The Role of Geotechnical Block Models in Mineral

Mineral Resource Management Colloquium - 2007

THE ROLE OF GEOTECHNICAL BLOCK MODELS IN MINERAL RESOURCE MANAGEMENT

Alan Bye - Anglo Platinum Megan Little - Anglo Platinum

1

INTRODUCTION

Geology, and the detailed understanding of its properties, are fundamental to the optimal design and successful operation of any mine. To that end extensive fieldwork has been conducted at Anglo Platinum’s PPRust operation to collect geotechnical information both from exploration boreholes and in-pit mining faces. Since 1998, geotechnical data has been bee n collected from over 337 km k m of exploration core, 7 km of exposed exposed mining faces and more than 15 600 field and laboratory tests, in order to define the complete set of geotechnical properties for each rock type in the mining area. The geotechnical information collected is stored in two databases, SABLE Warehouse and MineMapper3D before it is imported into the Datamine mining software package. The geotechnical parameters, mining rock mass rating system (MRMR), uniaxial compressive strength (UCS), fracture frequency per metre (FF) and rock quality designation (RQD), are modelled using geostatistics to generate a 3D geotechnical block model. Data is interpolated between exploration boreholes and exposed mining faces and modelling modelling is constrained, using wireframes, by rock type and structural features. By having detailed geotechnical information available in a 3D model that can be readily accessed and interpreted, significant production optimisations, feasibility studies and planning initiatives i nitiatives can be implemented. From a slope slope design perspective the model is used to target data deficient zones and highlight potentially weak rock mass areas. As this can be viewed in 3D, the open pit slopes can then be designed to accommodate the poor quality area before it is excavated. It also follows that geotechnical zones can be readily identified and the slopes optimised accordingly. Rather than viewing the drill and blast department as an isolated cost centre and focussing on minimising drill and blast costs, the fragmentation requirements of the comminution and load and haul business areas are studied. It is well understood that chemical energy is the cheapest form of comminution and that major downstream benefits can be derived by increasing drill and blast expenditure. 238 blasts were assessed to determine the optimum fragmentation requirements for ore and waste. Based on the study a mean fragmentation target of 150 mm was set for delivery to the crushing circuit and a mean fragmentation of 230 mm was set for waste loading from the pit. Substantial benefits have been realised in the drill and blast department by developing empirical correlations that relate the MRMR values in the geotechnical model to a blastability index, fragmentation, required powder factor and costs. As the geotechnical model can predict changes in geotechnical conditions, the blasting

1


parameters can be adjusted in advance to ensure the load and haul and comminution plant’s fragmentation requirements are met. PPRust operates autogenous mills which are sensitive to the fragmentation profile delivered. The harder zones occurring in the ore zone therefore have a major impact on the plant’s performance. Accordingly these zones are identified and additional blast energy is introduced in order to achieve the target fragme ntation for the plant. The 3D geotechnical model allows these optimizations to be undertaken proactively by the drill and blast department.

2

OVERVIEW OF PPRUST OPERATIONS

Potgietersrust Platinums Ltd (PPRust) is Anglo Platinum’s only open pit operation. It is located 35 km north of Mokopane (previously Potgietersrus), in the Limpopo Province of South Africa (Figure 1). It is situated in the centre of the northern limb of the Bushveld Complex, a saucer-shaped layered igneous intrusion. The northern limb hosts the Platreef orebody, which is a ~100 m thick tabular body that strikes north-south, dips 45° to the west and reaches a depth of at least 2 km. The Platreef is capped by the Main Zone hangingwall sequence consisting of gabbronorites. The mineralization is hosted predominantly within pyroxenite and parapyroxenite. The parapyroxenites are conformable with the footwall of the Platreef and are essentially a contaminated metamorphosed pyroxenite formed between the cold country rock and the Platreef intrusive phase. The Platreef transgressed sediments of the Transvaal Supergroup as well as Archaean granites and the interaction and varying degrees of assimilation has resulted in a metadolomite, known generically as "calc-silicate", and granofels footwall. The Platreef is a PGM deposit and contains economic quantities of platinum, palladium, rhodium, gold, copper and nickel, which are extracted and processed by Anglo Platinum. PPRust is the only operational mine on the Platreef, which has an economic strike length of 42 km.

2


Figure 1 Location map of the Bushveld Complex outcrop and Anglo Platinum operations (AP, 2006)

PPRust operates three open pits with a further three pits planned (Figure 2). Sandsloot open pit, the world’s largest platinum open pit, started in 1992 and is currently 1.6 km long, 600 m wide, 260 m deep. It is expected to operate until 2009 when it will reach a final depth of 300 m. Zwartfontein South open pit, situated 1 km north-west of Sandsloot, began in 2002 and is 1.5 km long, 500 m wide and 120 m deep and is expected to operate until 2015. PPRust North open pit had its first blast in August 2006 and is currently only 30 m deep. Together the three pits produced 67 Mt of rock, including 5 Mt of ore, in 2006. The operation is expanding and the production targets for 2007 and 2008 are 90 Mt and 120 Mt respectively. Ore with a grade greater than 3 g/t is crushed and fed into a 150 mm grizzly and is split into coarse material (+150 mm) and fines (-150 mm). The concentrator is then fed by a mix of coarse and fines in a 40% - 60% ratio at an average grade of 4 g/t. The ore passes through two autogenous mills and two ball mills before it enters the flotation cells. The milling rates are dependant on rock hardness and particle size (or fragmentation) thus good blasting can go a long way to optimizing plant performance. The concentrator has a capacity to process 400,000 tonnes per month. A second concentrator is under construction adjacent to the PPRust North open pit and it is designed to process 600,000 tonnes per month.

3


Figure 2 Farms, open pits and the Platreef on the PPRust lease area (AP, 2006)

3

GEOTECHNICAL DATA COLLECTION

The data collection at PPRust is done with the entire mining and processing operation in mind, both on a short term and long term scale. Primary geotechnical data is collected by borehole core logging, by surface and face mapping and by field and laboratory rock testing. This data is used to understand the rock mass that is being mined and to determine parameters that are used for analysis. If the quality a nd quantity of data is not sufficient then the analysis and resultant designs will be inaccurate. It is important that the correct data is collected according to international standards, that it is stored in secure databases and is optimally and appropriately used. At PPRust, all diamond drilling core is logged geotechnically and point load tested by contractors. Since 1997 over 337 km of core have been geotechnically logged and more than 11 000 point load tests have been done at PPRust. Face mapping is performed when time allows and point load tests are done for all face maps. Orientated drilling and lab testing is done as required and 350 uniaxial compressive strength (UCS) tests, the most common method for lab testing of rock strength have been done. S ite-specific conversion factors for point load index have been calculated for all the major rock types. This allows a UCS value to be determined more accurately throughout the mining area. Metallurgical tests have been commissioned by the geotechnical engineers to characterise the major rock types for purposes of the processing plant. The Bond Work Index (BWI) is a measure of the resistance of the material to crushing and grinding and

4


is used to determine the energy requirements of a ball milling process. The Drop Weight Test (DWT) measures the impact parameters for an autogenous mill where there are two main mechanisms of breakage, impact (high energy) and abrasion (low energy) (JKTech, 2006). 62 BWI tests and 90 DWT tests have been done on the main ore bearing rock types at PPRust. The logging and window face mapping are used to determine rock mass ratings and the point load test and UCS lab test results are also used in these calculations. All of this data is stored in two databases, SABLE Warehouse and MineMapper3D, which were customised for Anglo Platinum at PPRust. Validation checks, standardized lookups and site specific limits can be applied to every field in each table to ensure the data is accurate and usable. Until 2003, PPRust only calculated Laubscher’s Mining Rock Mass Ratings (MRMR) (1990). Anglo Platinum’s underground operations, however, used only Bieniawski’s RMR (1976) and Barton et al .’s Q (1974) systems. A logging standard was required for Anglo Platinum thus all three rock mass rating systems were combined into one log. Using a single log instead of three separate logs was chosen for the time and cost savings that it would provide. Using three rating systems enables quality control and auditability and ensures that the standard can be used at all Anglo Platinum operations. Hoek (2001) recommends that at least two methods be used at any site during the early stages of a project as different classification systems place different emphases on the various parameters. All three rock mass ratings are calculated and stored in the SABLE and MineMapper3D databases. In order to integrate and interpret the testing, mapping and logging geotechnical data, it is imported into Datamine, PPRust’s modelling software program. The implementation of SABLE and MineMapper3D databases has made this process easier and quicker as well as more auditable. Links are setup to import the data in a set format from SABLE and MineMapper3D. The facemaps are treated as horizontal boreholes for viewing and modelling purposes. In Datamine, the boreholes can be viewed in 3D, which is not possible in SABLE, and this improves the ability of the geotechnical engineer to analyse the logging data as well as apply adjustments for Bieniawski and La ubscher’s systems more accurately. The unadjusted ratings for boreholes are imported into Datamine and the final ratings are calculated based on the boreholes’ location in the open pit. A Datamine script (Figure 3) was written to simplify the importing, viewing and calculation processes as well as to standardise them so any geotechnical user can manipulate the data. The script also allows the user to view the relevant pit designs and blast patterns so that the field data can be compared to the actual designs and conditions in the open pit.

5


Figure 3 Datamine script developed at PPRust and visual display of boreholes for the proposed PPRust North open pit.

4

GEOTECHNICAL BLOCK MODELLING

Over a period of six years, from 1998 to 2003, a geotechnical block model for Sandsloot north pit was developed in Datamine. Since 2003, geotechnical block models for Zwartfontein South and PPRust North open pits have been created. The concept used is the same as that of an ore reserve model. In this case, instead of interpolating grade, the geotechnical parameters, UCS, RQD, FF and MRMR are interpolated into 15m x 15m x 15m blocks. Data is interpolated between exploration boreholes and exposed mining faces and modelling is constrained, using wireframes, by rock type and structural features. Calculations are performed on the model cell values to determine parameters used for slope design, blast design and plant design. Figure 4 illustrates the flow of information for creating a geotechnical block model.

6


Proto model (pit design)

Field data: boreholes, facemaps, point load tests, UCS

Lithology & fault wireframes

Geotechnical zones

Ore/waste zones

Interpolation: RQD, UCS, FF, IRMR, Q, RMR, PLI

Calculate MRMR, Slope angles, BI, EF, Cost, DWT

Plant design

Slope design

Blast design

Figure 4 Illustration of the flow of information in the geotechnical block model

The primary, and most common, use of geotechnical data in an open pit is slope design. Stewart and Kennedy (1971) showed that it was not only the steepness of the ultimate slopes in an open pit mine that had an influence upon the overall profitability of an operation. They contended, on the basis of cash flow calculations, that there is frequently considerable economic advantage to be gained from using steep slopes during the initial stripping programme. This is particularly the case at PPRust where the stripping ratio plays a large role in profitability. The model can be used to optimise the pit slopes rather than applying a single design to the entire pit. A one degree slope optimisation generates an additional R300 million in revenue during the life of Sandsloot pit. In Datamine, Laubscher’s adjustments are applied based on the geotechnical zone in the model to produce a MRMR value per block. Laubscher’s MRMR can be used to calculate an ideal slope angle using Haines and Terbrugge’s (1991) slope design chart. This chart allows the user to calculate a slope angle based on MRMR, slope height and factor of safety (FOS). In mining, a FOS of 1.2 is usually used for slopes while a FOS of 1.5 can used for ramps. A formula for a 100 m stack at FOS=1.2 was derived from the chart. A slope angle (100 m stack, FOS=1.2) for each block in the model is then calculated. Equations for other stack heights at FOS=1.2 were also derived so that slope angles for different stack heights can be calculated. The required slope height is selected and the slope angle is calculated for each model cell. The model can then be colourcoded on slope angle and viewed on plan or in section with the pit design (Figure 5). The ‘Visualise’ button in the script places the model data on a chosen pit design, highlighting areas of high risk (Figure 6). Standard cross- sections can be used or any user-defined sections can be displayed. A ‘Slope Risk’ is calculated by comparing the design versus actual slope angle and highlights where the slopes are over or under designed. This helps to identify where slope monitoring and/or support needs to be

7


installed or increased and where slope design may need to be changed. The slope design functionality has highlighted the fact that the Zwartfontein South pit is under-designed and the slope angles could be increased by at least 5 degrees. This translates to a cost benefit of over R1 billion for the life of the mine. It must be noted that although this functionality aids in the slope design process, geological structure, groundwater, monitoring and mining method must be taken into account for the final design.

Slope too flat

Pit design Figure 5 Vertical slice through the Zwartfontein South block model coloured on Slope angle for a 100m stack at FOS=1.2 indicating the western wall is under-designed.

nd Figure 6 Visualisation of MRMR on the pit slopes with the 2 script

Apart from calculating slope angles, the model is used to refine the geotechnical zones. Other rock properties, such as cohesion and friction angle, can be included in the model (based on rock type) and used in failure analysis. Even with this functionality, the geotechnical block model currently only provides a guide to slope design. The full design process includes numerical modelling. The block model does improve the confidence in the data used for design though, and it provides a tool for much more regular review and incorporation of new data. It is a cheap and effective way of increasing the reliability of design. Another major strength of the geotechnical block model is that it crosses the departmental barriers between geotechnical engineering, blasting and processing.

8


Geotechnical data can be used to determine Lilly’s (1986) blastability index (BI) which is simply a rock mass rating for blasting personnel. It enables them to customise a blast pattern design based on the rock mass properties of the blast in question. This is used to optimise blasting and produce ideal fragmentation profiles required by the processing plant. All the geotechnical data necessary to calculate a BI is modelled in Datamine as described earlier. It is a simple step to calculate a BI per block which is then used to calculate an energy factor, drill and blast cost, burden and spacing that will produce the required fragmentation. A reworked Kuz-Ram equation is used to calculate the ideal energy factor while a correlation was developed to determine a drill and blast cost per cubic metre for PPRust (Bye, 2003). All the equations are written into a Datamine script where the user can also import any blast design boundary into the model and query the model for all the information within that boundary. A results file is created with the average model information for each rock type. The blast design results produced in the block model are only useful to those people who have access to Datamine. In order to make the data available to the survey, draughting, planning, drilling and blasting personnel, it is exported in csv format from Datamine and saved to a specified network location. From there it can be imported into AutoCAD, which is used for the draughting of all the blast requests and designs. Customised menus have been created by Peter Nathan of CGSS for importing the block model data and colour-coding it in the same way as in Datamine. The blasting department can then overlay the information on the official blast request plans (Figure 7) and adjust their design accordingly for every blast. A window appears (Figure 8) that shows the blaster the parameters used in the model to calculate the energy factor, burden and spacing and allows him to adjust these numbers. A blast design template also appears in the pattern boundary which he can rotate or translate. As he changes the blast design parameters, so the template changes on the screen. This window also shows the total number of holes that need to be drilled, the pattern area and the average UCS for the pattern. This is used by the drillers for drill bit selection and prediction of penetration rates based on rock type.

waste

ore

AutoCAD menus

blast pattern

Model filtered on EF Figure 7 Model bench slice filtered on energy factor (EF) with imported blast boundary overlaid and AutoCAD menus for importing and colour coding the model.

9


Figure 8 Customised blast design window and automated pattern design in AutoCAD

5

MINE TO MILL

Open pit mining involves a process of controlled destruction of the rock mass so that the waste may be stripped and the ore extracted. The blasting engineer is faced with the conflicting requirements of providing large quantities of well-fragmented rock for the processing plant, reducing drill and blast costs and minimising the amount of damage inflicted upon the rock slopes left behind. A reasonable compromise between the conflicting demands can only be achieved if the blasting engineer has a very sound understanding of the factors which control rock fragmentation, highwall damage and slope stability. This understanding was significantly enhanced through the use of the geotechnical block model. The definition of the fragmentation targets at PPRust was a lengthy process based on processing plant and face shovel loading requirements, and what was practical in the pit. The measure for the plant requirements is mean fragment size (P50 ) which is the sieve size that 50% of the fragmented material will pass through. The fragmentation profile is also of high importance as the plant uses autogenous mills which rely on the rocks to break each other into smaller fragments. They require a coarse feed (+125 mm) and a fine feed (-25 mm) for optimum performance. The particles between 25 mm and 125 mm, defined as the critical size, are discharged as pebbles and have to be mechanically crushed. Digital fragmentation analysis was done using Split-Desktop® which is an image-processing program designed to calculate the size distribution of rock fragments by analyzing digital greyscale images taken by a digital camera in the field (Split®, 1999). Split-Desktop® software ‘provides an economical alternative to manual sampling and screening and an objective quantitative measure rather than subjective qualitative estimates’ (BoBo, 2005). Eighteen blasted muckpiles from a wide range of powder factors (0.94 – 2.28 kg/m 3 ) and hole diameters (171 mm – 311 mm) were analysed to determine the fragmentation profiles from both ore and waste. The percentage of coarse and fine fragment sizes is important for the autogenous mill as breakage is a result of particles colliding. The results showed that a mean fragment size of 150 mm for ore should be delivered to the primary crusher (-250 mm) on the way to the plant. This target was met by increasing

10


the energy factors and fine tuning the blast designs, which shifted the ore fragmentation profile (Figure 9). The blasting costs in 2003 increased from R1.31 per tonne to R1.89 per tonne of rock which totalled R231 538 per month. The cost of rock fragmentation by blasting equates to only 10 -15 % of the overall comminution costs of the operation at PPRust. This extra cost is far outweighed by the savings gained in the plant. The crushing costs in 2003 were reduced from R0.72 to R0.31 per tonne of ore which equates to a cost saving of R168 877 per month for the primary crusher alone. The plant milling rates are largely dependant on feed size and hardness. They were therefore increased by 8.8% due to the finer fragmentation. This accounted for R2.1 million per month in additional revenue (Bye, 2003). 100

(251) 5.5 x 6.3 90

(251) 6.5 x 7.5

80

70

(251) 6.0 x 6.9 60

g n i s s 50 a P % 40

(251) 5.0 x 5.8

(271) 5.0 x 5.8 30

(271) 6.0 x 7.0

20

10

0 1

10

100

1000

Fragment Size (mm)

Figure 9 History of ore blast fragmentation curves at Sands loot pit

A primary measure for the Load and Haul efficiencies is instantaneous loading rate (ILR) which is the tonnes per hour loaded by a face shove l. It is recorded at PPRust b y the Modular Mining Truck/Shovel Dispatch® system. The ILR of 238 blasts were assessed to determine the ideal fragment size for loading. A target ILR of 3200 t/hr for waste and 3300 t/hr for ore was set based on the analysis. This was to coincide with the 150 mm mean fragmentation target of the plant and equated to a 230 mm fragmentation target for waste. With the implementation of the geotechnical block model, the instantaneous loading rates improved by 8.5% over a two-and-a- half year period. A more detailed analysis of all the customer performance measures was undertaken for the period from January to June 2003 (Figure 10). There is a clear improvement across all the performance indicators, which include the following: •

•

•

•

Average Average Average Average

plant milling rate (AG and Ball mills) - 18% improvement AG milling rate - 16% improvement instantaneous loading rate (Ore and Waste) - 13% improvement instantaneous loading rate (Ore) -11% improvement

These performance improvements represent a substantial value add to the overall business and the associated financial benefits are significant in terms of millions of Rand per month. The improvement in the autogenous grind milling performance for the

11


18-month model application period from January 2002 to June 2003 was recorded. The average milling rate for 2001 was 156 t/hr while this improved by 5.5 t/hr to an average of 161.5 t/hr in 2003. The additional revenue generated by this increased efficiency for the eighteen- month period was R30 million. This research illustrates the improvement in business efficiencies that has been realised at PPRust, not from a restructuring but by assessing the company’s total business process and defining a customer focus for the drill and blast department. This customer focus was facilitated by the use of a fragmentation model. 300

3,900 Average Plant Milling Rate (t/hr) [ 18% Improvement]

280

Average AG Milling Rate (t/hr) [ 16% Improvement] 3,700

Average Instantaneou s Load Rate Ore & Waste (t/hr) [ 13% Improvement] 260

Average Instantaneous Load Rate Ore (t/hr) [ 11% Improvement] r 3,500 h

240 r h / t ( 220 e t a R g n i 200 l l i M

3,300

3,100

180 2,900

/ t ( e t a R g n i d a o L s u o e n a t n a t s n I

160 2,700 140

120

2,500 Jan-03

Feb-03

Mar-03

Apr-03

May-03

Jun-03

Figure 10 Graph illustrating the performance of the drill and blast department’s two customers during 2003. The loading rates include ore and waste and the milling rates include both the AG and ball mills.

The use of the block model for blast design obviously has a big impact on plant efficiencies. There is also potential to use the block model for daily predictions on mill throughput. From the lab test results for UCS and DWT, a correlation with an 80% reliability was made between the two values. As UCS is interpolated into every block in the model, a DWT estimate is calculated for each block based on this formula. Also, the elastic properties of each rock type are known from lab testing and these are calculated into each block. This information can be used by the metallurgists to predict mill throughput rates and to prepare accordingly. Improved mill throughput increases the plant recovery and ounces produced. This has huge financial implications and the utilisation of the geotechnical data in this way is the next step that is being taken at PPRust. 6

FUTURE WORK

Mine design interrogates an ore reserve model to generate an economic mining shell using Datamine® NPV Scheduler or Whittle 4D®, which are based on the LerchsGrosman algorithm. The scheduling software proposes pit outlines and preferred areas of mining, based on grade and mining cost-per-tonne. At PPRust, and most other mines, a single mining cost-per-tonne is given for the entire pit, except for consideration of the

12


hauling costs related to pit depth. A practical pit is then designed around the economic mining shell, generated by the scheduling software, using an average stable slope angle. The schedule is based on the mining costs versus the grade and tonnage for each ore reserve block. A further application of the geotechnical model would be to use the scheduling software to generate economic mining shells based on the comprehensive geotechnical and mining information contained in the geotechnical model. As the geotechnical model has a stable slope angle contained in each cell, it is feasible that the scheduling software could generate far more accurate economic shells, based on this information, rather than a single slope angle for the entire pit highwall. Feasibility studies for new projects require an expected budget expenditure estimate to within 5% of actual costs. This level of accuracy is expected by mining companies and investors but cannot be achieved witho ut a sound knowledge of the rock mass variability. By developing a 3D geotechnical model as part of the feasibility study and using the same geological model exploration boreholes, a significantly better understanding of the rock mass conditions can be obtained. This rock mass information can be used for equipment selection, economic pit layouts, processing plant design, underground mining layouts and support requirements, to name but a few design parameters which are dependant on rock mass conditions. There is the potential for the application of geotechnical models to be equally successful in underground operations, as well as considerable scope for the implementation of these methods as a tool for mine evaluation and feasibility assessments of new ore deposits. The ore reserve model has gained widespread acceptance as an invaluable tool for a mining operation. Certainly most financial organizations will not invest in a mining project that does not have an ore reserve model. There is the potential for geotechnical models to be accepted as vital tools to the mining process, as is the case with ore reserve models. The development of a geotechnical model facilitates the provision of geotechnical information well in advance of the mining face. Using the model, mining areas were evaluated not only for grade and tonnage predictions but also for predictions of rock mass quality. Blast design and explosive requirements were derived from the rock mass quality predictions. This information was used for overall mine planning and evaluation, costing, production optimisation and slope design. This allows the full range of mining activities and costs to be inter-connected, thereby lowering costs and improving efficiencies through the application of the geotechnical model. The cumulative benefit delivered to the PPL operation by the development of the 3D geotechnical model was significant. The direct, quantified financial benefits were in the short term, in excess of R 29 million and in the long term over R 900 million. Other mining operations would be well served through the development of a similar geotechnically driven, business initiative.

7

CONCLUSIONS

A 3D geotechnical model has application to any major civil or mining venture that requires a detailed understanding of the variability in rock mass conditions. A geotechnical model does not propose to generate solutions by creating information from a limited data set. It does, however, give the engineer a tool whereby he can assess the

13


spatial variability of the rock mass information and thereby identify data-deficient or high risk areas. There are numerous case histories detailing the failure or significant over-expenditure of civil, tunnelling and mining projects caused by a lack of knowledge of the variability of the in-situ rock mass. Manual information systems used for design require significant dedication and time commitments and can be onerous to continually update. They often rely on the commitment of a single individual and are therefore not sustainable. The 3D geotechnical model is a user-friendly and sustainable tool, which can be readily updated and therefore does not suffer from the limitations of a manual system. The 3D geotechnical model provides information well ahead of the mining face, which can then be used for rock quality prediction, production optimisation, slope evaluation and design, as well as planning and costing. Using a similar query function as the ore reserve model, mining slots can be evaluated and not only grade and tonnage figures derived but predictions of penetration rates, powder factors, presplit and blast designs, as well as equipment and explosives requirements. The mining costs could be broken into drilling, blasting, crushing and milling costs, based on expected powder factors, penetration, crushing and milling rates, thereby further optimising pit planning and expenditure. More detailed costing and budgeting can be undertaken especially in respect to comminution and drill and blast costs. Steffen (1997) suggested classifying slope designs into three classes in the same way that geological models are classed, as specified by the SAMREC code. The three classes in order of increasing confidence are inferred, probable and proven. The geotechnical block model provides much of the information needed to qualify the resource as ‘proven’ which is the goal. The plan for the future is to incorporate all the geotechnical data available – groundwater, monitoring etc – into one model so that the slope can be designed with the highest confidence. There is the potential for a similar geotechnical programme and the application of a geotechnical model to be equally applicable to underground operations. Additionally, there is considerable scope for the implementation of these methods as a tool for mine evaluation and feasibility assessments of new ore deposits. The ore reserve model has gained widespread acceptance as an invaluable tool to a mining operation. Certainly most financial organizations will not invest in a mining project that does not have an ore reserve model. There is the potential that the same acceptance as a vital tool to the mining process will follow the development of geotechnical models. In the race for reduced mining costs and increased productivity, the development of a geotechnical model provides a cost-effective tool to improve the overall profitability of a mining operation. 8

REFERENCES

AP (2006) Anglo Platinum website. www.angloplati num.com Barton, N., Lien, R. and Lunde, J. (1974) Engineering classification of rock masses for the design of rock support. Rock Mechanics. 6, pp.189-236. Bieniawski, Z.T. (1976) Rock mass classification in rock engineering. Exploration for rock engineering . A.A. Balkema. 1, pp.97-106.

14


BoBo, T. (2005) What’s new with the digital image analysis software split-desktop®? JKTech website. www.jktech.com Bye, A.R. (2003) The development and application of a 3D geotechnical, model for mining optimisation, Sandsloot open pit mine, South Africa . Unpublished PhD Thesis, University of Natal. Cunningham C. V. B. (1986). The Kuz-Ram model for prediction of fragmentation from blasting . Proc. 1 st intern. Symposium on rock fragmentation by blasting, Lulea, Sweden. pp 439 -452. Haines, A. and Terbrugge P.J. (1991). Preliminary estimation of rock slope stability using rock mass classification systems. Proc. 7th Int.Cong. Int. Soc. Rock Mech. Aachen. Balkema, Rotterdam, 2, pp 887-892. Hoek, E. (2001) Rock Engineering . Course Notes. JKTech Pty Ltd. (2006) Company website. www.jktech.com.au Laubscher, D. H. (1990). A geomechanics classification system for the rating of rock mass in mine design. J. S. Afri. Inst. Min. Metall., 90. 257-273. Lilly, P. A. (1986). An empirical method of assessing rock mass blastability. The Aus.IMM-IE Aust. Newman Combine Group, Large Open-pit Mining Conference. October, 1986, pp 89-92. Steffen, O.K.H. (1997) Planning of open pit mines on a risk basis. J. S Afri Inst. Min. Metall. March/April 1997. pp 47-56. Stewart, R. M. and Kennedy, B. A. (1971). The role of slope stability in the economics, design and operation of open pit mines. Proc. 1st Symposium on Stability in Open pit Mining, Vancouver 1970. Published by A.I. M.E. New York, 1971: pp 5-21.

15

The Role of Geotechnical Block Models in Mineral

Recommend Documents