 1985. The SPOC Manual Simulated
Processing of Ore and Coal, CANMET EMR Canada, Ch.
5.1 (Part B).
LYNCH, A.J. 1965. The characteristics of hydrocyclones and their
application as control units in comminution circuits, AMIRA
Progress Report No. 6, University of Queensland
(unpublished).
LYNCH, A.J. and RAO, T.C., 1965. Digital computer simulation
of comminution systems. Proc. 8th Comm. Min. Metall.
Congr., Aust., N.Z., Vol. 6:597-606.
NAGESWARARAO, K., 1978. Further developments in the
modelling and scale-up of industrial hydrocyclones. Ph.D.
Thesis (unpublished). University of Queensland.
PLITT, L.R., FLINTOFF, B.C. and STUFFCO T.J., 1987. Roping
in hydrocyclones. 3rd International Conference on
Hydrocyclones, Oxford England, Elseveir, pp21-23.
YOSHIOKA, N. and HOTTA, Y., 1955. Liquid cyclone as a
hydraulic classifier. Chem. Eng. Jpn., Vol. 19:632-640.
Particle Size
100
Steel
10
Rubber
1.0
0 F W1 FW2 High
The explanation of Figure A3.2 is that when the feed rate to screens
with rubber decks is low the particles move independently,
accumulate energy, take large bounces and have little opportunity
to pass through the screen aperture.
An increase in feed rate causes an increase in inter-particle
collisions, reduction in particle energy and bounce lengths, and an
increase in number of trials. Hence, the screen efficiency increases.
A further increase in feed rate causes more particle interference, a
decrease in the number of trials due to particles not reaching the
screen surface, and a decrease in screening efficiency.
With steel screens, however, the coefficient of restitution is low
and particles do not accumulate energy. Particle bounces are small
and high efficiencies occur at low feed rates. As the feed rate
increases the inter-particle interference increases and this reduces
the number of trials and the screening efficiency.
Fines Factor The fines factor is used to describe the "piggyback" effect of fines
on coarse material.
and SF* Area is the t/h of fines which are carried into the oversize
product.
Moisture For damp ores, the behaviour of moisture can be very important.
Behaviour There are sometimes several kinds of moisture. The only one of
interest to this model is in the fines, that is, fractions finer than the
"Moisture Split Critical Size XM".
Scaling The model allows scaling of screen length by linear scaling of the
number of trials parameter, TRN.
A3.4 Symbols
Symbol Meaning
xi size of particles in the ith size fraction
E(x) fraction of particles in the feed of size
x which enter the coarse product
PARAMETER MENU
Ap Length
Ap Width
OA %
A int (TRN)
B*FW (TRN)
D*FW (TRN)
U*P1 (TRN)
U*P2 (TRN)
E int (FF)
F*PSF (FF)
G*PSF (FF)
XF
XM
Aperture Length Where screen data do not provide precise aperture and wire
and Width dimensions, the screen aperture can be fitted to the data.
Note that for slotted screens, effective aperture length depends on
the shape of the particle because the size data are measured using
square mesh screens.
Selection of feed Screen performance can be affected by the feed size distribution.
size parameters This is usually a secondary dependence compared with feed rate.
X1 to X4 However the model does allow it to be incorporated. X1 and X2
are the upper and lower sizes of a critical size fraction (or
fractions). If a particular range of sizes in your feed data is highly
variable use X1 and X2 to bracket it.
Critical Size If your Trials (TRN) versus feedrate data are erratic and your data
Dependencies are a good fit (less than 2 stream SD with Whiten weights), then it
is worth trying P1 and P2 dependencies.
Submesh Factor This is the other important dependence. For many operations, SF is
Modelling small and more or less constant. However, for operations with
damp ore, it can be crucial to a good model.
Once again, plot your best fit SF values against the calculated PSF
and TSF values from the parameter screens and draw a linear
regression line against any one variable. Print out the graph with a
fine grid for this slope (for G) and the intercept E. points.
Running the Input your estimated values back into the screen menu and import
Screen Model to each of your data sets. Try a simulation and check agreement
on product streams. Expect to make errors in this procedure the
first few times.
A3.8 References
WHITEN, W.J. and WHITE, M.E., 1977. Modelling and
simulation of high tonnage crushing plants, XII
International Mineral Processing Congress, Brazil, Volume
II, 148-158.
WHITEN, W.J., 1984. Models and control techniques for crushing
plants, Control 84, Mineral/Metallurgical Processing,
(Editor, J A Herbst), Publishers - AIME, New York, 217-
225.
A4.5 Symbols
Equivalent
JKSimMet
Symbol Meaning
parameter
Reduced efficiency curve sharpness
α Alpha
parameter.
Reduced efficiency curve dip
β Beta
parameter.
Parameter for describing the
β* Beta*
reduced efficiency curve.
Range of Validity The highly simplified form of these models means that
extrapolation away from the conditions at which the parameters
were determined will significantly decrease the accuracy. If a wide
range of data is available, it may be possible to use Model 251 (see
Appendix A5) which has a variable cut point.
PARAMETER MENU
A4.8 References
LYNCH, A.J., 1977, Mineral Crushing and Grinding Circuits.
(Elsevier, Amsterdam) pp. 124-126.
Efficiency Curve The efficiency curve used in this model is given below:
Relationship
. . . (A5.3)
A5.4 Symbols
Symbol Meaning
α reduced efficiency curve sharpness parameter
β reduced efficiency curve dip parameter
β* reduced efficiency curve calculated parameter
W,X,Y,Z regression constants in the d50c equation
d50c size of particle in the feed which has equal probability
of going to fine or coarse product
C % water split to fine product
SW slot width (mm)
FW volume flow rate of water in the feed (m3)
FPS % solids in the feed
Screen Wear DSM screens are sensitive to wire wear condition. The screens are
usually reversed on a regular basis. If possible, test data should
record the wear condition. If this is not possible, test at both new
and worn to obtain a range of likely operation.
A5.6 Fitting
PARAMETER MENU
W * Slot
X * FPS
Y * FdWater
Z (int)
Sharpness α
Dip β
C
Multiple Data If the data cover a range of feed rates, feed percent solids, slot
Sets widths and screen widths, proceed as follows:
Note: If the slot width does not have a strong effect on d50c, then
the data are very questionable.
A5.7 References
LYNCH, A. J., 1977, Mineral crushing and grinding circuits,
(Elsevier, Amsterdam), pp 124-126.
f x
Classification
Function
p
Appearance
A*C*x Function C*x
Since the feed and product are expressed as size distributions, and
the properties of the internal classification and breakage
mechanisms are expressed with respect to particle size intervals or
mean sizes, it is convenient to represent these quantities as vectors
and matrices respectively.
Since f is known and p is the desired output, the problem resolves
itself into obtaining estimates of C and A for a particular machine
and feed material. These values can then be manipulated by
simulation to explore the effects of changing machine parameters,
material characteristics or operating conditions upon the product
size distribution.
An important limiting factor in crusher operation is the power
drawn by the machine. This model permits estimates of power
draw to be made for a given condition, so that the simulations can
be constrained by power requirements (by the user). The power
draw can be normalised to experimental data or estimated from
data for similar crushers in the Supplementary Parameters Manual.
Note: a single particle breakage test of the ore is required for either
type of power estimate.
C(x) = 1
(A6.4)
for x > K2
i.e. all particles are broken
K2 - x K3
C (x) = 1 - K - K (A6.5)
2 1
1.0
K3
0.0
K1 K2
Particle Size x
Note that only closed side setting and crusher throw will normally
be used. The other relationships require a very detailed
experimental database.
t
4
t
10
Y/10 Y/4 Y/2 Y
Particle size mm
First Typical cone crusher operation for secondary and tertiary crushers
Approximation will be at a t10 of 15 to 20. For a lightly loaded crusher (size
control on a primary jaw crusher) will operate at a t10 of 5-10. A
high reduction crusher (toothed roll or choke fed tertiary) may
achieve a t10 up to 25.
1.5
t10 = 30.0
t10 = 20.0
1.2 t10 = 10.0
0.9
0.6
0.3
0.0
10 14 18 22 26 30
Size mm
Reduction
Particle size mm
parameter
t10 14.50 20.63 28.89
Specific comminution
%
energy kWh/t
10.0 0.35 0.30 0.25
20.0 0.80 0.70 0.50
30.0 1.2 1.0 0.80
Power Prediction A power prediction method has been developed using energy –
Method size-reduction information from the pendulum test (Andersen &
Napier-Munn, 1988) and is also applicable to the Drop Weight test.
Using the ore-specific energy-size-reduction relationship from the
pendulum test, the breakage function, B the classification values Ci
(from the parameter fitting or model regression equations, the
model calculates the total energy required to reduce the feed size
distribution to the product size distribution as if all the feed was
broken in the pendulum or drop weight testing device, i.e. it
defines the energy which would have been used by the breakage
device to achieve the same degree of breakage observed in the
crusher. The sum of the products of the amount of material selected
for breakage in each size fraction, Ci*xi (tonnes) (from equation
(A6.3)), and the Ecs (kWh/t) for each size at the breakage
parameter value t10 (determined from the parameter fitting or model
regression equation (A6.11)), is the total comminution energy
calculated by the model, Pcalc (kWh).
Model 400
Model 405
A6.8 Symbols
Symbol Meaning
f feed size distribution (vector)
p product size distribution (vector)
A appearance function (matrix)
C classification function (diagonal matrix)
I unit matrix
K1 size below which C = O
K2 size above which C = 1
K3 exponent in the equation for C
CSS closed side setting (mm)
TPH tonnes/hour feed
F80 coarseness of feed, e.g. 80% - 25.4mm
t10 breakage distribution factor
total power consumed in size reduction using the
pendulum (from laboratory tests results)
Ai regression constants
Bi "
Ci "
Di "
Ei "
PARAMETER MENU
There are two distinct levels of use of the crusher model. The
different uses require different fitting strategies.
Limited Data One data set allows a (somewhat approximate) estimate of product
size for small variations in closed side setting.
For one data set:
fit A4 and B5
with A0 = 1.0, and B0 = 2.0 and (for cone crushers) K3 = 2.3.
Set other A and B values to zero.
Similarly for the breakage function:
fit D3
D0, D1 and D2 are set to zero.
Note that one data set does not provide useful information about
power dependencies.
Extensive Data The model is much more useful with a range of data. This means
5 to 10 data sets covering a range of crusher settings, feed rates and
(if possible) feed sizes.
JKTech can undertake breakage tests to characterize an ore as
shown in Figure A6.3 and Table A6.1 and to determine the size
single particle breakage/power as shown in Figure A6.5 and Table
A6.2.
HP Grinding Note that the value of K3 is generally 2.3 for cone, jaw and
Rolls gyratory crushers only. For other types of crushers, such as
(and others) grinding rolls and hammer mills, it is advisable to fit K3 also, with
2.3 as a good initial estimate
Master/Slave Master/Slave model fitting is available for the crusher model in the
Fitting general release version of JKSimMet.
A6.13 References
ANDERSEN, J. A., 1989. M.Sc. Thesis, University of
Queensland, (unpublished).
ANDERSEN, J.A. and NAPIER-MUNN, T.J., 1988. Power
prediction for cone crushers, Mill Operators' Conference,
Cobar.
AWACHIE, S.F.A., 1983, Development of crusher models using
laboratory breakage data, PhD Thesis, University of
Queensland.
MULAR A. L. & BHAPPU, R. B. 1978, Mineral Processing
Plant Design.
WHITEN, W.J., 1984, Models and control techniques for crushing
plants, Control 84, Minl./ Metall.Process
(Am.Inst.Min.Engrs. Annual Meet., Los Angeles, USA,
February), 217-225.
m=C.q+f
(I-S).m
Selection S Diagonal Matrix S
C.q
Classification C Diagonal Matrix C
pj
product vector from stage j
p = (X)v ⋅ f (A7.1)
or
p = X⋅X⋅X ... ⋅ f for v times
Non-integer numbers of stages can only be calculated by
interpolation.
Once A, S and C are known, any particular operating condition can
be represented by a value of v.
Feed Rate The key dependence is the variation of stages of breakage v with
mill feed rate F.
Experimentally:
F (v)1.5 = MC
where MC is the mill constant.
The mill constant can also be scaled as detailed later.
A(x,y) = (1-e-x/y)/(1-e-1)
Where A(x,y) is the proportion after breakage of particles of initial
size y which are smaller than size x. The appearance A is made up
of vectors of the differences in x for the specified screen interval.
For specific ores, JKTech can measure the appearance function. A
range of appearance functions for various ores is given with the
ball mill model description in Appendix A8.
1.0
SL
Selection
Function
IN
0.0
XC
Size
Scaling The rod mill model is scaled by modifying the mill constant
according to dimensions and operating conditions described below:
These scale factors only apply for rod mills with normal length to
diameter ratios, that is, 1.2 < simulated L/D < 1.6 and L ≤ 7m.
Media Load Load Fraction (i.e. volume of mill occupied by charge and media at
rest after grinding out)
(1 - LFSIM ) ⋅ LFSIM
FACTB =
(1 - LFFIT ) ⋅ LFFIT
Note 30% < LF < 45%
CSSIM
FACTC =
CSFIT
F90FIT
FACTD = ln F90 / ln 2
SIM
MCSIM2/3
vSIM = + FACTD
FRSIM
A7.4 Symbols
Symbol Meaning
f feed size distribution (vector)
p product size distribution (vector)
A appearance function (step matrix)
C classification function (diagonal matrix)
S selection function (diagonal matrix)
Si element of selection function S from size i
v number of stages of breakage of original mill
vSIM v for simulated mill
F90FIT 90% passing size for fitted mill feed
F90SIM 90% passing size for simulated mill feed
MCSIM mill constant for simulated mill
MC mill constant for original or fitted mill
SL slope of selection function
IN intercept of selection function
XC Size below which selection function is constant
DSIM diameter of simulated mill
DFIT diameter of fitted mill
LSIM length of simulated mill
LFIT length of fitted mill
LFSIM load fraction of simulated mill
LFFIT load fraction of fitted mill
CSSIM fraction critical speed of simulated mill
CSFIT fraction critical speed of fitted mill
WISIM work index of ore for simulated mill
WIFIT Work index of ore for fitted mill
Note: The fitted mill is the rod mill which provided the
experimental data.
Scaling Note the restrictions for scaling in the section on Model Equations.
Change in Feed The number of stages of breakage is calculated from the feed solids
Pulp Density mass flow. No account is taken of water in the feed. It is assumed
that rod mills operate at 75 to 85 percent solids in the feed.
Effect of Feed There is some doubt about the adjustment of number of stages of
Size breakage according to feed coarseness. Data from some operations
exhibit an effect while data from others do not. If the particles are
large enough and strong enough to resist a rod impact, the
dependence is reasonable.
A7.7 Reference
LYNCH, A.J., 1977. Mineral crushing and grinding circuits,
(Elsevier, Amsterdam), 51-60.
p = D•s (A8.1)
Within the mill, two factors control breakage. The first is the rate
of selection of each size for breakage. The second is the way in
which the selected particles are broken (or appear) in the mill
contents.
Selected = R•s
Appearance = A•s
At steady state, the mill feed minus the material selected for
breakage plus the material from breakage minus the material
discharged must equal zero. This can be written as:
Discharge Rates For overflow mills and most of the operational range of grate
discharge mills, the discharge elements can be approximated by:
*
D
i
0.
Discharge
Log (size)
Screen Size
Figure A8.1 - Typical graph of mill discharge function
Breakage Rates Breakage rates tend to increase rapidly with particle size, with the
increase tapering off at the feed top size.
Log (R )
i
Log (size)
Figure A8.2 - Typical graph of breakage rate factor
pi = Disi (A8.5)
where feed and product are related by R/D for a particular breakage
function.
Equation A8.3 can be used to scale for feed rate and mill
dimensions.
Scaling Scaling of the ball mill model is achieved by modifying the fitted
R/D* function according to dimensions and operating conditions as
described below.
dSIM
FACTA = d
FIT
Load Fraction The load fraction LF is the volume of mill occupied by charge and
media at rest when the load is ground out.
(1 - LFSIM ) . LFSIM
FACTB =
(1 - LFFIT ) . LFFIT
WISIM0.8
FACTD = WI
FIT
Ball Size Scaling By assuming that the reduction mechanisms of impact and attrition
occur in a ball mill, the following relationships can be derived from
theoretical considerations.
( 1
FACTE = Db
) = Db FIT
(1Db ) Db
SIM
SIM
FIT
New xm Old x m
Scaling These factors are applied to each fitted 1n (R/D*) knot as follows:
Calculation
R/D*SIM=R/D*FIT•FACTA•FACTB•FACTC/FACTD•FACTE
Scaling Using Where characteristic breakage functions have been measured (i.e.
Breakage pendulum tested) for both ores, these breakage functions may be
Functions used to predict performance. Note that it is not valid to scale this
way from the default breakage function.
A8.4 Symbols
Symbol Meaning
f feed size distribution vector
p product size distribution vector
s mill contents size distribution vector
A appearance function lower matrix
R breakage rate function diagonal matrix
D breakage discharge function diagonal matrix
D* normalised discharge function
R/D*SIM normalised R/D ratio for simulated mill
R/D*FIT normalised R/D ratio for fitted mill
dSIM diameter of simulated mill
dFIT diameter of fitted mill
LSIM length of simulated mill
v volume flow rate of feed
LFSIM load fraction of simulated mill
LFFIT load fraction of fitted mill
CSSIM fraction critical speed of simulated mill
CSFIT fraction critical speed of fitted mill
WISIM work index of ore for simulated mill
WIFIT work index of ore for fitted mill
Db ball diameter (top size)
DbSIM ball diameter for simulated mill (top size)
DbFIT ball diameter for fitted mill (top size)
K maximum breakage rate factor
xm maximum breakage size.
Critical Speed The critical speed dependence is approximately valid for 55-78%
Range of critical speed and incorrect outside of that range.
Predicting Rates If the ball mill model does not produce any of a coarse fraction (i.e.
at 'Missing Sizes' none in the mill discharge) then the effective rate of grinding is
'infinite'. One way to overcome this problem is to size the mill
contents and expand to the perfect mixing model used for the SAG
mill model.
High Mill The perfect mixing mill model only takes account of pulp density
Viscosity or variations as variations in mill volume. Therefore, higher pulp
Pulp Density density will always predict higher grinding rates. In practice, the
rates do improve until pulp viscosity begins to interfere with ball
action and rates decrease rapidly. This onset is difficult to predict
as it is highly ore type dependent. However, effective mill
operation of greater than 50% solids by volume is unlikely and
improbable at greater than 60% solids by volume.
Ball Size Scaling The ball size scaling relies on the R/D* function exhibiting a
maximum. If there is no maximum in the fitted R/D* function,
increasing the ball size will give optimistic results.
Wide Range Size The mill model assumes a constant breakage function for all size
Data fractions. This assumption simplifies the model but experimental
evidence suggest strongly that partial breakage increases in severity
with decreasing size - down to some limiting size. Therefore, if
more than (say) twenty size fractions are considered, an apparent
minimum rate may be produced in the finer ranges. This
phenomenon is more likely to be an artefact of an incorrect
assumption than to have any physical significance. Research work
continues in this area.
R/D* Spline Use three knots for normal grinding conditions and four knots for a
Knots wider than usual size range (such as SAG mill discharge or a very
fine product).
Knot Estimates Estimates for the function values at the knot positions are provided
as ln(R/D*) values. A simple ascending series provides a good
first estimate, for example:
Work Index, If you have several sets of data, use an operating Work Index for
Load Fraction each (calculated from mill feed rate, mill power, feed and product
80% passing sizes). If the major variation is hardness only, then
the average knots can be used.
The calculated R/D* values are displayed on the unit data entry
screen. There should be a smooth increase with size. Sometimes
the curve will have a maximum at the coarse end. If there are any
sudden changes or ups and downs, try adjusting the knot positions.
Graph Cumulative When nothing else works, plot the experimental feed and product
Simulated and on a coarse scale (say 0-30%) percent retained against log size. If
Experimental there are any large discontinuities, check your data very carefully,
Product and repeat your sampling if possible.
Master/Slave The perfect mixing ball mill model is well suited to fitting of
Fitting multiple data sets. The ln(R/D*) knot values can be fitted
simultaneously for a number of surveys. Ensure that you use the
same knot positions, and number of knots, for each mill in your
master/slave fitting test.
Size Interval Ball milling circuits from which the samples were collected:
Massive Massive Porphyry Porphyry Massive
Sulphide Sulphide Hard Soft Sulphide
(Ni) Coarse (Cu) (Cu) Fine
(Pb-Zn) (Pb-Zn)
1 0.000 0.000 0.000 0.000 0.000
2 0.0591 0.0505 0.08586 0.05220 0.1128
3 0.1052 0.0974 0.1248 0.09919 0.1490
4 0.1318 0.1276 0.1387 0.1288 0.1497
5 0.1295 0.1278 0.1278 0.1284 0.1250
6 0.1127 0.1128 0.1076 0.1129 0.09885
7 0.0927 0.09469 0.08722 0.09423 0.07866
8 0.07486 0.07810 0.06960 0.07727 0.06289
9 0.06082 0.06428 0.05540 0.06339 0.04943
10 0.05005 0.05316 0.04428 0.05239 0.03842
11 0.04166 0.04424 0.03574 0.04364 0.03003
12 0.03462 0.03666 0.02899 0.03623 0.02376
13 0.02723 0.02880 0.02278 0.02847 0.01865
14 0.02054 0.02171 0.01743 0.02146 0.01448
15 0.01537 0.01623 0.01325 0.01604 0.01120
16 0.01144 0.01207 0.01004 0.01192 0.00864
17 0.00849 0.00894 0.007581 0.00883 0.00664
Operating Work Index
12.8 9.0 13.6 12.2 15.9
Size Interval Ball milling circuits from which the samples were collected:
Quartzite Porphyry Massive Massive Standard
Sulphide Soft Sulphide Sulphide Function
Low Grade USA (Cu,Pb,Zn) (Pb, Zn, Cu)
(Cu) (Cu)
1 0.000 0.000 0.000 0.000 0.000
2 0.09514 0.05013 0.1171 0.1081 0.193
3 0.1322 0.0970 0.1537 0.1442 0.157
4 0.1417 0.1273 0.1522 0.1472 0.126
5 0.1267 0.1276 0.1247 0.1253 0.101
6 0.1049 0.1128 0.09723 0.1006 0.082
7 0.08477 0.09481 0.07685 0.08050 0.066
8 0.06778 0.07832 0.06131 0.06444 0.053
9 0.05371 0.06451 0.04810 0.05076 0.043
10 0.04244 0.05336 0.03729 0.03958 0.035
11 0.03379 0.04438 0.02911 0.03103 0.028
12 0.02709 0.03677 0.02303 0.02459 0.022
13 0.02127 0.02888 0.01810 0.01929 0.018
14 0.01637 0.02177 0.01407 0.01496 0.015
15 0.01254 0.01628 0.01089 0.01155 0.012
16 0.009565 0.01211 0.008413 0.008888 0.010
17 0.007279 0.008968 0.006483 0.006825 0.008
Operating Work Index:
14.1 10.2 14.1 13.5
A8.8 References
LYNCH, A.J., 1977. Mineral crushing and grinding circuits,
(Elsevier, Amsterdam), 309-312.
WHITEN, W.J., 1976. Ball mill simulation using small
calculators, Proc. Australas. Inst. Min. Metall., 258, 47-
53.
MORRELL, S. 1992. Ball size effects in ball mills. Chapter 2,
End of project report, AMIRA/JKMRC Project P9J.
"Simulation and Automatic Control of Mineral Treatment
Processes".
Load
Feed
Product
High Low
Energy Energy
(impact) (abrasion)
The model assumes that each size fraction experiences only one
energy level of breakage. (The reality will certainly be a
distribution of energy levels).
High Energy The relationship between the amount of breakage and the input
Breakage energy is described by
Low Energy One or more 3 kg samples of 50 mm natural ore are tumbled for 10
Breakage minutes in a small dry mill at 70% of critical speed. The products
of each run are sized and t10 is measured for each run.
Where 50mm material is not available, other sizes are used and
adjusted using a simple linear model.
The t10 data are fitted to
t10 = a0 + a1 * mean size + a2 * sample mass + a3 * time. (A9.2)
Low Energy The size distributions produced by ores tested to date have a similar
Appearance shape. This shape can be scaled to the ta factor that is the
Function - percentage passing one tenth of the original particle size.
Abrasion
A cubic spline function is used for smooth interpolation.
A 100mm particle is chosen as an example as particles of this size
will typically undergo abrasion rather than crushing breakage.
Parameter ta is taken as 1.0 to make the scaling obvious.
size (mm) % passing
Breakage Energy As the charge provides the grinding media, the level of available
energy is related to the coarse fraction of the mill charge.
The average size of the top 20% of the charge is used as the highest
energy reference level.
S20 = (p100 * p98 * p96 ... p80) 1/11 (A9.3)
and the potential energy at the full height of the mill
4
E1 = 3 π (S20)3 ρ g D (A9.4)
High Energy Once the energy of breakage is known, the distribution the particle
Appearance breaks into can be described by a cubic spline surface.
Function
(Crushing
Breakage)
50 100
t2 25 92.49
t4 12.5 61.58
t10 5 30.00
t25 2 15.62
t50 1 9.412
t75 0.67 6.893
Combined As noted earlier, the abrasion distribution does not vary with
Appearance particle size while the crushing breakage is highly dependent on
Function particle size.
tLE * a LE + t HE * a HE
a= tLE + t HE (A9.6)
Breakage To predict a product from the mill contents and the appearance
Rates function requires only a rate of selection for breakage for each size
fraction of the mill load.
These rates will be inherently scaled because the mill load will be
constrained by mill dimensions and the mill diameter (if the energy
versus breakage assumptions are correct). These rates will
certainly vary if mill speed is changed but this dependence is not
included in the Leung model.
To describe these rates, a five knot spline function is used.
Best fit values to data are tabulated.
Spline knots ln (Rate ln (Rate
(mm) of Breakage) of Breakage)
Autogenous SAG
Classification The mill grate is modelled as a very simple classifier. When this
model was developed the relationship between the classification,
discharge and the operating conditions was not well defined.
Hence, the classifier/discharge is assumed to be constant- for other
than minus grate size hold up. A simple form is used.
Pebble Port Pebble port allows a small discharge rate of substantially coarser
Modification particles. This modification affects the classification curve as
shown below.
1.0
fp
x x x
m Size g p
d = dmax * D (A9.8)
Mass Transfer The value of dmax is adjusted until the model prediction matches
"Law" the required one. That is, until it lies on the operating line of
L = m1 Fm2 where
m1 = 0.37 (A9.9)
m2 = 0.37
L is the fraction of the active volume of the mill occupied by minus
grate size material and F is the total volumetric feed rate per minute
divided by the active volume of the mill.
Perfect Mixing The perfect mixing model at steady state provides the structure to
Mill Model combine the various components of the model. It relates the
different parts in the following manner.
i
fi - ri si + ∑ rj s j a ij - disi = 0 (A9.10)
j=1
pi = di * si (A9.11)
where fi, si, ri, di and pi are feed rate, contents, breakage rates,
discharge rates and product rate vectors and aij is the combined
appearance function.
The form of equations (A9.10) and (A9.11) allows both the mill
load and the product to be calculated for any mill load and
discharge rate adjusted until equation (A9.9) is satisfied.
Calculation sequence
↓
Calculate breakage rates
↓
Calculate volume of below grate
size material in the mill, L
↓
Calculate discharge rate
↓
If error is acceptable exit else
make correction to Dmax
Mill Load This model is unusual because it uses an internal port to describe
the mill contents. This port is accessible from the model properties
drop down or from the model window. It does not appear as stream
equipment.
Scaling This model is inherently scaled for mill diameter and volume. This
scaling optimistic in capacity as mill diameter is increased. It is
reasonable for mills of up to 8 to 9m diameter.
The Grinding The gross power draw of the mill is that drawn by the mill
Mill Power motor(s), ie metered power. It is assumed that this has two
components, viz
Net Power Draw From photographic evidence, the charge shapes shown in Figure
A9.2 were assumed to occur in grate discharge mills.
Grate Discharge
o
90
θS
rm
o
θ o
180 0
ri
θT
o
270
Power can be defined in terms of torque (τ) and rotational rate (N)
as follows:
Power = 2π Nτ (A9.14)
rm θ S
Pnet = 2π gLenρ ∫ ∫N
ri θ T
r r 2 cosθ dθ .dr (A9.15)
No-Load Power The no-load power draw (i.e. that drawn by the mill when
completely empty), is associated with various electrical and
mechanical energy losses. The main ones are motor, gearing and
bearing losses. None of these are fixed over the full mill operating
range. Some, however, may have a fixed component. For
example, bearing losses due to friction will be dictated by the mill's
dead weight (though even this will vary as liners and lifters wear),
and the mill charge weight which will clearly vary with grinding
condition.
1000
800
Predicted (kW)
600
400
200
0
0 200 400 600 800 1000
Indicated (kW)
Power Calculation The most recent JKMRC database currently includes power data
Accuracy from 63 different mills. Details are shown in Table A9.1.
The power model has been applied to this database and was found
to give excellent results. The standard deviation of the relative
error of the model was calculated to be 6.5% for gross power..
The model therefore requires a knowledge only of mill dimensions
and speed, ball charge, volume occupied by balls and pulp, and the
ore specific gravity. Full details of the model are given in Morrell
(1991).
Because of the industrial database, the prediction of gross power is
the most reliable.
Restrictions This power model assumes the SAG mill grate and pulp lifters do
not limit pulp throughput. For a large diameter mill (say > 7m) in
closed circuit with hydrocyclones or fine screens, this assumption
may not be justified. A build up of fine slurry in the mill will
remove some of the charge imbalance and reduce the actual power
draw.
A9.6 Symbols
Symbol Meaning
Operating The model is numerically stable at any mill load (equation (A9.9)).
Limits Real world mills typically operate with maximum loads of 30 to
35% by volume of charge.
Feed Sizing The auto/SAG model 'forms its load' from the mill feed. If the mill
feed size distribution is smooth (ie. a reasonably straight line on a
Rosin-Rammler plot), simulated variations in feed sizing give
sensible results. If the coarse end of the feed distribution is
artificially adjusted for the feed is preclassified in some way, then
the S20 assumption that the load can be treated as a single number
becomes unjustified. Hence artificially adjusted top sizes will
cause the model to predict wide variations in performance.
(While these variations are excessive, it should be noted that real
auto mills are also sensitive to feed top size).
Ball Size Effects Ball effect is estimated by generating an equivalent load of ore
particles. As the top 20% of this load is used to find S20, only the
top one or two ball sizes can have any 'impact' on this calculation.
Manipulating the finer ball sizes (ie. < half top size) have very little
effect. In practice, it does change the fine grinding rates.
Ball Load Effects These have been investigated in some detail at pilot scale. In
general, the harder the ore (low b and low ta) the less the grinding
rates are affected. A soft ore however follows the accepted wisdom
that increasing ball load will produce a coarser product. This may
well be because the increased number of balls are now breaking the
ore particles in the load which were doing the fine grinding.
Discharge Rates Considerable work has been carried out by Morrell (1990) on
factors affecting discharge rates. These effects are also
summarised in Morrell and Morrison (1989). See A11 for details.
Mill Liner Effects The SAG mill model is valid for correctly designed traditional
'high/low' lifter type action. Wave liners or short lifters do not
provide enough lift to achieve the default rates. If poor lift is
combined with poor discharge, the mill only produces abrasion
with a very fine product at a correspondingly low throughput.
Mill Load The autogenous and SAG mill model does not include an explicit
Limits maximum for the mill load. However, a warning will be flagged if
the total load (ie. balls and pulp) exceeds 35% by volume. An error
will be flagged if the total load exceeds 40%.
PARAMETER MENU
XG
XM
Rate 1
Rate 2
Rate 3
Rate 4
Rate 5
Initial Values Use the grate width and 100 µm as initial estimates for xg and xm.
The default breakage rates for auto and SAG will provide a good
guess for each knot value.
Ore Type For accurate results, these are best derived from tests carried out on
Parameters representative samples at JKTech.
Mill Load If a reasonable estimate of load mass and sizing is available, then
fitting with a range of A and b values may provide a way of
estimating these values - that is - use the values which give the best
fit for ratio work.
Closed Circuit If the mill is being operated in closed circuit with hydrocyclones, it
Operation is better to reduce m1 from 0.37 to 0.25. This seems to provide a
better approximation of the mass transfer response for a large
recirculation of material finer than grate size.
Knot Positions The spline knot positions are better left where they are for the
'normal' range of SAG mill feed sizings, 80mm < F80 < 250mm.
However for very fine auto mill feeds, the limiting size fraction
will also be finer and it may help to scale down all the knots. That
is, reduce them by the same ratio. An alternative is to simply fix the
larger knots at their default values.
A9.9 References
A10.1 Introduction
(Blank Page)
A11.1 Introduction
The Leung AG/SAG model (A9) typically requires a full scale plant or
pilot mill survey combined with ore breakage testing to generate a set of
grinding rates. However research in the mid 1990’s using a large
database of pilot and full scale milling tests has lead to the development
of a correlation between model grinding rates and mill operating
conditions. A further correlation between mill feed sizing and ore
breakage characteristics has also been developed. These two correlations
now allow mill performance to be predicted for a wide range of mill sizes
and operating conditions. Hence the model can be used to evaluate
optimisation strategies in existing plants and to investigate (and
compare) grinding circuit configurations at the pre-feasibility stage thus
reducing the cost of pilot testing.
For model fitting, the original and simulated cases will usually be
identical. This is considered in detail in section A11.6.
where
Jp = fractional slurry hold-up
D = mill diameter (m)
2
A = total area of the grate apertures (m )
φ = fraction of critical speed
3
Q = volumetric flowrate out of the mill (m /hr)
γ = mean relative radial position of the grate apertures
γ =
∑ ri ai
rm ∑ ai
Range
New Feed F80 (mm) 35-140
Ball load (%) 0-12
Recycle load (%) 0-500
No. different ores 16
No. tests 52
The breakage rate distribution is described within the model using cubic
splines (Ahlberg, 1967). This gives rise to five breakage rate values each
of which relate to a particular particle size and which together
characterise the entire breakage rate distribution. The five standard
particle sizes chosen are 0.25, 4, 16, 44 and 128mm which have
associated with them breakage rates which are labelled R1, R2, R3, R4
and R5 respectively.
1000
R5
Breakage rate (hr^-1)
100
R2 R4
R1
R3
10
1
0 0 1 10 100 1000
Size (mm)
These rate curves exhibit a characteristic shape. The coarser (R5 and
R4) rates relate to abrasive breakage while the finer rates R1 and R2
exhibit similar characteristics to those of coarse ball milling, ie.
predominantly impact breakage. The pronounced dip in the rates at R3 is
associated with the critical size which may limit mill throughput by
building up to excessive levels. Typically it is in the 25-75mm range and
where
Sa = rpm scaling factor
= Ln (simulated mill rpm/23.6)
Sb = fraction of critical speed scaling factor
= simulated mill fraction of critical speed/0.75
DB = ball diameter scaling factor
= Ln (simulated ball diameter/90)
JB = % of total mill volume occupied by balls and
associated voids
Rr = recycle ratio
= (tph recycled material_-20+4mm)
(tph new feed) + (tph recycled material -20+4mm)
F80 = 80% passing size of new feed (mm)
kij = regression coefficients
It can be seen from the equations that the finer size rates are functions of
the rates of the coarser sizes. Hence R1 is a function of R2 and R3 etc.
The rates can be considered as falling into 2 groups which represent the
grinding media and product size fractions. Hence the grinding media
group contains the rates R4 and R5 (related to particles >30mm) the
magnitude of which affect the throughput. The product group
incorporates rates R1, R2 and R3 (related to particles < 30mm) and the
magnitude of these affects the final product size. It is of particular note
that the rates are interrelated in a complex manner and are best
understood by graphing the entire breakage rate distribution.
Ball Load The effect of changing ball load on the breakage rate distribution is
illustrated in Figure A11.3.
10000
0% balls
1000
4% balls
Breakage rate (1/hr)
8% balls
100
10
1
0 1 10 100 1000
Size (mm)
oxidised, surface ore. As the ore becomes harder it may well be possible
to replace balls with ore as grinding media for more power effective
operation.
Makeup Ball No significant dependence of the breakage rates on ball size was found in
Size the pilot mill database. The SAG model does account for ball size
changes in terms of the energy provided during impact. It does this by
changing the mean grinding media-size, which in turn changes the
‘energy level‘ of the mill. This ‘energy-level‘ term is used to determine
the specific energy of impact. As the ball size is increased, therefore, the
specific energy increases and hence for a given impact event a finer
product size distribution occurs. However, as the ball size is increased
the number of grinding media per tonne of charge will decrease. As the
breakage rate is related to the number of impacts provided by the grinding
media then a reduction in the breakage rate may be expected to occur. To
account for this a ball scaling factor is used. Figure A11.4 illustrates the
effect of the ball size correction factor on the breakage rate distribution.
1000
100
Breakage rate (1/hr)
10
94mm balls
125mm balls
1
1
10
1000
0.01
100
0.1
Size (mm)
Feed Size F80 The effect of F80 was found to be the most difficult one to evaluate as it
Effects for SAG interacted with the ball charge level. At relatively high ball charges (10%
Milling or more) high F80 values were detrimental as evidenced by the reduction
in the breakage rates illustrated in Figure A11.5.
10000
f80=75;Jb=10%
1000 f80=125;Jb=10%
Breakage Rate (1/hr)
100
10
1
0 1 10 100 1000
Size (mm)
Feed Size F80 However in the case of autogenous grinding the pattern is different. In
Effects for this case a higher F80 promotes breakage in the coarser size fractions
Autogenous (Figure A11.6). This is to be expected when it is considered that in
Milling autogenous milling large rocks are required to break ore in the R5 size
range (128mm). As the F80 increases, this will typically result in more
coarse rocks in the charge able to break R5-size ore and hence R5 will
increase. In SAG mills running with higher ball charges, the rock
component of the grinding media plays a lesser role in dictating the
breakage rate and contributes more to the rock ‘burden’ which has to be
ground down. Feeds with F80 values and hence more coarse feed rocks,
can thus be expected to reduce the breakage rate.
Caution needs to be exercised, however, as it has been found that the F80
is not always a good indication of the feed size distribution. This is
particularly noticeable with autogenous mills whose performance may
fluctuate considerably yet maintain a reasonably constant F80. In such
cases the distribution changes systematically with performance and that
typically higher proportions of 25-50mm material in the feed result in
lower feedrates, ie. less sub-grate size material is present in the feed and
more near size material has to be broken.
1000
f80=75;Jb=0%
f80=125;Jb=0%
10
1
0 1 10 100 1000
Size (mm)
It has been found that the amount of recycled material in the -20 + 4 mm
size range is inversely related to the amount of breakage that this
material is subjected to. This can be explained if one considers that these
rocks are broken by coarser rocks and balls whose frequency does not
appreciably change with changes in recycle load. However as the
amount of recycled -20 + 4 mm rock increases, the amount of this size
material in the load will increase. As the breakage rate in a given size
class is related to the ratio of the number of coarser rocks and balls to the
number of rocks in the given size class, then increasing the -20 + 4 mm
recycle will result in a drop in the breakage rate in this size range (R3
size = 16 mm). The changes in the breakage rate distribution as the
coarser recycle increases is illustrated in Figure A11.7. Interestingly,
recycle of fine material ie. –4 mm did not correlate with any of the
breakage rates. This may be related to the breakage mode of this
material which is believed to be dominated by attrition.
1000
100
10 Rec=0
Rec=.05
Rec=.1
1
1
10
1000
100
0.1
Size (mm)
Recycle Control in For the most part, the Version 5 model is identical with the
V5 Variable Rates SAG model in Version 4. There are, however, a
couple of important differences which relate to control of recycle
of –20 +4mm material on the grinding rates.
As for V4, the User inputs new feed rate tonnes per hour and 80%
passing size for Simulated and “Original” mills.
Version 5 uses this implied switch as well i.e. the fixed recycle
tonnage value is set to zero to allow for simulated recycle.
Hence the User uses the Ore Feeder size marker to estimate % -
20mm and % -4mm and enters the difference into the appropriate
field on the SAG model.
Mill Speed/Mill The breakage rate is related to the number of size reduction events
Diameter per particle, per unit time and is hence a frequency. This in turn
must be related to the frequency with which the mill rotates (rpm).
A scaling factor is therefore applied to account for changes in the
rotational rate. For a given fraction of critical speed the rpm
decreases with mill diameter0.5 and hence this scaling factor will
also change with mill diameter. All else being equal, therefore, a
larger diameter mill will have a lower breakage rate than a smaller
unit. However it is pointed out that the JKMRC model inherently
scales on the basis of breakage energy which it relates to mill
diameter. Therefore, whereas a larger diameter mill will have a
lower breakage rate it will have a higher breakage energy.
In a given mill as the rpm changes, apart from the rotational rate,
the shape of the grinding charge will also change in line with the
fraction of critical speed (Morrell, 1996). Typically as the fraction
of critical speed increases the charge is subjected to increased lift
and hence impact breakage is enhanced. It is at the expense of
attrition breakage which is normally associated with cascading
motion and which is prevalent at lower speeds. To account for
these effects a further scaling factor is applied which is based on
the fraction of critical speed. Figure A11.8 illustrates the predicted
changes in the breakage rate distribution as speed is changed.
10000
85% Cs
75% Cs
1000
65% Cs
100
10
Size (mm)
Figure A11.8: Predicted Effect of Changing Speed on the Breakage
Rate Distribution
Mill Power The variable rates model allows the user to specify the conical slope
inside the liners of each mill end. The mill power estimate includes
the conical ends (Morrell, 1996).
Fitting Single Model fitting the variable rates model is quite similar to fitting the
Data Sets Leung model (A9). The defaults for the original mill grinding rates
are all set to zero, ie. the intercepts of the rate equations (Table
A11.2) are included in the model.
Hence the fitted rates indicate how far the measured mill is
operating from “typical” conditions. The recommend strategy is to
first fit xg and xm with the grinding rate intercepts set to zero. If
the mill has pebble ports, set the initial pebble port size to the
largest measured particle in the mill discharge. If the xg and xm fit
is plausible, add the pebble port size (PPSize). Use the measured
open areas for pebble ports and grates and the measured weighted
radius/mean relative radial position for the grates. Note that the
grate open area includes grates and pebble ports. The recycle
streams are selected from the unit menu.
The new feed size (F80) should be noted and entered for both
Sim(ulated) and Org(inal) mills as should all of the other measured
mill data.
If the xg, xm and PP (pebble port) size fit is plausible, adjust the
scale factors on Breakage Rate “Constants” to 0.1 and include them
in the next fit.
The open area fractions (Grate OA) can be selected to fit. They are
only suited to matching wear conditions and should not be fitted
together with grate or pebble port sizes as the parameters are likely
to interact quite severely.
Given good data and ore characterisation this model will often
predict the measured results quite well and model fitting is very
simple.
Fitting Multiple A comprehensive pilot test program will produce data over a range
Data Sets of operational conditions. For sophisticated users, the variable rates
model allows several sets of pilot data to be analysed
simultaneously.
The first step is to analyse each set by using its own select list. This
should identify any data problems. Then add each data set onto a
combined select list for master slave fitting.
One of the pilot data sets is selected as a base case. For this set,
simulated and original inputs are the same. For the other sets,
change the simulated mill conditions as required (eg. Ball load)
and use the base case original mill conditions in all tests. Add all
of the measured load and product streams to the model fit data list.
Use the master/slave capability to simultaneously fit, xg, xm, PPort
size and grinding rate intercepts to all data sets at once.
Notes:
• a fast computer (Pentium 166 or better) is required for three or
more SAG data sets
• Multi-fit capability is not available in Version 5. However, the
number of fittable sets of port data will be expanded in later
releases.
A poor overall fit, particularly if the grinding rates are lower than
typical (negative intercepts) may indicate shortcomings in data
collection. More seriously, it may also indicate more significant
problems such as poor liner design or inadequate pulp transport
capacity (i.e. pulp lifters).
Recycle Streams Up to three recycle streams can be selected from the model menu.
These should be recycles which actually go into the mill, eg.
Recycle crusher product, not feed. The “Fixed Recycle” input
should be set to zero for simulation to allow the calculated flow of
-20 +4mm to be used. (Input the measured flow for model fitting).
NB: V5 handles recycle loads differently from V4. See pages 108-
109 for details of the differences.
Load Limits The feed trunnion diameter indicates the maximum volumetric load
limit. If the simulated mill limits at a lower level than the actual
mill, reduce this diameter.
Grate Flow The flow correlation detailed in A11.3 provides a maximum flow
Limits (Mass estimate at the simulated mill load. The user may enter a design
Transfer maximum load level for which a maximum flowrate estimate is also
“Law”) calculated. These estimates relate to flow through the grate. They
assume that the pulp lifters can remove all of the grate discharge.
This is not always true for mills operating in closed circuit with
cyclone or fine screens.
If the simulated flow exceeds the maximum, the mill will likely fill
up with fines and go into overload as the slurry pool reduces impact
breakage. This effect is not simulated by the model.
Feed Size The F80 values for new feed for both simulated and original mills are
Considerations entered by the user.
For a design case, the F80 of the feed can be estimated from the
measured ta with a standard deviation of about 10% of the primary
crusher closed side setting.
100
10
Coarse/hard ore
Fine/soft ore
1
.01 .1 1 10 100 1000
Size (mm)
The size converter model (see Appendix A10) can be used to adjust
from a similar ore to the target range for simulation.
With the large database of SAG mill test work, it is clear that
maximum throughput does not always correspond to maximum
mill power draw or maximum mill load.
For hard ores, maximum throughput requires sufficient impact
energy at the toe of the charge. Hence the maximum throughput (at
maximum discharge coarseness) will often occur between 20 and
30% volume mill load.
This F80 value should be used as the Reference F80 value on the
Recycles tab in the Variable Rates SAG Model equipment window.
A11.10 References
Morrell S & Morrison R D, 1989. Ore Charge, Ball Load and Material
Flow Effects on an Energy Based SAG Mill Model. Presented SAG
1989, University of British Columbia. Editors. Mular & Agar.
(Blank Page)
A12.1 Introduction
Pre-crusher If particles are bigger than a certain critical size they will be broken
directly by the roll faces as would occur in a conventional rolls
crusher. The breakage in this zone can be considered as analogous
to a ‘pre-crusher’, the products from which may subsequently pass
to a region where a bed under compression has formed. The
boundary between the pre-crusher and bed compression regions is
defined by a critical gap (xc).
Edge Effect Breakage at the edge of the rolls is different to that at the centre and
Crusher conforms more to that experienced in a conventional rolls crusher.
This is the so-called ‘edge effect’ which defines the proportion of
relatively coarse particles usually seen in HPGR products. Its
existence has been explained by the pressure gradient across the
width of the roll and the static confinement of the ore at the edges
of the rolls which the cheek-plates provide.
Compressive At some point away from the edges of the rolls, and extending
Bed Crusher upwards from the area of minimum gap (xg) to an area bounded by
the critical gap (xc), is a compression zone where breakage
conditions are similar to those experienced in a compressed packed
bed.
then combine with feed particles which are smaller than xc. A
proportion is then diverted to another single particle crusher stage
where all particles greater than the minimum gap (xg) are crushed
to below this size. The remainder are diverted to a compression
stage where all particles greater than xg are crushed below this size
but in a compressed bed mode.
Precrusher:
conventional rolls crusher;
gap = x c
Combiner
HPGR Model The model contains three breakage processes and one splitting
process between the edge and compressed bed zones. For the
breakage processes the JKSimMet crusher model is used to
describe the size reduction. Four model parameters are required for
each breakage process: K1, K2 and K3 and t10. The first three are
used to describe the probability that a particle will be broken whilst
the t10 is used to describe the product size distribution that results.
For a detailed model description, refer to Appendix 6.3.
t10 Definition The t10 is defined as the percentage passing one tenth of the
original particle size in the product after breakage. Other tn
parameters can be similarly obtained from a product size
distribution, eg. t2 is the percentage passing one half of the original
particle size. From breakage tests the t10 and a number of other tn
values are determined from the breakage products. These values
are stored in tabular form in the model which, given a value of t10,
uses spline interpolation to determine the associated tn values and
hence reconstructs the entire product size distribution.
0.5
4 ρ g Dx g
x c = 0.5 {(D + x g ) (D + x g ) 2 - } (A12.1)
ρc
xc
αc D
xg
Compressed Bed In the compressed bed crushing zone, on the other hand, size
Breakage reduction is assumed to be similar to that experienced by a bed of
particles in a piston press. The parameters used to describe size
reduction are determined from tests in a laboratory or pilot scale
HPGR machine combined with breakage tests in a piston press.
The piston press tests provide information on the relationship
between size reduction and energy input in a compressed bed.
They also provide a description of the characteristic shape of the
product size distribution. If the piston press tests are not available,
then the results from the single particle Drop Weight test may be
used to determine the Compressed Bed Breakage Function (Section
A12.4)
Edge Crushing The last sub-process in the model is the split to the edge and
Bypass compressed bed zones. The edge zones are associated with the
drop in pressure that is experienced towards the edge of the rolls.
Their extent is assumed to be a function of the working gap. The
fraction of feed which is crushed in the edge zones (f) can therefore
be expressed as:
xg
f = g L (A12.2)
where g is split factor and L is the roll width. Using pilot scale
HPGR test results where sizing data of both pure flake and total
product were available, the split factor g was found to be
approximately constant with a value of 3.4. In physical terms this
means that the edge effect zone extended from the edge of the roll a
distance equivalent to 1.7 times that of the working gap. By sizing
the pure flake and total products from lab/pilot test results f can be
determined experimentally. Recent work suggests that the fraction
of material being subjected to edge crushing is usually about 10%.
Thus, the model may be simplified by manipulating g (split factor)
to ensure that 10% of the feed reports to the edge crushing zone.
The resulting products are sized and fitted to a spline surface. This
surface can be regenerated by the model from a matrix of spline function
values. These values are input to the model as
It can be clearly seen that these breakage models are different. The
power requirements can also be characterised with particle size
dependence if required and also related to motor power (Section A12.6).
A12.5 Throughput
Throughput is controlled principally by roll dimensions, speed and
profile, and material characteristics such as size hardness and
particle-roll friction (and thus nip-angle). The profile and material
of the roll surface is important in controlling both wear and
machine performance, and various options are offered by the
different manufacturers.
where
It is realised that A12.3 does not take into account the slip between
feed material and the rolls surface, nor does the feed characteristics
(particle size and size distribution, moisture, etc). Figure A12.3
shows the deviation between the measured throughput and the
calculated one using Equation A12.1 for Primary diamondiferous
ore treated through a 100 mm Polysius laboratory scale HPGR. It
is obvious that Equation A12.3 over-predicts the HPGR throughput
at high rolls speed, which may indicate that slip exists in the HPGR
operation at these speeds.
12
0.38 m/s
10 1.50 m/s
2.50 m/s
3.10 m/s
8
tph (measured)
6
0
0 2 4 6 8 10 12
tph (calculated)
To correct for the slip effect it is considered that for a specific feed
the slip is a function of the rolls speed and the dimensionless
working gap which is defined as xg /D, where D is the rolls
diameter.
Qm
Figure A12.4 plots the correction factor c (c = Q , where Qm is
c
the measured throughput and Qc is the calculated by Equation
A12.3) versus the product of the speed and the dimensionless gap
xg
(U* D ) for the Diamondiferous ore using the laboratory HPGR
data. A linear regression on the plot was obtained and Equation
A12.3 was accordingly modified as:
Q = 3600 U L xg ρg c (A12.4)
2.0
1.5
c = Qm / Qc
1.0
0.5
0 0.01 0.02 0.03 0.04 0.05 0.06
U * (Xg / D)
15
LAB (D = 0.25 m)
KHD (D = 0.80 m)
POLYSIUS (D = 0.71 m)
10
Q (predicted tph)
0
0 5 10 15
Q (measured tph)
Figure A12.6: Prediction of Throughput for Two Pilot Scale HPGRs from
Equation A12.4 with Model Parameter c Calibrated Using Laboratory
Scale HPGR data
HPGR Crusher In the size reduction model the two parameters K3 and t10 were
Power fitted to the laboratory scale HPGR power data. It was found that
the fitted t10s for 24 sets of Diamondiferous ore tests under various
rolls speeds and feed size conditions fell on a t10 - Ecs master
curve, as shown in Figure A12.7
100
A = 100, b = 0.2084
80
60
Fitted T10 (%)
40
9.5 mm feed, 0.38 m/s speed
9.5 mm feed, 1.50 m/s speed
9.5 mm feed, 2.50 m/s speed
20 9.5 mm feed, 3.10 m/s speed
6.7 mm feed, 0.38 m/s speed
6.7 mm feed, 3.10 m/s speed
A,b fitted to Lab
0
0 2 4 6 8 10
Ecs of motors (kWh/t)
Figure A12.7: The Fitted t10 vs Specific Energy Ecs for Diamondiferous
Ore Treated through a Laboratory HPGR
4.0
3.0
2.5
2.0
0 2 4 6 8 10
Power Draw vs The prediction of the working gap xg is also required for simulation.
Working Gap The working gap depends on pressure and power draw.
2
0 2 4 6 8 10
Figure A12.9: Relationship between Working Gap and Specific Energy for
Diamondiferous Ore Treated Through a Laboratory Machine
SpFact
K1H
T10H
Power Coeff. H
Split Factor (g) This factor determines the proportion of material which is crushed
(SpFact) in bypass mode. This is usually 1.7 times the effective gap width
on each side for a default value of 3.4. (Section A12.3)
Pre and Edge Using the same feed material as for the pilot/lab HPGR test,
Crusher Model laboratory roll crusher is operated at close to the nipping gap and
Parameters the working gap of the HPGR.
(K1H & T10H)
The Whiten/Awachie/Anderson crusher model (Appendix 6) is
used to derive K1 and t10 where K2 is the crusher gap and K3 is set
at 2.3.
The ratio K1/K2 is the input to the pre crush and edge effect model
along with the fitted t10 values. It is unlikely that power can be
measured with sufficient accuracy in this test to justify using other
than the default power factor of 1.25.
The model defaults are for smooth rolls. It is highly likely that
different roll surfaces will generate different correction factors.
Compressed Bed Breakage within the compressed bed is assumed to be uniform and
Breakage (t10 able to be described by a single parameter t10 . The t10 parameter
HPGR) will increase as the reduction ratio increases. In compression, all
particles are assumed to be able to be selected for breakage ie. K1 =
0 and every particle larger than the working gap will always be
broken, ie. K2=cacluated Working Gap.
Power Model The HPGR model takes a somewhat circuitous approach to power
Fitting modelling. As noted in Section 12.6, the combination of piston
press tests and laboratory/pilot scale HPGR produces a relationship
between the compressive bed t10 and net motor power (Figure
A12.7).
To predict the performance of pilot scale and full scale HPGRs the
model is firstly calibrated using the results from the laboratory,
conventional rolls, single particle breakage and piston bed
breakage test. Figure A12.10 illustrates the scale-up procedures.
Also shown in Figure A12.10 are the values of the parameters
obtained from the calibration, which have been used to predict the
two pilot scale units and one full size machines treating a
Diamondiferous ore (Morrell, Shi and Tondo, 1997).
Input data
Rolls dimension: diameter, length
Select rolls speed, required specific energy Ecs
Feed size distribution
Bulk density of feed, flake density
Working gap: chose the lab working gap by Ecs from Figure 12.9,
multiplying the gap by the ratio of full scale to lab rolls diameters
Nipping gap calculated from Equation 12.1
Throughput calculated from Equation 12.4
Power draw = Ecs required x throughput
Pre-crusher
Single particle (parameters from the conventional rolls
breakage test test)
(using a drop weight K1p=0.64 K2p
device) K2p = nipping gap
K3p = 1.0
K3p = 1.0
t10p = 12.04
Mass Splitter
Fraction split to the edge effect
crusher is calculated by Equation
12.2 in which γ = 3.4
as determined from the KHD tests
Combined Product
Roll Surface Tests using a Krupp Polysius pilot roll (rolls diameter 0.71 m), with
4 mm profiles (on the rolls) resulted in a considerably larger
working gap than was observed for the KHD pilot tests using
smooth rolls. Therefore, laboratory tests must be conducted with a
rolls surface similar to that proposed on the full scale machine.
Limited Data As only limited production scale data were available, the models
Base need to be further tested and validated against more real data in the
future, and their capabilities explored in case studies.
Power Coefficient Ideally, this coefficient should be constant. A better understanding and
(kp) (possibly) a better representation need to be developed.
A12.11 Nomenclature
A12.12 Acknowledgments
A12.13 References
SchÖnert K.and Sander, U., 2000. Pressure and shear on the roller
surfaces of high pressure roller mills, Proc. XXI IMPC, Rome,
Italy, Sect A4, 97 - 103.
(Blank Page)
A13.1 Introduction
The concept of a degradation model has its origins in iron ore and
coal operations where particles may undergo significant size
reduction during mechanical handling such as dropping on to a
stock pile from a conveyor or perhaps at a conveyor transfer
point.
Where:
Ecs = specific comminution energy (kWh/t)
h = height of the drop (m)
Where:
t10 = Breakage Distribution Parameter
Ecs = specific comminution energy (kWh/t)
A & b are ore characteristic parameters
Conditioning In most cases, the damage inflicted by a second drop is less than
that inflicted by the first drop. This effect is known as
conditioning. Of course, the height of each drop is important as
well as the number of drops.
Example The A and b values from the JKMRC Drop Weight test for the
example ore are 50 and 0.5 respectively.
Ecs = 0.00272 * h
= 0.00272 * 20
= 0.054 kWh/t
this value of t10 is then entered into the model for the second drop.
Use for the The degradation model can be used to represent a lightly loaded
Vertical Shaft Vertical Shaft Impactor. In this case, the energy of an impact is
Impactor calculated from the velocity of the particle imparted by the rotor.
This energy must be converted to units of kWh/t before equation
A13.2 can be applied.
A14.1 Introduction
APPENDIX B
Error Messages
B. ERROR MESSAGES
These messages occur during operation of JKSimMet. The display
warns that an error has occurred and provides the error number.
The descriptions provided here give more information about the
possible cause of the error message.
ERROR 58 Not enough size distribution data in the feed to an equipment item
for spline interpolation to work. Check combiner ports with Exp
TPH Solids values > 0.0 with limited or no Exp Size Distribution
data.
One of the combiner ports of one of the equipment items on the
select list has Exp TPH Solids greater than zero but limited or no
size distribution data. Thus JKSimMet is not able to perform the
required Spline Interpolation. Either add some size distribution
information or zero the Exp TPH Solids.
ERROR 121 Two or more streams have different sieve series. Please correct.
For Mass Balancing with GSIM format data, all selected streams
must have the same screen series.
ERROR 122 Sum of % component does not equal control value. Please
correct.
If you have specified a component sum (on the active
COMPONENT LIST) your assays must total this sum or less. A
component sum of zero turns off this constraint. Normally a
REMAINDER TERM is added to achieve the sum in the
experimental data. To force a constraint, omit one of your
categories and it will be the remainder term.
ERROR 126 No stream input to unit. Please check current SELECT list.
One of the units selected for inclusion in the balance has no input
stream. You should check that input streams are selected for all of
the selected units.
ERROR 127 No stream output from unit. Please check current SELECT list.
One of the units selected for inclusion in the balance has no output
stream. You should check that output streams are selected for all of
the selected units.
ERROR 128 Stream is not connected to unit. Please check current SELECT list.
There are no units selected on the SELECT list. At least one unit
and its associated streams must be selected to run a mass balance.
ERROR 130 Morrison solution error. See Morrison solution error Section 6.10.
The simple solution has not worked correctly. Check your data
carefully and then read section 6.10. If you can find no data
problems, try increasing the number of steps. Note that only one
flow rate should be tightly constrained - not all of them.
ERROR 131 Morrison solution convergence error. Please increase step number
ERROR 132 Consistent assays convergence error. Please increase step number.
Check the SELECT screen to see that the step count has reached its
nominated maximum. If so you may increase that count by a few
steps. Caution: such long searches indicate poorly defined flows
or some variation of the middlings problem. Refer to section 6.10.
ERROR 133 Adjusting sum convergence error. Please increase step number.
ERROR 134 Main loop convergence error. Please increase step number.
ERROR 136 Balance size distributions cal. err. Increase adj. sum step number.
WARNING 143 Model-Fit data do not include this stream. Normal editing only.
The Stream selected is not included for Model-Fit. Therefore, the
extra information cannot be edited.
ERROR 144 Model-Fit data do not contain any streams. Fitting requires data.
No streams are selected for Model-Fit. The list of streams, whose
data must be fitted, is empty.
ERROR 146 No streams in the circuit have data. Please add some.
None of the Streams selected for the Model-Fit have data entered
for them. At least one stream must have data. See ERROR 51.
ERROR 147 No parameters are selected for fitting. Please select some.
Model fitting works by adjusting parameters of models until
simulated results match experimental data. You must specify at
least one parameter to adjust.
ERROR 148 There are too many streams on the circuit. Please simplify!
Please simplify the circuit or break it into two circuits, The Model-
Fit function has strict limits on the number of units and streams
allowed. Please reduce the number you have selected, then try
again. Refer to the manual for the current limits (there is a limit of
10 units, 20 streams).
ERROR 149 There are too many units on the circuit for fitting. Please simplify.
The Model-Fit function has strict limits on the number of units and
streams allowed. Please reduce the number you have selected, then
try again. Refer to the manual for the current limits (there is a limit
of 10 units, 20 streams).
ERROR 152 Fitting is not getting anywhere. Try again with better guesses.
The Model-Fit has not been able to make effective use of the data
given. Please enter new parameter guesses and try again.
ERROR 154 No errors were calculated. Only non zero SDs contribute an error.
WARNING 155 No SDs have been entered on a stream. Unit SDs are assumed.
See ERROR 154.
ERROR 156 You have duplicate data entries. Please remove duplicate.
If you want one stream to have a greater significance reduce its
SDs.
ERROR 157 You have duplicate parameter entries. Please remove duplicate.
Each parameter entry is independently adjusted. This becomes
nonsense if one parameter is repeated.
ERROR 158 Only two parameters may be fitted per stream. Please correct.
There are only two independent parameters for a stream. See
ERROR 157.
ERROR 159 Only one water parameter may be fitted per stream. Please correct.
Once stream's water parameters are independent. See ERROR 157.
All water parameters control the water content of a stream. There
is only one way this may be selected. See Error 157.
ERROR 164 That unit doesn't have experimental data suitable for model-fitting.
The unit selected has no parameters and can not be fitted.
ERROR 165 No units in the circuit have data. Please add some.
None of the units in the circuit have data. You must supply some.
WARNING 167 The coarsest particles in the feed to the AG/SAG mill are in a size
range either finer than 200 mm or coarser than 300 mm. This
affects the calculation of energy values and the results of
simulations using this feed are likely to be unreliable. Please
modify your feed size distribution.
The energy calculations in the Variable Rates SAG model were
based on data from mills with the top size of the feed in the region
200 – 300 mm. Simulating with feeds outside this region using the
default rates will be unreliable.
APPENDIX C
JK BREAKAGE
TESTING
C JK Breakage Testing
Perspex
enclosure
5kg lead
weights
Guide rail
Adjustable
height (energy)
Rock
Steel anvil
Where:
Ei = energy used for breakage
M = drop-weight mass
g = gravitational constant
h = initial height of the drop-weight above the anvil
xM = final height of the drop-weight above the anvil.
where:
Eis = specific input energy
Ecs = specific comminution energy
m = mean particle mass
In the manner described above, a set of t10 and Ecs values are
produced for the 15 energy/size combinations.