MIME 413 – Strategic Mine Planning with UncertaintyMIME 513 – Mine Planning Optimization under Uncertainty Lecturer: R. Dimitrakopoulos Fall 2014 Class Project – Part 1 Optimization in Mine Design using Whittle Software Due date: September 26th 2014 Early bird submission bonus – by September 23st 2014: +10% Late submission penalty: 10% per day OUTLINE 1. Introduction 2. Task 3. Data Files 4. Instructions Page 1 MIME 413 – Strategic Mine Planning with Uncertainty MIME 513 – Mine Planning Optimization under Uncertainty Lecturer: R. Dimitrakopoulos Fall 2014 1. Introduction Part 1 of the project covers the following topics: 1. Definition of the ultimate pit limit. 2. Determining nested pit shells. 3. Designing pushbacks. 4. Effects of extraction sequence of ore/waste blocks on production scheduling and long term planning. Concepts behind Whittle Nested Pits Whittle uses a Model File containing details of the contents of each block, but, for optimization purposes, we need a single value for each block. This value is the cash flow (positive or negative) that would result from mining the block. For optimization purposes it is important to assume that the block has been uncovered. It is unnecessary to allow for stripping costs, because the optimizer does this implicitly. If we calculate the block values for a particular model we will get a certain set of block values that, when used in a pit optimization, will lead to a particular pit outline, the optimal outline. Example 1 For the purpose of this example, assume the pit outline is outline "A" in the following diagram. Figure 1. Pit outline obtained from a single optimization. Within an optimal outline, every block is "worth mining". Each block consists of zero or more parcels and, possibly, some undefined rock that can be regarded as another parcel. Page 2 MIME 413 – Strategic Mine Planning with Uncertainty MIME 513 – Mine Planning Optimization under Uncertainty Lecturer: R. Example 2 In addition. No block value will go down. typically.) and the associated prices and costs with the method of treatment. Consequently. and this is. the new outline (shown as outline B below) is certain to include the whole of A. large tonnage differences between adjacent pits can occur. It multiplies all of the prices by a series of 50 to 100 "Revenue Factors" ranging.3 to 2. An increase in all commodity prices while keeping the costs constant may cause the parcel to be treated differently (e. but which are now worth mining. the value of every block within outline A will increase or stay the same. doing an optimization for each. Increasing the prices will not decrease the cash flow obtained from mining a given parcel. in effect. Thus.e. in this case. The Whittle Pit Shells node (Optimization program. from 0. FXOP) If we step the prices through a series of values. it may now be processed rather than treated as waste). A parcel’s cash flow depends on the way we treat the parcel (i. Figure 2. Pit outlines obtained from two optimizations. what the Pit Shells node in Whittle does. The outlines are often very close together and form an almost continuous "spectrum". every block within outline A is still worth mining.0. etc. However. It may also include extra blocks that were not worth mining before. if the grade increases sharply with depth or the ore body is discontinuous. Since all the outlines obey the pit Page 3 . mill.g. the cash flow will always either stay the same or increase. The reason for producing outlines for the smaller values of Revenue Factor is that we want to produce inner pit shells that highlight the best positions to start mining and to assist with the sequencing. where the change in tonnage from one outline to the next is quite small. Dimitrakopoulos Fall 2014 The resultant cash flow from mining a block consists of the sum of the cash flows generated be mining the block’s parcels. we obtain a set of nested pit outlines. leach. if we do another optimization using the new values. if we increase the prices. and produces a pit outline for each factor. d. g. pit number 20 from one run may be very similar to pit number 25 from the previous run. e. It is therefore reasonable to simulate mining with wide ranges of prices and costs using the same set of nested pits. b. it is simple to determine which sequences are feasible when mining out a particular pit. Once a set of costs has been settled on. k. Regardless. grades and cash flows. b. given a set of nested pits. Dimitrakopoulos Fall 2014 slope requirements. g. if pit 5 is the ultimate pit. Page 4 . and thus can be used to show various mining schedules in order to obtain projected tonnages. but are shifted relative to those from the first run. f. another possible sequence is a. a final optimization using those costs and a repeated simulation can be run as verification. For example. f. etc. If the pits are sufficiently far apart to give working space. c.MIME 413 – Strategic Mine Planning with Uncertainty MIME 513 – Mine Planning Optimization under Uncertainty Lecturer: R. Although each set of outlines is only strictly optimal for a particular set of costs. Schematic of nested pits. their usefulness goes far beyond this. In Figure 3. c. Provided that the costs used are of the same order. these sequences are clearly defined and easy for the computer to trace. Example 3 Figure 3. h. we can clearly mine in the sequence a. etc. another optimization run with different costs will usually produce a set of pits of similar shape. the Whittle software must be used along with the parameters provided in Table 1 and the orebody models in the files described below.8 $/tonne Processing cost – Fresh Ore 16. Find the optimal ultimate pit limits. design pushbacks and generate a Life-of-Mine (LOM) production schedule for a small gold deposit. Slope angle 54 ° Selling price 19. for the mine’s yearly pit design and mine planning study.MIME 413 – Strategic Mine Planning with Uncertainty MIME 513 – Mine Planning Optimization under Uncertainty Lecturer: R. Project (Part 1) Task & Deliverables Deliverable for Project Part 1: A properly documented report that amalgamates the deliverables from each of the three Steps below. refer to the following sections.2 Mt/y Extraction capacity 0.3 g/tonne Model Au unit Grams Processing capacity 0.195 $/tonne Recovery – Fresh Ore 84 % Recovery – Oxide Ore 90 % Cut-off 0. Parameters for the optimization of the gold deposit.5 Mt/y Discount rate per period 8% Page 5 . so as to maximize its Net Present Value. Report all findings and justify your choices. for information on the related files and steps.29 $/unit Selling cost N/A Mining cost 1. For this task. All deliverables for this assignment are discussed in this section. Table 1. Dimitrakopoulos Fall 2014 2.862 $/tonne Processing cost – Oxide Ore 8. Dimitrakopoulos Fall 2014 Page 6 .MIME 413 – Strategic Mine Planning with Uncertainty MIME 513 – Mine Planning Optimization under Uncertainty Lecturer: R. explain the differences and related effects. c) Discuss the implications of the number of pushbacks used and what they physically represent. d) Plot the required graphs using bar style. waste. where the pit number is plotted on the X axis and the discounted economic value on the Y axis. b) Report in pushback-by-pushback graphs ore. where the pit number is plotted on the X axis and the net economic value on the Y axis. Page 7 . d) All choices must be justified and implications briefly explained. Milawa Balance and fixed lead. Deliverables from Step 1: a) A pit-by-pit graph of undiscounted cash flow per pit shell. metal production. use the automatic pushback selection in Whittle to choose the pushback design. b) A pit-by-pit graph of discounted cash flow per pit. Step 3: Generate the LOM production schedule of the mine Using the ultimate pit and pushback design from Steps 1 and 2. b) Describe the differences among the available schedule generating options: Milawa NPV.MIME 413 – Strategic Mine Planning with Uncertainty MIME 513 – Mine Planning Optimization under Uncertainty Lecturer: R. Dimitrakopoulos Fall 2014 Step 1: Defining the ultimate pit limit Define the final pit by choosing amongst the available nested pit shells generated by Whittle. Deliverables from Step 2: a) A table that associates a pushback design to its NPV. ore and waste where the pit number is plotted on the X axis and the metal on the Y axis. c) All choices must be justified and implications briefly explained. e) All choices made need to be justified and implications briefly explained. metal and cash flows. and cumulative NPV) for the mine. Step 2: Design pushbacks given the ultimate pit limit in Step 1 Using the final pit as defined in Step 1. including the selection of the scheduling option. Deliverables from Step 3: a) Produce the LOM production schedule and report all related graphs (ore. c) Pit-by-pit graph of recovered metal. generate the LOM production schedule. Use the same processing and extraction capacities as in Table 1. waste. Try at least three different and meaningful numbers of pushbacks and assess the possible implications on the final NPV of the mine. MIME 413 – Strategic Mine Planning with Uncertainty MIME 513 – Mine Planning Optimization under Uncertainty Lecturer: R. Dimitrakopoulos Fall 2014 Page 8 . “the second line” is written for each of the parcels for that block. Page 9 . Figure 4. If the number of the parcels is 1 or more. Processing CAF. “the second line” doesn’t exist for that block. Format: The formats of the parameter and model files are provided in the Whittle software’s manual and the summarised below. A) Model File: Orebody block model files: These types of files are identified by their extension as “mod. Y. ore and waste tonnage etc. Positional mining CAF. Dimitrakopoulos Fall 2014 3. metal grade. The first line: X. Z. Tonnage.” They contain the specification of the block model of the deposit such as block coordinates. deterministic) orebody model and a parameter file. Z. Data Files The folder named “MIME 413 Project Part 1” contains a single estimated (kriged.MIME 413 – Strategic Mine Planning with Uncertainty MIME 513 – Mine Planning Optimization under Uncertainty Lecturer: R. Metal Content If the number of the parcel for a block is zero. tonnage The second line: X. Example of a mod file when opened in a text editor. Number of Parcels. Y. Rock Type. processing CAF(1=use CAF. restart interval and restart time] 4 1 nx 1 ny 1 nz (sub-region block limits) 5 1 30 0.000 1. total ton. pit slope) 12 4 0 4 4 0 $ (decimal places to write: block ton.MIME 413 – Strategic Mine Planning with Uncertainty MIME 513 – Mine Planning Optimization under Uncertainty Lecturer: R. all the necessary parameters for the design and planning of the pit are provided. For more information. 0:quantity of unprocessed material not printed). dilution factor. number of benches to generate structure arcs. currencycharacter) 13 1.800 1 (general block tonnage.1657506 (revenue factor as single number or range start-step size-end) Page 10 . see the “Whittle Parameters” document included or the Whittle help manual. The parameter file: parfile. printing index (1:quantity of unprocessed material printed.par Format: The format of the parameter file is summarised as below. 0=do not use). reference mining cost.0 (number of slope regions. y and z directions and the origin of model) 2 nx ny nz (number of blocks along x. 1 dx dy dz x0 y0 z0 (block dimensions along x.0 (slope bearing angle. air flag B=air blocks not included in the result file (1) or air blocks within the ultimate pit are included (2)-Air flag A must be 1. y and z directions) 3 ABI MCAF PCAF PRNT RSTINT RSTME [active block indicator. Dimitrakopoulos Fall 2014 B) Parameter File In this workshop. default block tonnage) 6 0. revenue factor.000 1 2 1.] air flag A=consider air blocks in optimization(1) or do not (2). the selling cost ratio. or all air blocks in the model file are included (3).0 54. not used. positional mining CAF(1=use CAF. 0=do not use). ore selection by cutoff (1) or by cash flow (2)) 14 0. currency total. [recovery. Page 11 . mining CAF. Task Instructions The computer session of this first part focuses on the basics of using the Whittle software. price/unit) 21 OXOR 1.900 0 0. processing cost/ton) 26 GOLD C 0 0. The estimated orebody model will be used as input and the resulting schedules and pushbacks obtained in the first part are referred to as the “base case” during analysis performed in subsequent assignments. rock type.000 0 1. processing throughput factor—speed of processor for a particular rock type) 25 MILL OXOR 8. cutoff controlled (C) or not controlled (N). process recovery fraction. position in the file. encouraged to explore these features if desired. minimum and maximum grade). recovery threshold. such as varying commodity price and processing capacity are also included in the notes.000 (rock type. This section covers importing data into Whittle. selling cost/unit. however. This is only useful to those who have not used Whittle previously. The data set used in this assignment is generated from an estimation method (ordinary kriging).195 (processing method. grades and cut-offs for this element. designing pushbacks from pit shells and optimization with the Milawa-NPV option.2900 (element. processing cost/unit. decimal places for gold in the block: total unit of element.300 0 (element type. Some of the traditional sensitivity analysis. but is not necessary for this assignment. generating pit shells and ultimate pit limits. 20 GOLD 0 19. rehabilitation cost/ton. You are.MIME 413 – Strategic Mine Planning with Uncertainty MIME 513 – Mine Planning Optimization under Uncertainty Lecturer: R. 4. Dimitrakopoulos 18 GOLD 1 4 4 Fall 2014 4 (element type. select the project directory by clicking on the directory icon on the right side . Make a new directory in your personal network drive (P:\ drive) called “Project Part 1”. c. Type project name: Project1. If there are any additional windows that appear asking for parameters. Dimitrakopoulos Fall 2014 1 Import Deposit Data in to Whittle a. Copy the files: ‘parfile. for import type.mod” as the model file to import.P:\Project1\working_Project1. b.par” as the parameter file to import. Then. Page 12 .par and kriging. Click on create a new project and press OK.P:\Project1\. Browse the folder you created to find “kriging. Open Whittle and import parfile. choose “Whittle block model”. click on finish and “yes” to confirm.mod’ from the folder provided into the “Project Part 1” directory you just created. Whittle should automatically create a “working” directory . simply keep selecting “Next” or “Finish” until they go away.MIME 413 – Strategic Mine Planning with Uncertainty MIME 513 – Mine Planning Optimization under Uncertainty Lecturer: R. e. Click on “next”. and “parfile.par’ and ‘kriging.mod files as follows: d. Under “New Block Model”. A green check mark should appear beside “New Block Model”. select the “Description” tab and change the block model’s description from “New Block Model” to “Kriged Model”. h. Check the parameters loaded and run the program by clicking on the third man icon . Page 13 . g. click on each of the tabs and check what information is provided. i. Press the “Accept” button to accept the changes. Now. j. Not setting the correct units in Whittle can lead to misleading results. k. Dimitrakopoulos Fall 2014 f. When finished. Under the “Formats” tab. set the “Units” for Element GOLD to “gram”.MIME 413 – Strategic Mine Planning with Uncertainty MIME 513 – Mine Planning Optimization under Uncertainty Lecturer: R. select the “Check Data” button to ensure that all data has been correctly entered. ) When finished. All of the parameters for this node should have been read from the parameter file but you should verify them all.MIME 413 – Strategic Mine Planning with Uncertainty MIME 513 – Mine Planning Optimization under Uncertainty Lecturer: R. Click on “New Pit Shells” on the left and description tab on the right. grades expressed as decimals. Type “Kriging Slope Set” for the description and click on “Accept” keeping all the values as default. Type “Kriging Pit Shells” for the description. press “Accept”. Dimitrakopoulos Fall 2014 2 Generate Ultimate Pit Limits a. as per the assignment requirements (Section 2). Page 14 . it is important to be mindful of your units (% recovery. etc. b. Once again. Click on “New Slope Set” on the left and description tab on the right. units for selling on market. Click on “Profile” tab and check the parameters to ensure that the slope angles are correct. Dimitrakopoulos Fall 2014 c. and choose “Open Pit Page 15 . Then add a new value to be displayed by clicking in ‘add’ select “Output” on the left and select “Undiscounted revenue and cash flow” on the right. Additionally. ensure that “fixed lead” is selected with “0” value. click on “accept”. Click on “description” tab and type “Pit by Pit Graph-1”. d.” Click on “time cost” tab and change the discount rate. Then. Click on description tab of the “New Operational Scenario” node and type “Operational Scenario-1. ensure that the units for “Element limits” is set to grams. e. delete all the lines below “tonnage of waste rock”. In the “Definition” tab. Add Pit by Pit Graph: Right click on “Operational Scenario-1” and add “Pit by Pit Graph”. Then.MIME 413 – Strategic Mine Planning with Uncertainty MIME 513 – Mine Planning Optimization under Uncertainty Lecturer: R. Repeat for the “Pit by Pit Graph” node. Right-click on “New Schedule Graph” on the left side and choose “Cut Branch” from the menu. On schedule tab. Click on the “limits” tab on the right side and change the value of the extraction and processing limits. Dimitrakopoulos Fall 2014 Value”. click on OK. click on OK and accept it by clicking “accept” in the general window. You will likely need to explore the various options to be able to get the correct graphs. Then.MIME 413 – Strategic Mine Planning with Uncertainty MIME 513 – Mine Planning Optimization under Uncertainty Lecturer: R. Click on add to selection list and best case. In order to generate various graphs (as outlined in the assignment deliverables). Do not hesitate to include a new graph not outlined in the assignment deliverables if you believe it helps you justify your points in your report’s discussion. Page 16 . Then. you will need to repeat this step and select the correct properties to save. run the program by clicking the running man icon. MIME 413 – Strategic Mine Planning with Uncertainty MIME 513 – Mine Planning Optimization under Uncertainty Lecturer: R. Dimitrakopoulos Fall 2014 f. Select the directory you created and change the name to Page 17 . Generating the pit list: Right click on the Kriging Pit Shells and select “Other”“Export Pit List”. You will need these pit shells for future assignments. Finding the optimal pit: Click on “Pit by Pit Graph-1. specified and worst case scenarios. Click on “OK” for the message. It shows the best. g. click on run. The optimum pit can be found from there (checking in the discounted cash flow it may be seen which pit shell has the highest value and therefore may be chosen as the ultimate pit). Page 18 .” Click on “output” tab on the right side.pil”. Dimitrakopoulos Fall 2014 “kriga.MIME 413 – Strategic Mine Planning with Uncertainty MIME 513 – Mine Planning Optimization under Uncertainty Lecturer: R. Dimitrakopoulos Fall 2014 h. Click on Graph tab to see the modified plot.” Change the Style to Bar. Click on the “Graph” tab and press zoom button. Click on “Style preferences” and select “Open pit value for the best case.MIME 413 – Strategic Mine Planning with Uncertainty MIME 513 – Mine Planning Optimization under Uncertainty Lecturer: R. Type DCF on the Title of Y-Axis. Page 19 . Repeat these steps to generate your graphs required for the assignment deliverables. Be sure to comment on your results. Select only “Open Pit Value” and unselect the other options. Click on the preferences. and Colour to red. Click on the 2nd Y-Axis preference tab and unselect “Use multiple Y-Axis” because you have chosen only one display (there is no need to display the 2nd axis). Dimitrakopoulos Fall 2014 Page 20 .MIME 413 – Strategic Mine Planning with Uncertainty MIME 513 – Mine Planning Optimization under Uncertainty Lecturer: R. as they will be generated in more detail in the following step. You will not need to worry about the graphs in this step. These tell you the optimal pit shell combinations that Whittle has chosen. You can now enter the final ultimate pit shell to be considered (the limit of mining). as they will be required in the following step (Step 3 b). given the number of pushbacks you have selected. After the algorithm has finished running. Press the running man icon to run the pushback designs. Select the “Definition” tab. and changing the number of pushbacks. you will see numbers listed in order beside “Pushbacks used for last run”. under the “Schedule” tab. and under “Specified case pushback definitions”. Right-click on “Pit by Pit Graph-1”. We will first create pushbacks using the automatic pushback selection tool. enter the ultimate pit shell number defined in Step 2 (g. Dimitrakopoulos Fall 2014 3 Generating Pushbacks from Pit Shells a. Change the name of the node (“Description”) and re-name it to “Pit by Pit Graph.MIME 413 – Strategic Mine Planning with Uncertainty MIME 513 – Mine Planning Optimization under Uncertainty Lecturer: R. and press “Edit”.h) in the “End” textbox.# PB” where # is the number of pushbacks you wish to test. Right-click on “Operational Scenario-1” and select “Paste”. and select “Copy Node”. Repeat this step for various numbers of pushbacks by copying the node and pasting under “Operational Scenario-1”. You will need to write down these numbers. select “Auto” and enter the number of pushbacks you would like in your design. Under the “Schedule” tab. Page 21 . in “Pushback definition”. Then. These were made available to you in Step 3 (a) under “Pushbacks used for last run”. You will need to repeat this step for each pushback configuration you wish to test (# of pushbacks in design). It is recommended that you save these as new scenarios Page 22 . If you are unsure of any of the parameters in this menu. Click on the “Definition” tab and type in the desired mining width (justify your parameters in your report). click on the “Description” tab and type “Kriging Mining Width . please refer to the help file in Whittle.MIME 413 – Strategic Mine Planning with Uncertainty MIME 513 – Mine Planning Optimization under Uncertainty Lecturer: R.# Pushbacks” (where # is the number of pushbacks you want to consider). click on edit pushbacks and type in the desired pit shells numbers to be considered as pushbacks. Dimitrakopoulos Fall 2014 b. Run the program to combine pit shells to pushbacks. then. Adding Mining Width (combine pit shells to pushbacks): Right-click on “Kriging Pit Shells” select “Add” and “Mining Width”. Dimitrakopoulos Fall 2014 Page 23 .MIME 413 – Strategic Mine Planning with Uncertainty MIME 513 – Mine Planning Optimization under Uncertainty Lecturer: R. Dimitrakopoulos Fall 2014 c. In the “Definition” tab. Generating Pit list Krigb. When ready. Feel free to add whatever graphs necessary to justify your decisions in your report’s discussion. Right click on your new operational select and select “Add” then “Pit by Pit Graph”. Compare what happens in pushback tonnages. however if they are incorrect. if you have 4 pushbacks. you should now be able to see your graphs for the pushbacks (i. ensure that Whittle will generate the correct graphs that you require. In the “Specified Case Scheduling Algorithm” section of the “Schedule” tab.g. Choose “Qty of <element> output from <method> (<method><element> /UO*) to have quantity of gold in the output file. and you will need to add your pushbacks.pil: Right click on “Kriging Mining Width”. as per Step 2 (d) . you would enter “1 2 3 4”). Add “Manual” pushbacks by clicking “add” on the right side (Add as many pits as needed e. etc. be sure to put in the correct parameters (Discount rate and Processing limit). Click on “add to selection”. Click “accept”. 4 Optimization with Milawa Algorithm a. New Schedule Graph: Create a production schedule by right clicking on the new operational scenario (whichever was deemed optimal from the previous step). not pit shells. Whittle should add the correct parameters. Page 24 .# Pushbacks). type “Kriging Schedule – Milawa NPV”. press the running man button. d. Here.MIME 413 – Strategic Mine Planning with Uncertainty MIME 513 – Mine Planning Optimization under Uncertainty Lecturer: R. run the program by clicking on 2nd running man. Discuss the differences between the output schedules. Then. choose “GOLD”. if you have 4 pushbacks. toolsexporting pit-list. You will need to do this for which ever design (# of pushbacks) you deem best. ensure that “Specified Case Pushback Definitions” is set to “Manual”. as shown in the previous step). delete all the lines below “Discounted open pit value” and click on add. export it. and then enter your pushbacks in increasing order (e.select the working directory and change the name to krigb. and on the right choose “element and grade”. Press “OK”. and OK. Repeat this step for “Milawa Balanced”. selecting “Add” then “Schedule Graph”. Right click on your pushback scenario created in the previous step (Kriging Mining Width .g. Then. click on “output” on the left. Under the “Schedule” tab. which one you think is more effective. Press the “Add” button. In the “Description” tab. to justify your decision on number of pushbacks in your design before going on to the next step. from “MILL” for “specified case”. You will need to re-do this analysis according to the various # of pushback scenarios you have chosen. this pit list contains block coordinates and the number representing the pushbacks. The output file is required for the next part of the project. Under the “Definition” tab. you would enter “1 2 3 4”). Under the “Graphs” tab. ore tonnages. select “Milawa NPV”. and select “Add” then “New Operational Scenario”. etc.pil.e. Page 25 .csv” and Ok .MIME 413 – Strategic Mine Planning with Uncertainty MIME 513 – Mine Planning Optimization under Uncertainty Lecturer: R. See that the directory file will be created (directory can be changed by clicking at the end of directory name box). Click on the square beside “Spreadsheet File” to put a check-sign and type the name as “krigpb. Dimitrakopoulos Fall 2014 5 Exporting the Results to an Excel Worksheet a. Choose “Options” then “Export Files” options. Right-click on the menu-button on the optimal pushback design you have chosen on the left side. Dimitrakopoulos Fall 2014 Page 26 .MIME 413 – Strategic Mine Planning with Uncertainty MIME 513 – Mine Planning Optimization under Uncertainty Lecturer: R.