Optimisation of Cut-Off Grade at Mount Isa Mines Limited s Enterprise Mine

Transcription

1 Optimisation of Cut-Off Grade at Mount Isa Mines Limited s Enterprise Mine J Poniewierski 1, G MacSporran 2 and I Sheppard 3 ABSTRACT The practical application of Lane s theory on optimum cut-off grade policy has to date proved difficult for selective stoping in large-scale underground mining. The key to determining an optimum cut-off grade is the ability to rapidly perform complex mine layout designs combined with rapid output of multiple potential schedules. This paper describes aspects of the detailed investigation into estimating a practical static cut-off grade for the two main orebodies in Mount Isa Mine s Enterprise Mine. Rule based automated scheduling software was used to evaluate thirteen different stoping layouts (including design infrastructure and capital development) of the Enterprise Mine s two main orebodies, with over 200 schedules generated. The investigation illustrates the application of commercially available software to solve the problems of generating multiple cut-off grade stope designs and achieving the otherwise insurmountable task of rapidly generating and valuing multiple scenarios for each stoping layout. Each scenario was evaluated using modelled net cash flows of the mining, concentrating and smelting operations of the company and calculating NPV. The cut-off grade for each orebody was selected based on the cut-off grade that achieved maximum NPV with an appropriate production rate. Underground mine scheduling is a complex task involving the adherence to a number of simple scheduling rules and the balancing of several competing drivers. The study also found that: while each scheduling rule itself is simple, the compounding complexities of their interaction preclude manual scheduling from being a realistic option. there is a maximum sustainable rate of production possible for any given stoping layout, and that the actual maximum production rate is also a function of the required number of years of sustainable production. the optimal cut-off grade was different for each orebody a recognition of the impact of the different characteristics of each orebody. INTRODUCTION The calculation of the optimum cut-off grade for any particular mine is a complex procedure involving a considerable amount of geological interpretation, mine design, scheduling, and economic and financial modelling (Hall, 2000). At any given production rate there is generally a particular cut-off grade that will maximise the Net Present Value (NPV) of the mine a grade that will invariably be higher than the operating breakeven grade. As the production rate at a mine is changed, it is also generally found that the cut-off grade will vary a result of fixed capital and fixed operating costs being spread over greater or lesser tonnages. 1. MAusIMM, Business Analyst/Senior Mining Consultant, Runge Ltd, GPO Box 2774, Brisbane Qld jponiewierski@runge.com.au 2. MAusIMM, Formerly Mine Engineering Superintendent 2002 Copper Business Study, currently Senior Mining Engineer, SRK Consulting, PO Box 943, West Perth WA gmacsporran@srk.com.au 3. MAusIMM, Manager 2002 Copper Business Study, MIM Copper Study Group, Mt Isa Mines Private Mail Bag, Mount Isa Qld ikshep@isa.mim.com.au In most studies where the changes in cut-off grade with respect to production rate are assessed, the derivation of the maximum sustainable production rates for an underground orebody have by necessity required a number of simplifications. One such simplification is to calculate the tonnes per vertical metre and assume the mine production rate is limited by the vertical advance rate. There can be a large range of rates used and a typical one used in Western Australian gold mines is 30 to 50 vertical metres per annum (McCarthy, 1993). While useful in conceptual studies, the validity of such simplifications is highly dependent on the mining method used. For a large-scale multiple level open stoping operation such as Mount Isa Mine s Enterprise Mine, an assumption of vertical depletion of an orebody bears no resemblance to the actual mining operations. The need to make such assumptions has in the past been a result of the complexity and time-consuming nature of designing an orebody at different cut-off grades and then manually scheduling each orebody design at different production rates. Iterative mine planning has therefore rarely been undertaken for optimising cut-off grade, although previous authors (Wheeler and Rodrigues, 2002; Hall, 2000; Hall, 2002) have however highlighted the need for a comparative analysis at different cut-off grades and different production rates. This paper describes aspects of the detailed investigation into estimating a best practical static cut-off grade for the two main orebodies in Mount Isa Mine s Enterprise Mine a deep underground copper mine. The investigation was undertaken as part of a major review for the copper business at Mt Isa. The approach described shows the first step towards achieving a practical application of the cut-off grade theory of Lane (1997) in selective underground mining. Commercially available software was used to generate simplified stope layouts for various cut-off grades for each of the two orebodies in the mine. The mine designs for these different cut-off grades were then repeatedly scheduled at various target production rates, again using commercially available software, in order to determine a suitable sustainable production rate for each cut-off grade stoping layout. The resulting production tonnage and grade profiles were used as input into financial models of the mining, concentrating and smelting operations of the company. Based on the maximum Net Present Value obtainable, the ideal cut-off grade for each orebody was thus determined. Further development of the software tools used for this study, to a point where they will allow the rapid automation of an optimum cut-off grade policy for underground mining should be an industry goal. A significant concept that emerged from the study was that for each orebody design, the complex interaction of individually simple scheduling rules that account for the unique characteristics of each orebody, result in there being a maximum sustainable production rate for a given stoping layout design. To exceed that maximum sustainable production rate, major mine design philosophy changes are required. THE ENTERPRISE MINE Located in mineral-rich north west Queensland, Mount Isa Mines (MIM) operates one of the world s largest underground mining complexes, mining copper and lead-zinc-silver orebodies. Mine Planning and Equipment Selection Kalgoorlie, WA, April

2 J PONIEWIERSKI, G MacSPORRAN and I SHEPPARD FIG 1 - Schematic cross-section of Enterprise Mine. The MIM copper stream processes ore from the X41 Copper Mine and Enterprise Mine at Mount Isa and concentrate from MIM Holdings Ernest Henry Mine which lies about 40 km north of Cloncurry. The Enterprise underground copper mine is the most recent copper ore source at Mount Isa and was developed beneath existing operations at a cost of $A 370 million. It commenced commercial production on 1 July 2000, and was officially opened in August The Enterprise Mine is Australia s deepest mine, with mining expected to a depth of 1800 metres below the surface. Ore from the mine has a head-grade of around four per cent copper. The Enterprise Mine required construction of three major fixed plant and equipment systems: a new ore handling system to access the 3000 and 3500 orebodies, a refrigeration plant and a new paste fill plant. A schematic cross-section of the mine is shown in Figure 1. The Enterprise ore handling system consists of an underground jaw crusher, a conveyor connecting to the M62 internal shaft and another conveyor connecting from the top of the M62 internal shaft to the pre-existing U62 shaft and copper ore handling system. The M62 internal shaft is 5.3 metres in diameter and 713 metres long. Its base lies approximately 1660 metres below the surface. Virgin rock temperatures at Enterprise which reach more than 60 C, along with the autocompression of ventilation air at the depth of mining, required installation of a major refrigeration plant. The refrigeration plant provides up to 22 MW of refrigerative cooling to condition the air entering the mine to 14 C. Four large ammonia compressors provide 540 L/s of cold water to the bulk air cooler, where 580 kg/sec of downcast air passes through large spray towers before being sent underground via the R67 fresh air intake. A major paste fill plant was also constructed for the Enterprise mine. The paste fill plant is among the largest of its type in the world and is designed to produce 450 dry tonnes of paste per hour continuously as a 76 per cent solids-by-weight slurry. The paste is transported underground through two vertical boreholes, each 200 mm in diameter and 1100 m deep. The mining method utilised is sublevel open stoping (SLOS), which has evolved over the last 70 years of operation, in the copper and lead orebodies, to the present day design standards. Stope dimensions for the Enterprise Mine are typically 35 by 35 metres in plan and are extracted vertically to the full height of the orebody. The sublevel interval used is approximately 40 metres which limits the length of blast holes. The mining method is discussed in detail in numerous technical papers (Hall, 1992). CUT-OFF GRADE THEORY APPLIED The accepted theory on optimum cut-off grade policy is that by K Lane (1997). Practical application of Lane s theory has however proved difficult for selective stoping in underground mining. The availability of automated scheduling software for open pits has allowed significant advances in cut-off grade analysis in surface mining. However, difficulties in scheduling of underground selective mining operations has deterred most operations from undertaking such an investigation of optimum underground cut-off grades. The work presented in this paper is based on the assumption that the corporate objective in cut-off grade selection is to maximise the Net Present Value (NPV) of the mine, as is the case in the cut-off grade theory expounded by Lane. In many cases this is not the real objective of the corporation. Frequently longest mine life, maximum Mineral Resource recovery or conversely achieving bottom quartile position on the industry cost curve, are cut-off grade selection objectives. These alternative objectives are not always openly admitted. They may be equally valid objectives alongside NPV for individual mine owners. Some Australian mining consulting firms have made comment on these issues over several years (Hall, 2000). In the study undertaken for the Enterprise Mine, NPV was chosen as the primary criteria for cut-off grade selection. At a given point in the life of a mine and with a given future production schedule, there is a particular cut-off grade that will maximise the NPV. Lane has described three cut-off grade equations that can be used to calculate the optimum cut-off grade for maximum NPV. The three equations are required to cover the three main operation scenarios possible: marketing limited operations, treatment limited operations, and mining limited operations. In practice for long-term planning we are concerned with only two cases, the mining limited and the treatment limited cases. 532 Kalgoorlie, WA, April 2003 Mine Planning and Equipment Selection

3 OPTIMISATION OF CUT-OFF GRADE AT MOUNT ISA MINES LIMITED S ENTERPRISE MINE FIG 2 - Optimum cut-off grade with respect to mining limited and treatment limited grade-npv curves. Figure 2 shows an idealised concept of how these equations may apply. The mining limited curve shows that for an orebody constrained only by its own mining rate, there will be a cut-off grade that gives the best NPV. Application of a treatment capacity limit introduces a second intersecting curve. Where the curves intersect is the physical capacity balance point, where the mining and treatment capacities are balanced. In Case 1 of Figure 2, the treatment capacity available is large and the peak of the mining limited curve is at a higher cut-off grade than the intersection point. Hence the mining limited curve defines the best cut-off grade and we can ignore the treatment capacity limit. In Case 2 of Figure 2, the treatment capacity is smaller. The intersection of the curves is at a higher cut-off grade than the peak of the mining limited curve. In this case we have to consider the treatment capacity limit and the best cut-off grade is the capacity balance point which is higher than for the simpler Case 1. The difference in the two solutions is the mathematically explained by Lane s opportunity cost concept. Other possible combinations of interaction between these curves exist. But these simple cases explain the approach used in the Enterprise Mine study described here. The reader is referred to Lane (1997) for more detail. The Enterprise Mine cut-off grade study assumed that the situation described in Case 1 of Figure 2 existed for the orebodies being reviewed. The Enterprise orebodies are the highest grade copper deposits on the lease and processing capacity is twice the likely mining capacity. Hence the problem was simplified to establishing the nature of the mining limited curve. It also needs to be noted that the above analysis is valid for only one point in time. A true optimised cut-off grade policy will consider the change in NPV of the mine over time. Lane (1997) describes how this cut-off grade policy can be estimated. The software to simulate this optimum cut-off policy is available for open pit mines. The problem is more complex for selective underground mining operations. A fundamental concept in the optimum cut-off grade policy theory is that cut-off grade for a mine should start high and then decline over time reflecting the fall in the mine s NPV as it is worked out. A continuously falling cut-off grade is the ideal. A major issue for this concept for many mining operations is that adopting a high cut-off grade early in the mine life may result in sterilisation of Mineral Resource. Shortened mine life and loss of Mineral Resource are the result. This is one of the reasons that companies may not adopt maximum NPV as the selection criteria for cut-off grade. Open pit mines can avoid the sterilisation cost by stockpiling low-grade ore. Underground mines mostly do not have this option. Additionally, the concept of changing cut-off grade every year in large underground mines in order to develop the optimum cut-off grade policy is often not practical. The skilled geology and engineering resources to redesign every stope for a new cut-off grade each year are simply not available in the average mine. A practical solution is to select either a few cut-off grade steps over the life of mine or only a single life of mine cut-off grade that shows the best NPV solution. In the study presented here the decision was made to estimate the first step in the cut-off grade for the mine. To estimate this first step a life of mine design and schedule over a range of cut-off grades is required. Scheduling and selectivity are the two key issues to be solved in determining the mining limited cut-off grade NPV curve. Underground mine scheduling is a complex task involving the balancing of several competing drivers. The same problem applies to selecting the volume of rock from amongst the Mineral Resource that can be mined as Ore Reserve. When the selection and scheduling needs to be done at many cut-off grades the problem can seem just too big. As stated in the introduction, simplifying assumptions must be made to control the size of the problem. Too many or too crude assumptions however can destroy the sensitivity of the orebody to the calculations involved. The aim is to achieve the right balance between the simplifying assumptions and the real world complexity. For any one mine design there are many possible production schedules. Each possible schedule will generate a different NPV result for the orebody. The more complex the stoping design the more variation on NPV between alternative schedules. In simple narrow single orebody mines the schedule choices are limited, so the variation in NPV will be small. For complex geometry and large orebodies the schedule choices are greater and the NPV results will have more variation. In this study sophisticated, but commercially available software was used to solve the issues of selectivity and scheduling. CUT-OFF GRADE DETERMINATION TECHNIQUE The steps taken to determine the cut-off grade for each orebody were in summary (MacSporran, 2002): A. Selectivity solution Create simple mine designs at each cut-off grade. 1. establish the range of cut-off grades of interest. Simple economic analysis showed the likely range; 2. select StopeSizor Smallest Mining Unit (SMU) parameters; 3. establish stope modelling criteria; 4. generate psuedo Mineral Resource ; 5. design stopes for each cut-off grade model; 6. establish dilution criteria for stope models; and 7. design infrastructure and capital development required for each cut-off grade. Mine Planning and Equipment Selection Kalgoorlie, WA, April

4 J PONIEWIERSKI, G MacSPORRAN and I SHEPPARD B. Scheduling solutions Generate life-of-mine schedules for each cut-off grade mine design. There are many possible schedules for a mine. The use of rule based semi automatic scheduling software allows the identification of the best schedule for a given cut-off grade. 1. build XPAC Autoscheduler model for each design; 2. run, interrogate and re-run XPAC Autoscheduler scenarios until a maximum sustainable production rate is achieved; and 3. output XPAC Autoscheduler life-of-mine schedule details. C. Simple economic calculations Evaluate all cut-off grade design maximum rate schedules in a cash flow model to determine maximum NPV cut-off grade, involving: 1. establish costs and pricing/revenue assumptions; 2. establish a cash flows analysis model, including costing of mining, concentrating and smelting operations including both fixed and variable capital and operating costs; 3. conduct a sensitivity analysis on the key assumptions; 4. Mineral Resource cut-off grade is selected as the grade that gives the highest cumulative cash value; and 5. Ore Reserve cut-off grade is selected as the grade that gives the highest Net Present Value. Thirteen different stoping layouts were generated in the study of the Enterprise Mine s two main orebodies. For each stoping layout at least eight and up to fifteen different production rate targets were investigated. Additionally, a number of What If scenarios were reviewed to establish the sensitivity of the production schedules to desired but flexible mining constraints. In total, over 200 schedules were generated. This has allowed the mine to determine its Mineral Resource and Ore Reserve cut-off grades. These cut-off grades were optimised from those previously used and are now specific for each of the two Enterprise Mine orebodies recognition of the impact that the unique characteristics of each orebody has on the final cut-off grade decision. A number of the steps taken to evaluate the cut-off grade will be discussed in more detail in the following sections of the paper. CREATING THE CUT-OFF GRADE DESIGN MODELS Establishing a range of cut-off grades to evaluate The geological resource block model was used to establish a range of cut-off grades to be evaluated. The grade distribution of this selection ensured that the expected optimum cut-off grade was between the minimum and maximum. To reduce the amount of work required, no more than five or six cut-off grades were selected for evaluation. Generating a psuedo Mineral Resource A pseudo Mineral Resource refers to material that is considered mine-able at the selected cut-off grade. To establish a pseudo Mineral Resource for each cut-off grade model, Snowden s StopeSizor software was used (Thomas and Earl, 1999). This software filters the block model and lumps the geology model blocks together within a specified Smallest Mining Unit (SMU). Refer to for further information. The SMU represent a block size that assist in establishing shapes for stope designs. In the case of the Enterprise Mine the SMU is 30 mn by 30 me by 10 mrl. The results of the StopeSizor analysis were output in a form suitable for review in Microsoft Excel. Figure 3 shows a typical cross-section (representing a 30 m of Northing) for one of the orebodies. The value within each grid represents the grade (per cent copper) of the SMU block. FIG 3 - Typical cross-section of StopeSizor output in Excel. The lumped blocks that match or exceed the cut-off grade being reviewed are tallied as part of the pseudo Mineral Resource. Designing stopes for each cut-off grade model The conversion from a pseudo Mineral Resource to a pseudo Ore Reserve required stope models to be created for each cut-off grade being reviewed. A consistent set of rules established to construct the stope models had to be followed at each cut-off grade. This was to ensure that the analysis was relative to the change in grade and not the change in the stope model design. The rules used in constructing stopes with a SMU of 30 m by 30 m by 10 m were: 1. smallest stope height is 30 m (three SMU blocks high); 2. bands of low-grade blocks can be stoped but no more than 20 m (two blocks) thick; and 3. sill pillars between stopes must be 30 m (three blocks) thick. Since the stopes are effectively a series of regular blocks, the stope models were constructed in Excel by aggregation of the SMU blocks with grades above the designated cut-off. A typical cross-section through one of the orebodies is shown in Figure 4, with the constructed stopes outlined for a cut-off grade in this case of 1.7 per cent Cu. The SMU blocks greater than or equal to the cut-off grade (used to create the pseudo mineral resource) are highlighted. The corresponding tonnage and grade for each stope were tallied below the stope column lower in the Excel spreadsheet. Establishing the dilution criteria for stope models Two principal conversion factors by which mineral resources are modified to produce ore reserves are mining dilution and mining recovery (Stephenson and Vann, 1999). For this study the dilution for the stopes was based on estimates of the Equivalent Linear Overbreak/Sloughing (ELOS) for the stope crown and any fill walls exposed. The stope crown generally has mineralised material, the grade of which was estimated by using the grade of the SMU block directly above the stope crown. The fill material has an effective zero grade. 534 Kalgoorlie, WA, April 2003 Mine Planning and Equipment Selection

5 OPTIMISATION OF CUT-OFF GRADE AT MOUNT ISA MINES LIMITED S ENTERPRISE MINE 1. a description of each segment of the capital development including length and appropriate development rates for that segment; 2. a listing of the pre-requisite capital development segment(s) for each segment of the capital development; and 3. a listing of the pre-requisite capital development segments for each stope. For the largest of the cut-off grade designs (the lowest cut-off grade design), the number of capital development segments (horizontal and vertical) was of the order of 100 individual segments (in comparison to about 330 segments for the detailed final design model). FIG 4 - Typical cross-section of lumped SMUs to create stope designs at 1.7 per cent cut-off grade. The total effect of the Fill dilution is dependent on the number of exposed fill masses or the Sub Level Open Stope (SLOS) state. The sequence in which the stopes are mined dictates this state. A series of dilution possibilities were therefore assigned to each modelled stope one for each possible number of fill exposures. During the XPAC scheduling process, the appropriate dilution value was then assigned dynamically, dependent upon the fill exposure state at the time of stoping. Designing infrastructure and capital development for cut-off grade model The capital development requirements (level and decline) and infrastructure requirements (ore-passes, ventilation shafts and raises, and trucking loops) were needed to be determined for each cut-off grade design. This involved determining a detailed but conceptual level of design. The level of detail required (for use in the XPAC Autoscheduler model and in the cash flow analyses) was such that three lists were required: Refer to for further information. SCHEDULING THE CUT-OFF GRADE DESIGNS Preliminary The next major step in the cut-off grade analysis was to determine the maximum sustainable production rate for each cut-off grade model. The schedule that best represents the likely production rate is the one that has consistent sustainable production over the shortest period duration. In order to complete the required scheduling tasks in an appropriate time frame, manual scheduling of the cut-off grade designs was out of the question. A rigorous and automated scheduling method was required one that could allow large numbers of schedules to be run quickly as well as obey the scheduling rules for each orebody and enable evaluation of the dynamic and static scheduling factors for each orebody. Having had some earlier experience with the heuristicscheduling program XPAC Autoscheduler (Poniewierski, 2002a, 2002b, 2002c), this software was chosen by MIM as giving the best representation of achievable production and grade profiles. Thirteen different stoping layouts were generated as part of the study. For each stoping layout at least eight, and up to 15 different production rate targets were investigated. Additionally, a number of the specific layout production rate scenarios were investigated for varying schedule objectives. In total, approximately 200 schedules were generated and analysed within a two-week period following model construction (Poniewierski, 2002d). FIG 5 - Plan view of XPAC database stope outlines for three of the six cut-off grade designs used for one of the orebodies shaded for contained copper metal tonnages. Mine Planning and Equipment Selection Kalgoorlie, WA, April

6 J PONIEWIERSKI, G MacSPORRAN and I SHEPPARD Building the XPAC Autoscheduler model Model database In essence, XPAC Autoscheduler is a programmed suite of tools (a toolbox ) which quickly converts a spatially defined XPAC database (a mine s reserves) into a time-based database (ie the definition of scheduling). Autoscheduler determines its own mining sequence based on user specified targets, rules, constraints and objectives. The constraints and rules instruct Autoscheduler on what it cannot select in a given situation, and the objectives guide Autoscheduler in the choice of the next stope to schedule from the alternatives available. The Autoscheduler model built for this study was based on an XPAC database that contained details about each stope in the deposit. The database was used to store a number of attributes for each stope tonnes, grade (including those for different dilution alternatives), and factors such as LHD tram distance, stope height, extraction level and ore-pass allocation. Programming scripts were used to calculate more complex attributes (eg allowable production rates for each stope). Figure 5 shows stope outline plans and their contained metal tonnages for the XPAC database layouts for three of the six cut-off grades used for one of the orebodies. Model logic The logic built into the Autoscheduler model faithfully replicated the majority of the logic that mine schedulers had used in the past, and extended that logic when previously used simplifications were no longer needed owing to the computing power available. Many different activities must be performed during the course of mining a stope. The Autoscheduler model built for this study allowed for five sequential activities to be scheduled replicating the stoping life cycle for each stope. These five activities were: Preliminary activities: being the near fill development mining, the cut-off slot raising and the production drilling of the stope. The time for completion of these activities obviously varies dependent upon the size of the stope. Stope production ( stoping ). Pre-fill activities: representing the time delay from completion of stope production to commencement of backfilling effectively the time to prepare the stope for filling primarily fill-wall construction (particularly on the extraction level). Backfilling with paste fill. Backfill curing: representing the time to cure the cemented fill of a stope before preliminary activities are allowed to occur in stopes adjacent to the filled stope. Specifying the spatial characteristics of each stope was an essential part of this model s set up. This involved identifying the other stopes that were adjacent to each stope and whether each stope was accessible externally or only accessible via other stopes. This allowed the scheduling behaviour of the model to honour three important spatial rules: First Adjacent Stope Rule, ensuring that two adjacent stopes could not be mined simultaneously. Second Adjacent Stope Rule, which prevented two stopes being mined simultaneously if they exposed two sides of the same filled stope. Accessibility Maintenance Rule, which prevented selection of a stope for mining, if selection would prevent access to other stopes in the orebody (development mining through fill not practised). Examples of other rules and constraints used to control the scheduling behaviour included the following: A capital development time constraint, which was used to ensure that a stope did not become available until sufficient time to perform the prerequisite development, had lapsed. Use of variable stope-specific production profiles, which modelled the production ramp-up to a maximum production rate for an individual stope, followed by a production ramp-down. The maximum rates achievable were a function of the tramming distance to the assigned ore-pass. FIG 6 - Report graph outlining the monthly achieved production based on an annual target rate showing periods of no production and low production. 536 Kalgoorlie, WA, April 2003 Mine Planning and Equipment Selection

7 OPTIMISATION OF CUT-OFF GRADE AT MOUNT ISA MINES LIMITED S ENTERPRISE MINE Use of stope sequence dependencies in order to impose a specific predecessor successor sequence on certain stopes. Dependencies were used to ensure compliance with the mine s pre-determined ventilation restrictions, rock mechanics requirements and maintenance of tramming access corridors. Each rule itself is simple, but their interaction quickly becomes quite complex. Determining maximum sustainable production rates The constructed XPAC Autoscheduler model for each cut-off grade design was then scheduled at increasing production rates in order to find the maximum sustainable production rate for that cut-off grade design. Various objectives with varying competing priorities were tried for each schedule scenario in order to: maximise the sustainable production rate; maximise the time for which this production rate could be achieved; maximise the copper grade earlier in the schedule; minimise any production sequence tail resulting from single stope pillar retreats; and minimise the number of total fill exposures for any individual stope. The model s scheduling results were interrogated through the use of graphs, tables and animations of the mining sequence. Figure 6 provides an example of an XPAC report graph outlining the monthly achieved production based on an annual target rate. This figure shows the periods of no production or low production caused through an attempt to produce from the orebody at too high a rate. These no production / low production periods occur when no stope is available for production, due to combinations of the spatial constraints of the stoping sequence and activity rate limits of the operation. This highlights the fact that there is a maximum sustainable rate of production possible for a stoping layout. Schedules were repeatedly re-run with different objective priorities and manual modifications to stope dependencies until a satisfactory scenario that met the previous listed objectives was achieved. Figure 7 shows the annual tonnage and grade production profiles for three decreasing annual target rates. As can be seen, the problem of choosing a suitable scenario becomes one of deciding how many years at the sustainable maximum rate is sufficient. In general a sustainable maximum production rate of at least five years was deemed necessary. This period was believed to be a commensurate time frame with an efficient use of a set mobile-equipment fleet size, and a time frame in which further resources can be delineated by future diamond drilling with the expectation that the rate could then continue to be maintained. Output life-of-mine schedule details The main output required from the schedules, for use in the cash flow analyses were the tonnes and grade production profiles (in time). Additional data required however included backfill quantities schedule, capital development schedule, operating development schedule, production-drilling schedule, and a raise-boring schedule. OPTIMUM CUT-OFF GRADE DETERMINATION Each schedule created for each cut-off grade model and production target rate was put through a cash flow analysis model, modelling the mining, concentrating and smelting operations of the company, with assumptions of prices and exchange rates for revenue determination. FIG 7 - Annual tonnes and grade production profiles at three decreasing annual tonnage target rates. The analysis resulted in a Cumulative Cash Value (a zero discount rate assumption) and Net Present Value. Each of these values was plotted against the other cut-off grade models and the peak of these graphs represented the optimum cut-off grade values. An example of such a curve for one of the orebodies is shown in Figure 8. It was also found that the optimal cut-off grade for each orebody was different from each other a recognition of the impact of the different characteristics of each orebody. The peak of the cumulative cash value curve represents the cash breakeven cut-off grade for the orebody. Below this grade, the total cash generated is lower and therefore some of the material being mined must be loss making (and therefore below the break-even cut-off grade). Above this grade, the total cash generated is lower as a result of potentially profitable material not being mined. This cut-off grade has been used as the Mineral Resource cut-off grade, as resources can be quoted if they have reasonable expectation of economic extraction, ie if they are cash breakeven. Mine Planning and Equipment Selection Kalgoorlie, WA, April

8 J PONIEWIERSKI, G MacSPORRAN and I SHEPPARD This has allowed the Enterprise mine to determine its Mineral Resource and Ore Reserve cut-off grades. These cut-off grades were optimised from those previously used and are now specific for each of the two Enterprise Mine orebodies recognition of the impact that the unique characteristics of each orebody has on the final cut-off grade decision. FIG 8 - Graph of cumulative cash value and Net Present Value for cut-off grade designs modelled for one of the Enterprise Mine orebodies. The cut-off grade at the peak of the NPV curve is a higher cut-off grade than for the cumulative cash curve a result of influence of the opportunity costs (as per Lane s theory). This is the grade used for Ore Reserve calculation and mine design. A sensitivity analysis was also undertaken and indicated that the operating cost was the most sensitive parameter in selecting the optimum cut-off grade. The discount rate used in calculating the NPV, the copper price and the exchange rate, while modifying the NPV and cumulative cash values, did not however change the optimum cut-off grade selection. CONCLUSION The determination of an optimum cut-off grade in a large-scale selective underground mine requires the ability to rapidly perform complex mine layout designs and multiple scheduling runs. By establishing a simplistic but representative model of an orebody it is possible, with commercially available software, to make a defendable estimate of the optimum cut-off grade to meet the business needs. In this study, such commercially available software was successfully used to generate multiple cut-off grade stope designs and to achieve the otherwise insurmountable task of rapidly generating multiple scenario evaluations of these multiple stoping layouts. This task has been achieved in a manner that honoured the real world rules and complex trade-offs of a sublevel open stoping mining system. The scheduling software also provided the mine management and operational personnel with a strong understanding of the compounding complexities of the interaction between the orebody and the mining rules of the operation. REFERENCES Hall, B E, Copper ore mining at Mount Isa Mines Limited, Mount Isa, Qld, in Proceedings Underground Mass Mining Conference 1992, pp (South African Institute of Mining and Metallurgy: Johannesburg). Hall, B, Practical cut-off grades for underground mining maximising value, or marginalising resources? Digging Deeper (AMC Newsletter), February. Hall, B, Cut-off Grades Theory and Practice Workshop, mineability, Townsville, 1 August. Lane, K F, The Economic Definition of Ore, Second Edition (Mining Journal Books: London). MacSporran, G, Enterprise Mine cut-off grade analysis and results, unpublished internal report to MIM 2002 Copper Business Study. McCarthy, P L, Economics of Narrow Vein Mining, in Proceedings Narrow Vein Mining Seminar 1993, pp (The Australasian Institute of Mining and Metallurgy: Melbourne). Poniewierski, J, 2002a Orebody XPAC Study Results and Model Description, unpublished internal report to MIM 2002 Copper Business Study Report #4282 Runge Pty Ltd, January. Poniewierski, J, 2002b Copper Business Study XPAC Autoscheduler Model Refinements, unpublished internal report to MIM 2002 Copper Business Study Report #4327a Runge Pty Ltd, February. Poniewierski, J, 2002c Copper Business Study 3500 Orebody XPAC Scheduling, unpublished internal report to MIM 2002 Copper Business Study Report #4327b Runge Pty Ltd, February. Poniewierski, J, 2002d Copper Business Study XPAC Scheduling of Various Cut-off Grade Stope Layouts for Enterprise Mine, unpublished internal report to MIM 2002 Copper Business Study Report # 4327c, Runge Pty Ltd. Stephenson, P R and Vann, J, Common sense and good communication in mineral resource and ore reserve estimation, in Proceedings Pacrim 99, pp (The Australasian Institute of Mining and Metallurgy: Melbourne). Thomas, G and Earl, A, The application of second-generation stope optimisation tools in underground cut-off grade analysis, in Proceedings Strategic Mine Planning Conference 1999, p 6 (The Australasian Institute of Mining and Metallurgy: Melbourne). Wheeler, A J and Rodrigues, R L, Cutoff-grade analysis at Fazenda Brasileiro: mine planning for declining gold prices, Trans Inst Min Metall, 111, January April, A(Min Tech): Kalgoorlie, WA, April 2003 Mine Planning and Equipment Selection