A new short- and medium-term production scheduling tool -

Document Sample
A new short- and medium-term production scheduling tool - Powered By Docstoc
					 A new short- and medium-term production scheduling tool –
          MineSight Schedule Optimizer (MSSO)

                          Zhanyou Huanga, Wenlong Caia, A. Frederick Banfielda
                                      a
                                          Mintec, Inc., Tucson, Arizona, USA



The MineSight Schedule Optimizer (MSSO) is a new production scheduling tool developed by
Mintec, Inc. using mixed integer linear programming (MILP) techniques. This tool was
developed to solve the short- and medium-term schedule problems for surface and underground
mining operations that involve multiple models, multiple processes, multiple destinations and
blending requirements. MSSO finds the optimum schedule in each period that achieves the
objective while satisfying comprehensive product quality and quantity requirements as well as
physical and technical constraints. Using MineSight Attributed Geometry Data Model (AGDM)
as a central database, MSSO seamlessly works with MineSight 3D (MS3D) block models and
third-party data, MineSight Interactive Planner (MSIP), MineSight Haulage (MSHaulage) and
MineSight Activity Planner (MSAP) to prepare data, formulate MILP models, calculate, report,
fine-tune, save and visualize schedules. This paper presents the features and capabilities of
MSSO. A case study is also given to demonstrate the applications of this new tool.

Keywords: Scheduling/planning, optimizer, MILP, MineSight.





    MineSight is a registered trademark of Mintec, Inc.




                                                                                           1
Revised November 2009
1. Introduction


A suite of short- and medium-term production scheduling tools has been developed by Mintec.
All of these tools are interconnected through a central database hub that enables them to
seamlessly work together to enhance the mining engineer’s ability to produce timely plans that
have sufficient detail to address most of the operational considerations.

         MineSight currently provides five packages for short- and medium-term production
scheduling [1]. These packages are:

    •   MineSight Interactive Planner (MSIP)

    •   MineSight Operations (MSOPS)

    •   MineSight Haulage (MSHaulage)

    •   MineSight Schedule Optimizer (MSSO)

    •   MineSight Activity Planner (MSAP)

          MSIP is the primary tool in MineSight for cut design and reserve calculations. MSOPS
is mainly used for daily ore control. MSHaulage is a tool that builds haulage networks, provides
haulage equipment requirements and calculates cycle time for other planning tools. MSSO is
dedicated to finding the optimum operational cut mining sequence in each period that achieves
the objective and satisfies the product quality and quantity as well as technical, physical and
marketing constraints. MSAP is a tool that provides detailed scheduling by adding activities,
such as drilling, blasting, mining and hauling, to each cut and calculating the duration of each
activity.

          All the above planning tools are connected with the MineSight Attributed Geometry
Database Model (AGDM). The AGDM is an ODBC compliant, open database that provides pre-
defined tables for storing attributed geometry (mining shapes/cuts, default and user-defined
attributes and atomic reserves) as well as tables for storing equipment data. MineSight AGDM
takes advantage of the new technology for data management by using Microsoft SQL 2005,
SQL Server 2005 Express (a free, easy-to-use, lightweight version of Microsoft SQL 2005) and
future releases.

         Figure 1 shows the connections of the MineSight scheduling tools with the AGDM.




                                                                                              2
Revised November 2009
                        Figure 1. MineSight AGDM and scheduling tools



2. What is MSSO


MSSO is a new short- and medium-term scheduling tool that involves multiple models, multiple
processes, multiple destinations and comprehensive product quality, quantity as well as blending
requirements.

          As Figure 2 shows, suppose cuts will be mined from n models. These models may have
distinct rock material types and grade items. The mined materials will be sent to the destinations,
which consist of m mills, s stockpiles, p leach pads and d waste dumps. Each destination has its
capacity and feed material requirements. Shovel and truck fleets are available. The goal is to use
the available equipment to mine the cuts and send to appropriate destinations so that objectives
will be achieved while product quality and quantity requirements as well as physical and
technical constraints are satisfied. The objective is to maximize net present value (NPV), metal
content, or minimize stripping ratio, haulage distance or shovel/truck hours. This kind of short-
and medium-term production scheduling problem can be solved with this new tool - MSSO.


                                                                                                 3
Revised November 2009
BHP Billiton’s multi-mineral Olympic Dam expansion project, about 550km north of Adelaide,
Australia, is the first site to run this new tool [2].




                                Figure 2. Prototype of MSSO



3. Workflow of MineSight scheduling tools


Figure 3 shows an entire workflow of the existing MineSight scheduling packages. Some of the
processes are optional.




                                                                                          4
Revised November 2009
                        Figure 3. Workflow of MineSight scheduling tools



         Generally, mine planning starts from long-range and then under the guidance of long-
range plans does medium- and short-term scheduling. Following is a step by step walkthrough of
this workflow:

    1) MineSight Economic Planner (MSEP) and MineSight Strategic Planner (MSSP) are used
       for long range planning and design. MSEP is used for pit and pushback design, and then
       MSSP is used to schedule the benches in each pushback to be mined in each period
       (usually 1 year).

    2) Under the guidance of long-range plans obtained in step 1), MSIP is used to design cuts
       and calculate the reserves on those benches and pushbacks provided by long-range plans.
       MSIP saves cuts into AGDM databases.

    3) When cuts are ready, MSHaulage is used to define destinations, build haulage networks
       and calculate the cycle time from each cut to each destination. The cycle time data will
       also be saved into AGDM databases.

    4) MSSO then imports cuts and cycle time data from AGDM databases, defines phase
       precedence, calculates cut precedence , sets up material mappings based on cutoff grades,
       defines destinations and equipment parameters, sets up objectives, constraints and


                                                                                              5
Revised November 2009
        economic parameters for each period and then does schedule calculations. After getting
        optimum schedules, user can check the schedule reports. If some of the scheduling results
        are not satisfactory, user can go back to adjust material mappings, constraints and/or
        economic parameters, and runs the schedule calculations again. The above process will
        be repeated until satisfactory schedules are obtained. Finally, the obtained schedules can
        be saved back to AGDM databases or exported to Microsoft Excel or CVS files. MSSO
        schedules also give the mining sequence of each of the cuts to be mined in each period.

    5) After the period by period schedules have been obtained from MSSO and saved into
       AGDM databases, they can be visualized in MS3D or MSIP, or exported to MSAP for
       more detailed scheduling. The cut mining sequence decided by MSSO can help lay out
       the cuts on the MSAP Gantt-Chart. MSAP can then add drilling, blasting, mining and
       hauling activities to each cut and calculate the duration of each activity. MSAP saves the
       detailed plans into AGDM databases.



4. Features and capabilities of MSSO


4.1 Constraints

MSSO conducts schedule calculations by taking into account a comprehensive set of geometrical, mining,
processing and equipment constraints.

Geometrical constraints

          Phase precedence, i.e. one phase must be mined before another phase can be defined. MSSO
automatically generates the precedence rules for the cuts to prevent undercutting, which is a real bonus
that can save hours of work for the mining engineers. The bench advance (sinking) rates in each phase,
i.e. the maximum number of benches that can be mined from each phase in each period can also be
controlled.

Mining constraints

          Lower and upper limits can be defined for the tonnage and/or volume of the total and/or
individual rock materials (mills, wastes, stockpiles and leaches) that can be mined from all and/or
individual phases in each period. Stripping ratio can also be controlled in each period. In any case, all the
destination capacity restrictions will be honored.

Processing constraints

         Lower and upper limits can be defined for the amount of total and/or individual ores that can be
processed through a processing plant in each period. Blending of materials will be automatically



                                                                                                           6
Revised November 2009
conducted by MSSO to satisfy the ore grades requirements. Blending also honors grade ratio constraints,
such as Zn/Pb >= 3.5 and Zn/Fe >= 2.5.

Equipment Constraints

           The shovel and truck usages can be controlled in each phase in each period. The number of
shovels and trucks, efficiency, availability, capacity and operating cost data can be input interactively.
The cycle time from each cut to each destination is calculated by MSHaulage and imported from the
AGDM. MSSO assigns the shovels and trucks to each cut during the schedule calculations in order to
efficiently utilize the available equipment resources.

4.2 Stockpiles

MSSO provides very flexible mechanisms on the mining and reclaiming of stockpile materials. It allows
for the dynamic filling and reclaiming of multiple stockpiles during the schedule optimization. Users can
interactively control what materials are sent to each stockpiles, and when, where and how much of the
existing stockpile materials will be reclaimed in each period. Stockpile reclaiming methods include first-
in-first-out (FIFO), first-in-last-out (FILO) and average.

4.3 Objectives

The objective of production scheduling is usually to maximize the net present value. However, MSSO
also allows for maximizing metal contents, minimizing stripping ratio, or minimizing haulage distance or
minimizing shovel/truck hours. The objectives can also vary with periods.

4.4 Economic parameters

Following economic parameters can be input in MSSO:

    •   Annual discount rate

    •   Haulage and loading costs (based on $/ton or cycle time)

    •   Fixed mining costs of mill, waste, stockpile and leach materials

    •   Processing cost of each plant

    •   Reclaim cost of each stockpile

    •   Recovery of each metal or mineral in each mill

    •   Selling price of each metal or mineral from each mill




                                                                                                        7
Revised November 2009
4.5 Optimization engines

MSSO formulates the available cuts into mixed integer linear programming (MILP) models and uses
commercialized optimization engines to find the solutions. The default optimization engine is LINDO
API, which has been evaluated internally by Mintec and found to be a very reliable engine for solving
lower- and medium-end scheduling problems. Other optimization engines, such as ILOG CPLEX, were
also evaluated and may be supported by MSSO in the near future.

4.6 Infeasibility analysis

MSSO automatically conducts infeasibility analysis to find the violating constraints when no feasible
solutions can be found in some period. This is a very handy feature that can provide users with some
guidelines regarding how to modify the constraints.

4.7 Reporting

MSSO provides a comprehensive reporting tool with options to export to Microsoft Excel and CVS files
as well as full graphical visualization via MSIP or End-of-Map (EOP) tool in MineSight 3D (MS3D).
With this reporting tool, customized reports can be generated by defining the number and sequence of
columns in each report. A filter or a combination of filters can be applied so that only cuts in selected
models, periods, phases, benches, schedule materials and destinations will be included in the report.

4.8 Highly interactive graphical user interface

MSSO provides user-friendly and highly interactive graphical user interface that allows mine planning
engineers to communicate with the tool. The interface is designed based on the fact that mining
production scheduling, especially short- and medium-term, should be an interactive process between
planning engineers and software. With MSSO interface, user can input the cutoff grade, mining and
processing requirements and economic parameters for each period. User can stop the schedule calculation
progress at some period, make some changes, and restart the schedule calculation from the previous
period. MSSO remembers the period at which the schedule calculations were interrupted by a user.

           Another feature of MSSO interface is that tasks can be processed in parallel. For example, user
can work on other MSSO panels, such as input constraints and economic parameters, while MSSO is in
the process of loading cuts, which may take quite a while if there are a few thousand cuts. Also, user can
check and analyze the obtained schedules using MSSO report while MSSO is still calculating schedules
for other periods. If user wants to change the available schedule parameters, then just cancels the schedule
calculations, changes some parameters and restarts the calculations.




                                                                                                          8
Revised November 2009
5. Case study

Figure 4 shows the pit configuration of an existing open pit mine project. This model has four phases that
are marked with different colors. The total material is about 276.7 million tons. There are six rock
material types and six grade items. The material types are hard hematite, soft hematite, hard itabirite, soft
itabirite, non-estimated and other waste. The grade items are Fe, SiO2, P, Al2O3, Mn, and Mg. Tables 1
and 2 summarize the material tons and average grades.




                                Figure 4. Case study: pit configuration



                                Table 1. Case study: rock materials tons




                                                                                                           9
Revised November 2009
                               Table 2. Case study: average grades




          A 12-period schedule is needed in this case study. In these schedules, both hematite
(hard and soft) and itabirite (hard and soft) are treated as ores. Hematite materials with Fe higher
than 40% will be sent to Mill 1, and itabirite with Fe higher than 30% sent to Mill 2. Hematite
and itabirite materials with lower Fe% will be sent to waste dumps or stockpiled. Other materials
(non-estimated and other waste) will be sent to waste dumps.

        In this project, Fe will generate revenues whereas SiO2, P, Al2O3, Mn, and Mg are
contaminants. The mine plans should increase Fe grade and in the meanwhile decrease
contaminant contents in each period.

          Tables 3 and 4 are the mining and processing requirements. Specifically, period 1 to
period 3 is pre-production stripping. Total wastes should be around 7.4 million tons from period
1 to period 6, and 12.2 million tons from period 7 to period 8, and 4.0 million tons from period 9
to period 12. As for processing, total mill materials should be about 11.4 million tons from
period 4 to period 12, and Fe in Mill 1 should be higher than 64% and less than 70%, Fe in Mill
2 should be between 39% to 45%. The contaminants are capped, i.e., SiO2 in Mill 1 should be
less than 3.7%, P in Mill 1 less than 0.059%, Al2O3 in Mill 1 less than 0.87%, Mn in Mill 1 less
than 0.3%, and Mg in Mill 1 less than 0.3%. All these constraints must be satisfied in each period
by appropriately blending the cuts.




                                                                                                 10
Revised November 2009
                            Table 3. Case study: mining constraints




                             Table 4. Case study: grade constraints




         MSIP was used to design cuts on all the benches in each phase. Figure 5 shows the
designed cuts in each phase. Each cut is about 200,000 tons. A total of 1,458 cuts were designed.
Phase 1 has 443 cuts, phase 2 has 408 cuts, phase 3 has 364 cuts and phase 4 has 243 cuts. The
designed cuts were stored in an AGDM database.




                                                                                              11
Revised November 2009
           Figure 5. Case study: all cuts in phase 1, 2, 3 and 4 (gray-color polygons)



           Figures 6 and 7 are snapshots of MSSO panels that show the input of the mining
constraints and grade constraints. Tables 5 and 6 are period by period summary report of the
schedules calculated by MSSO. It can be seen that all the mining and grade constraints are
satisfied in each period.

         Figure 8 is a 3-D visualization of the 12-period schedule with different colors
representing cuts mined in different periods. Figures 9, 10 and 11 are 2-D views of the schedules
on bench 1147.5, 1141.0 and 1134.5.

        Tables 7 and 8 are two sample reports generated by MSSO. As shown, miscellaneous
customized reported can be generated by checking the items on the left filter panel.




                                                                                              12
Revised November 2009
                        Figure 6. Case study: MSSO mining constraints panel




                                                                              13
Revised November 2009
                        Figure 7. Case study: MSSO grade constraints panel




                                                                             14
Revised November 2009
                        Table 5. Case study: schedule results - mining




                        Table 6. Case study: schedule results - grades




                                                                         15
Revised November 2009
    Figure 8. Case study: a 3-D view of the 12-period schedule marked with different colors




          Figure 9. Case study: a 2-D view of the 12-period schedule on bench 1147.5




                                                                                              16
Revised November 2009
         Figure 10. Case study: a 2-D view of the 12-period schedule on bench 1141.0




         Figure 11. Case study: a 2-D view of the 12-period schedule on bench 1134.5




                                                                                       17
Revised November 2009
                        Table 7. Case study: MSSO report - mining




                                                                    18
Revised November 2009
                        Table 8. Case study: MSSO report - grades




                                                                    19
Revised November 2009
6. Conclusions


The MineSight Schedule Optimizer (MSSO) is a new tool for short- and medium-term
production scheduling. It formulates scheduling problems into MILP models and uses a
commercialized optimization engine to get the solutions. It finds the optimum cut mining
sequence while achieving the production objectives and satisfying quality, quantity, geometrical
and blending requirements. In case no solution is available in some period, MSSO will conduct
infeasibility analysis to find the violating constraints. MSSO works seamlessly with other
MineSight production planning tools via MineSight AGDM databases. It also provides
customized reports with options to export to Microsoft Excel and CSV files.



References


[1] Mintec, Inc. 25th Seminar: May, 2008.

[2] Australia’s Mining Monthly (AMM): April, 2008, Keeping projects on schedule, 128 pp.




                                                                                             20
Revised November 2009