a systematic approach to the optimization of extra low

A SYSTEMATIC APPROACH TO THE OPTIMIZATION OF EXTRA LOW PROFILE (XLP) 267

IntroductionThe application of extra low profile (XLP) mining methodsin narrow reef platinum mines is a relatively new anddeveloping method within the South African platinummining industry; if found to be successful, the cost benefitsfar outweigh those compared with standard low profilemining methods due to substantial reductions in wastedilution. The demand from the mining industry tosimultaneously improve mine safety and optimize miningoperations is continually growing; mechanization ofoperations is often seen to be the combined solution toachieve the above objectives. In the past, customers andstakeholders often held the perception that any form ofmechanization would automatically yield the desiredresults. With added experience relating to mechanizedmining, the majority of the industry has come to realize thattailored mechanized solutions are necessary in order tomake mechanized mining productive, economical, andhence profitable. This requires an overall systems approachto be adopted focusing on all components within the valuechain of the mining system including machinery,infrastructure, layout, and resource management. XLPmining has the added complexity of attempting tomechanize within confined stoping widths of 1.2–1.3metres resulting in constrained movement for men andmachinery to work within as demonstrated in Figure 1.

Given the added complexity of the XLP mining systemand the associated costs involved with this added

complexity, it is imperative that production output isoptimized. This means that the operation needs to beobserved and analysed in terms of the overall layout andconfiguration, together with the machinery and personnelworking within the layout. These components of the systemand their requirements then need to be integrated andassessed in unison. In terms of XLP mining, theseconstraints can be identified and mitigated by ensuring:

• Goldratt’s Theory of Constraints in applied to allactivities associated within the XLP mine cycle,thereby addressing system bottlenecks which will havea direct influence on machine and subsequent mineoutput.

• Optimization of resource availability to improve overallresource utilization and reduce sunken costs associatedwith redevelopment, re-ledging, and non-optimalplanning.

The details below attempt to quantify the impact that theabove constraints have on overall section and system outputand attempt to address these issues using a systemsapproach.

The need to simultaneously mechanize andoptimize operations

Industrialization and automation within a range ofindustries, together with a shift in demand for labourtowards service-related industries, means that personnelwho are willing to work within confined underground

ANDREWS, M. and PICKERING, R.G.B. A systematic approach to the optimization of extra low profile (XLP) mine productivity for narrow reef platinummines. The 4th International Platinum Conference, Platinum in transition ‘Boom or Bust’, The Southern African Institute of Mining and Metallurgy, 2010.

A systematic approach to the optimization of extra low profile(XLP) mine productivity for narrow reef platinum mines

M. ANDREWS and R.G.B. PICKERINGSandvik Mining and Construction

Previous trials conducted by Anglo Platinum and Lonmin highlighted the importance of the seriesnature of XLP mining and the impact of addressing bottleneck constraints on overall mineproductivity and output. Prior trials completed at Anglo Platinum Bathopele Mine and Lonminidentified the performance of the XLP roof bolter as the primary system bottleneck. SandvikMining and Construction undertook to evaluate current XLP system performance, quantifyindividual machine performances and address bottleneck constraints, applying Goldratt’s Theoryof Constraints. The roof bolter was indeed found to be the primary system bottleneck and thispaper initially attempts to define the measures undertaken to improve XLP bolter performance andquantify the impact of addressing the bottleneck on overall mine productivity. Once XLP bolterimprovements were implemented, it was seen that the bottleneck had shifted to the LHDperformance within the cycle. By optimizing the loading and hauling activity, improvements wereexperienced, however, their effect was less pronounced than the XLP roof bolter initiative,implying that additional system constraints existed. Further studies and analyses highlighted panelavailability as the next primary system factor, directly influencing overall mine output. A revisedsection layout and configuration is therefore proposed which, if tested and implemented, couldlead to an additional 32% increase in system output. This paper therefore attempts to describe themeasures undertaken to eliminate a combination of machine and application bottleneck constraintsin current XLP mining system and highlights the importance of applying a systems approach toXLP mine optimization.

PLATINUM IN TRANSITION ‘BOOM OR BUST’268

environments are often limited. Add a cumbersome jacklegto the mix and people are even less inclined to performunderground, which means that demand and subsequentcosts within the conventional mining industry haveescalated substantially. The increase in HIV/AIDS withindeveloping countries exacerbates this problem further. Theintroduction of mechanized mining methods often alleviatesthe above problems by:

• Reducing the physicality associated with mineactivities.

• Improving the environment in which personnel operate(air-conditioned cabins for example)

• Substantially reducing the risk of injury and harm topersonnel by removing them from the working face andmechanizing a number of the activities

• Reducing labour requirements although personnelworking in mechanized environments often have tohave additional and scarcer skill sets to operate theequipment correctly.

Although the above are imperative prerequisites for anysocially conscious company, it would make no sense to gointo business unless the above could be achievedeconomically and optimally. The substantial reduction inPGM prices, attributed primarily to the decline in theplatinum price, as depicted in Figure 2, has forced miningcompanies to focus on efficiencies and effectiveness withintheir operations.

The shift from standard profiles to low profile mining andto extra low profile mining is one means of optimizing theextraction of PGMs due to a substantial reduction in stopingwidth; this significantly reduces dilution, therebyminimizing costs associated with handling and processingof waste material. This is emphasized in Figure 3,comparing XLP mining net operating profit after tax(NOPAT) to standard low profile and conventional miningNOPAT respectively. Figure 3 clearly shows that the

NOPAT associated with XLP mining is far higher thanstandard LP mechanized mining methods and conventionalmining methods respectively over the life of mine duration.

Figure 1. XLP roof bolting within a 1.2 m stoping width

Figure 2. Historical platinum price trends from 2005–20091

Figure 3. Net operating profit after tax2


The series nature of XLP miningThe series nature of XLP mining within the current mineconfiguration implies that each panel needs to have someactivity being performed at any given time. This isillustrated in Figure 4.

Figure 4 shows that;• One panel needs to be drilled with the XLP drill rig,

and charged for blasting.• The preceding panel needs to be cleaned using the XLP

dozer and the LP LHD.• The panel above should be supported with roof bolts

using the XLP roof bolter.• Secondary support needs to be installed at predefined

blast intervals using either wooden props or groutpacks. Although this can, at times, be performedsimultaneously with the activities above, it isrecommended that a fourth panel is made available,particularly if grout packs are installed.

• Ideally one additional panel needs to be available as abuffer to the system in case the cycle does not takeplace according to plan. It also allows development tooccur simultaneously within the advanced strike drives(ASDs) while simultaneously stoping the panels.

The cycle above provides a typical example of a systemin series and implies that any manageable system is limitedin achieving more of its goal by a very small number ofconstraints, and that there is always at least one constraintas depicted in Figure 5.

According to Goldratt, five focusing steps should beadopted to address such a system, defined as:

• Identify the constraint (the resource or policy thatprevents the organization from obtaining more of thegoal).

• Decide how to exploit the constraint (make sure theconstraint’s time is not wasted doing things that itshould not do).

• Subordinate all other processes to above decision (alignthe whole system or organization to support thedecision made above).

• Elevate the constraint (if required or possible,permanently increase capacity of the constraint; ‘buymore’).

• If, as a result of these steps, the constraint has moved,return to Step 1.

The above methodology was adopted for the XLP miningsystem in an attempt to optimize the efficiency of the XLPcycle as described below.

XLP activity bottleneck identificationA detailed time-and-motion study was conducted at AngloPlatinum’s Bathopele Mine by Sandvik Mining andConstruction’s trans4mine™ department;—a team whichhas been assembled as part of Sandvik’s BusinessDevelopment offering to assist mines in optimizing currentoperations from a systems and applications’ perspective.

Results of the time-and-motion study, conducted over atwo-month period, are shown in Figure 6. Note thatBathopele Mine operates three by eight hour shifts and it isassumed (for this study) that five of the eight hours iseffectively availably for productive operation of XLPequipment (indicated by the red line in Figure 6)

The results clearly show that the XLP roof bolter is theobvious activity bottleneck within the XLP cycle and,according to Goldratt’s theory of constraints, should beaddressed first if we expect any improvement in overallXLP system performance.

In order to understand the XLP bolter constraint further,each task within the XLP bolting process was measured andis illustrated in Figure 7.

Issues and identified solutions associated with optimizingeach of the tasks above were analysed and are summarizedbelow:

• The rod coupling procedure is cumbersome and time-consuming, contributing 21% additional time to thatassociated with a standard non-coupled bolt. Time isrequired to couple and screw the threaded R23 rodstogether, prior to relocating the drill string on the drifterand raising the feed back into the hole. This is a manual

Figure 5. Goldratt’s theory of constraints

Figure 4. The XLP stoping cycle

Figure 6. XLP activity durations


process, thereby compromising operator/assistantsafety (the assistant runs the risk of injury whencoupling the bolt and is in the vicinity of potentialrockfalls during the process).

• The need to use four rods in order to attain the 1.6mbolt hole length poses the same safety and timeconstraints as those discussed above with the couplingprocess. Any possibility to reduce couplingrequirements would substantially improve cycle timeand reduce the potential of personnel injury/harm.

• Due to the nature of the tramming system, positioningof the machine directly over the target where the bolt issupposed to be installed is difficult to achieve;tramming currently lacks proportionality, which makesmachine movement relatively ‘jerky’ and difficult tocarry out to the required levels of accuracy.

• The bolt installation process is completely dependenton the XLP roof bolter due to the added requirement ofinserting, spinning, and securing the bolt; the forcesrequired with resin bolts requires mechanical assistanceto perform these functions. The injection andinstallation process therefore reduces the machinescapacity to drill by 36%.

XLP bottleneck optimizationIn order to address the above constraints associated with theXLP roof bolter activities, the following initiatives wereimplemented:

Drilling optimization A new XLP roof bolter feed was designed with a longerstroke ensuring that three rods could achieve the desired1.6m hole length instead of the four rods which werepreviously required (refer to Figure 8).

The revised design resulted in a reduction in drilling timeof 27 seconds per bolt (19%) and a 6% improvement inoverall bolt cycle time. Incidentally, the feed was alsomodified from a chain feed to a rope-cylinder feed which, todate, has yielded a 50% improvement in reliability.

Bolt installation optimization:A decision was taken to attempt to ‘de-couple’ the drilling ofthe bolt hole from the installation of the bolt. Afterconsidering several possibilities, the selected solution was totest New Concept Mining’s Hydrabolt™, a type of frictionbolt which encapsulates water (shown in Figure 9). The boltis manually installed using a separate water pump and can becarried out after the XLP roof bolter has completed drillingthe hole, allowing the XLP machine to proceed to the nexthole and start drilling while the previous hole’s bolt isinstalled. Although the implementation the Hydrabolt™ doesnot necessarily optimize the number of personnel at theworkface, it is viewed as an interim solution, whilst amechanized solution is being developed.

Implementation of the above resulted in a reduction inoverall XLP roof bolter cycle time of 156 seconds per bolt(35% improvement in bolt cycle time).

Rod coupling optimisation In order to reduce coupling time associated with theextension drilling process, a newly designed drill rod andcouplings were developed termed the Infinity™ range(shown in Figure 10). The Infinity™ drill steels were namedsuch due to their figure-eight coupling shape which wasrelatively easy to manufacture and far easier to couple duringthe drilling process of the bolt hole. The introduction of thissteel resulted in a 61-second improvement in overall cycletime per bolt, representing an overall improvement of 13% tothe bolt cycle time.

Machine/boom movement optimizationIn order to accurately and efficiency position the machineunder the required hole, proportional tramming wasdeveloped allowing for precise location of the machine over

Figure 7. XLP roof bolter activity times

Figure 8. Revised rope cylinder feed with longer stroke length

the bolt hole with minimal effort. This is currently beingtested at Anglo Platinum Bathopele Mine. However,provisional results to date show an overall improvement of10 seconds per bolt (2% improvement in overall cycle time).

The above initiatives are summarized in Figure 11,highlighting the improvement in each activity relative to theprevious XLP roof bolter configuration:

Applying the above to all activities within the XLP cycle,it can be seen that the XLP roof bolter has shifted from thebottleneck activity to the best performing activity within thecycle (Figure 12). The optimization improvements statedabove should also yield a 30% improvement in overall fleetproductivity attributed to these improvements.

According to Goldratts’ theory of constraints, the nextactivity that needed to be addressed was the LHD machine

as this had shifted to the new bottleneck activity. This wasrelatively easily remedied as Sandvik Mining andConstruction had just recently tested and implemented anew 8 tonne LHD (6.4 tonnes with ejector bucket asrequired by XLP mine layouts) as shown in Figure 13.

The above prototype was run as a trial at Anglo PlatinumBathopele Mine and yielded an improvement from 3.9tonnes/bucket to 5.2 tonnes/bucket (median values) whilemaintaining the same number of loads as the smaller EJC115 LHDs. This resulted in an overall loading cycleimprovement of 33%. It is important to note that the mine isconsidering using only one LH208L in place of the twosmaller EJC 115 LHDs, implying that the defined 33%improvement will not be realized; however, mine capital andoperating cost outlays for LHDs will essentially be halved.Assuming for consistency purposes that two LH208L LHDsare used (as per the previous system definition), the overallsystem improvement is illustrated in Figure 14.


Figure 9. Schematics of the Hydrabolt™

Figure 10. Infinity™ drill steels

Figure 11. Previous and optimized activities relating to the XLProof bolter

Figure 12. System analysis with XLP roof bolter improvements

Figure 13. LH208L LHD prototype tested at Anglo PlatinumBathopele Mine

Figure 14. System analysis with LHD improvements


The above machine fleet improvements can besummarized in terms of expected blasts per day andexpected square metres achieved on a daily basis as definedby Table I.

Although substantial improvements were initiallyachieved with the optimization of the XLP roof bolter, itcan be seen that incremental improvements thereafter wereseen with the introduction of the larger LHD; the reason forthe small improvement in overall system performance isdue to the XLP face rig almost immediately becoming thenew bottleneck with a small improvement in LHDperformance despite the substantial overall performanceimprovement with the introduction of the larger LH208LLHD. This is expected to occur due to the fact that therange between the bottleneck machine and the bestperforming machine is beginning to converge. It is alsoobvious from Figure 14 that although all machines arecapable of comfortably achieving their shift targets withinthe predefined effective shift time other system constraintsoutside of machine capabilities alone may be hamperingsystem output, implying that machine performance alonemay not be the ultimate system bottleneck.

Mine configuration and layout optimisationThe above results demonstrate the application of Goldratt’sTOC model while focusing on the XLP machine fleetalone. One limitation with the above method is that itassumes infinite system capacity in terms of panelavailability. The unpredictable nature of undergroundmining means that production is often lost due to panelsbeing ‘lost’ as a result of poor ground conditions such asfaults, dykes, and potholes. This is closely related to thecurrent mine and section configuration which is brieflysummarized below, together with identified constraints. Arevised mine and section layout is then proposed intendedto address some of the key constraints associated with thelayout.

Current mine and section configurationAnglo Platinum has defined a comprehensive set of minedesign criteria (MDC) which provide guidelines for XLPmining for dip angles of 10º and below. In general, theoverall XLP layout consists of 4 strike sections, separatedby a five-set decline as indicated in Figure 15.

It is intended that three of the four strike sections are inproduction at any given time with the fourth strike sectionbeing upgraded and advanced with activities such as serviceinstallations, belt extensions and borehole radar drilling.

Each strike section consists of nine panels (36 metres inlength, skin-to-skin) with two suites of equipment operatingwithin each strike section. An additional ‘swing suite’ ofequipment is provided to ensure that production continueswhile machines are being serviced. A comprehensive

conveyor system is located at the centre of the strike sectionwith an extensive set of dip belts and fish belts to reduceLHD tramming distances and a satellite workshop isavailable for machine services. A schematic of each sectionis provided in Figure 16.

System constraintsMajor system constraints associated with the MDC layoutare summarized below:

• The need to develop a 9-panel section prior toproduction commencing requires substantialdevelopment to occur within the strike section whichincurs significant time and represents an opportunitycost.

• A substantial amount of decline development isnecessary to reach the bottom of the four strikesections, which again increases the ramp-up period tofull production.

• Lack of predevelopment and use of borehole radarimplies limited orebody knowledge, resulting in feweroperational panels due to geo-losses.

• Re-raising beyond potholes is expensive and time-consuming.

• The conveyor configurations are expensive and time-consuming to move.

• The above introduces the need for an additional strikesection to be available for preparation andestablishment. This is expensive and has no immediatereturns in terms of revenue.

The above constraints all contribute to a decrease in thepossibility of optimizing panel availability.A summary ofpotential panels available versus panels lost per section issummarized in Table II for two of Anglo PlatinumBathopele Mine’s XLP sections from August 2009 to April2010. Note that panels were lost for a range of reasons andmay have not have been available for only a period withinthe defined month.

Table ISystem improvements attributable to machine cycle optimization

Base case Roof bolter optimization Roof bolter and LHD optimisationAvailable hours/day (h) 15 15 15Bottleneck activity XLP roof bolter LHD XLP face rigBottleneck duration (h) 5.6 4.3 4.2Achievable blasts/day 2.67 3.49 3.57m2 per blast 50.16 50.16 50.16Achievable m2/day 134 175 179% improvement - 31% 34%

Figure 15. XLP macro-layout1


It is interesting to note the following observations fromTable II:

• On average 32% of all potential panels were notconsistently available to be stoped over a 9 monthperiod; this represents an average of approximately 2‘unproductive’ panels per section at any given timewithin each section per month.

• Based on Figure 4, it can be seen that a minimum of 4panels per section is required to optimally carry outXLP mining; shaded examples in Table II show thosemonths where fewer than 4 panels were continuallyavailable for stoping, implying that the XLP cyclecould not be optimally implemented (1 out of 9 monthsfor Section A versus 5 out of 9 months for Section B).

• The positive impact of having an additional panelwithin an XLP section (Section A with 6 potentiallyavailable panels versus Section B with 5 potentiallyavailable panels) can clearly be seen when observingthe number of occurrences where XLP cycles cannot be

effectively implemented (shaded examples). Onaverage a six-panel section can be effectively stoped89% of the time (8 out of 9 months) whereas a five-panel section can only achieve an optimal cycle 44% ofthe time (4 out of 9 months). This will more than likelybe case specific for different sections. However, it doeshighlight the relationship between section size, panelavailability, and the ability to continually achieve theXLP cycle.

The wish listIn order to address the following constraints, and panelavailability in particular, the following wish list wasdefined:

• Provide a similarly sized section to the MDC solutionwith reduced development requirements (metres andcost) and an equal (or greater) number of panels persection.

Figure 16. Strike section layout defined by Anglo Platinum

Table IIAvailable and lost panels per XLP section

Section A Section B TOTALMonth Total panels Panels lost Total panels Panels lost Total panels Panels lost % Panels lostAug-09 6 3 5 1 11 4 36%Sep-09 6 2 5 2 11 4 36%Oct-09 6 2 5 2 11 4 36%Nov-09 6 2 5 2 11 4 36%Dec-09 6 2 5 1 11 3 27%Jan-10 6 0 5 1 11 1 9%Feb-10 6 1 5 1 11 2 18%Mar-10 6 2 5 3 11 5 45%Apr-10 6 2 5 3 11 5 45%Total average 32%


• Reduce panel establishment time to increase net presentvalue (NPV) and internal rate of return (IRR) results.

• Promote extended orebody knowledge prior to mining(predevelopment) to promote reliable section designsand subsequently optimise mine planning.

• Ensure all sections which have been developedareproducing at any given time.

• Limit panel loss (or its effects) due to geotechnicalconstraints.

• Provide a conveyor system which is easy to extend anddoes not impact the mining operation during transfers.

The proposed solutionIn order to achieve the above wish list and eliminate thedefined constraints, a revised mine layout was developedwith equal or improved criteria as opposed to the MDCparameters. Summary illustrations of the refined strikesection and mine layout are shown below and discussed infurther detail:

• It is proposed that each strike section is adaptedcomprising four panels with skin-to-skin panel lengthsof 36.4 m as per the MDC specifications.

• Each strike section would be predeveloped, two raiselines ahead, to ensure that ventilation is optimized andthat no vent columns are located near to productionblasting. Predevelopment will also increase orebodyknowledge allowing for optimized planning andscheduling to occur.

• By predeveloping two raise lines ahead of stoping,panels can be stoped from both sides, therebyconverting a typical four panel strike section into aneight-panel equivalent section.

• A conveyor would be located at the top and bottom ofeach strike section with allocated locations for stopingand development respectively. This will reducetramming distances for LHDs and allow belt upgradesto occur independently.

In order to ensure that production output from the mineequals or improves on the existing MDC model, it wasdecided that the number of panels should equal or exceedthe MDC criteria from a macro-scale. It is thereforesuggested that 3 sets of the strike sections defined in Figure17 are combined to form the overall production layout asillustrated in Figure 18.

Figure 18. Proposed macro mine layout

Figure 17. Proposed strike section layout


It is suggested that a single dedicated suite of equipmentis allocated to each strike section with an additional ‘swingsuite’ to be shared between pairs of strike sections assummarized and compared in Table III.

The suite complement comparison above emphasises thefact that the revised mine design has exactly the same fleetsize as the MDC model implying that the revised minedesign will not compromise capital and operating costsassociated with equipment.

A comparison of MDC system parameters and proposedPredevelopment design parameters are listed in Table IV.

Revised system benefitsBenefits of the revised system are summarized as follows:

• Each suite of equipment can operate within an eight-panel layout. Even with panel losses of 32% ascalculated above, this amounts to a minimum of 5panels per suite, which is ideal for XLP machine andprocess cycling.

• Once the macro-section has been developed, each strikesection is continually in production ensuring thatrevenue is maximized in terms of NPV and IRR.

• The inclusion of 4 dedicated and defined tipping pointsfor production and development ensures that trammingdistances are always maintained below 120 metres (oneway).

• The belt configuration allows for belt upgrades to takeplace without interrupting production.

• Predevelopment greatly improves orebody knowledgeand subsequent production planning, potentiallyimproving production by 25% plus.

• The stoping and development activities take placeindependently in separate sections within each strikesection, ensuring that interference does not occurbetween these functions.

• Less decline development metres are required,however, more strike and raise development will benecessary to predevelop each section.

• Longer section length along the strike direction (90 mversus 45–60 m) implies less raise line development,which leads to a lower overall development ratio.

Revised system constraintsThe primary argument with the revised model is theadditional development (strike and raise) necessary to pre-develop the section as apposed to stoping and developingsimultaneously within the same panel region. It is estimatedthat the MDC ramp-up to steady state takes approximately17 months whereas the predevelopment model will takeapproximately 23.5 months to full production. It is,however, felt that the impact on NPV and IRR attributed tothe longer lead time to full production would be faroutweighed by the proposed improvement in productionoutput attributed to the machine cycle and panel availabilityimprovements stated above. Several alternatives do,however, exist to improve this ramp-up period stated abovewhich will ultimately improve NPV and IRR results:

• Predevelopment could take place to the point of oneraise line ahead (as apposed to two raise lines asoriginally proposed). This would reduce the ramp-upperiod by approximately 6 months, aligning targetswith the MDC model. It would, however, complicatethe ventilation requirements.

• The inclusion of an additional two development rigs,shared amongst all six strike sections, will reduceramp-up time by 6 months. Introducing the LH208L inplace of two EJC 115 LHDs, may justify the inclusionof two additional rigs while maintaining pre-development two raise lines ahead of stoping.

The base case predevelopment model therefore needs tobe tested further using an NPV and IRR model and ifidentified to be uneconomical, then the above twoalternatives could be reassessed further.

ConclusionsThe above analysis highlights the importance of consideringmine system constraints when attempting to optimizemechanised machine solutions. A substantial expense can beincurred by focusing on perceived constraints which may notultimately form the overall system bottleneck. It is thereforeessential to assess each mine as a system, step through theprocesses involved utilizing a process mapping approach,and apply a methodology similar to Goldratt’s theory ofconstraints in order to optimize a particular operation. Theinitiatives highlighted above attempt to optimize machinecycles by analysing the activities associated within the cycleas well as the environment in which the machinery operates.Although several of the machine cycle constraints have beenaddressed, additional effort is still required to fully justify theproposed mine solution to ensure that it suitably addressessafety, operational and economic constraints. It is felt that theincorporation of such a layout will, however, ultimatelyalleviate the next hurdle/bottleneck related to panelavailability, thereby shifting the bottleneck and providing theopportunity for further optimization to take place.

References

1. JOHNSON MATTHEY INSTITUTE. PlatinumToday. Available from: http://platinum.matthey.com/pgm-prices/price-charts/ [Accessed 08 May 2010].

Table IVSystem parameters MDC vs. Pre-developed design

MDC Revised design design

Number of strike sections 4 6Number of strike sections in production 3 6Panels per strike section (100% available) 9 8Panels per strike section (33% geo loss) 6 5Panels per suite of equipment 4.5 8(100% available)Panels per suite of equipment (33% geo loss) 3 5Total panels in production per macro-layout 27 48(100% available)Decline development metres for 668 501macro-layout (m)

Table IIIMachine suite requirements - MDC vs proposed

MDC Pre-design developed

designNumber of strike sections 4 6Number of strike sections in production 3 6Number of dedicated suites/strike section 2 1Number of swing suites/strike section 1 0.5Total number of suites of equipment 9 9


2. PICKERING, R. and MOXHAM, K, The developmentand implementation of the Lonmin mechanised breastmining. Journal of South African Institute of Miningand Metallurgy, 2007. vol. 107, no. 1, pp. 05. Availablefrom: http://www.saimm.co.za/ component/ fabrik/

details/3/3/2349. [Accessed 22 June 2010].

3. CHING, C. Introducing Goldratt’s Theory ofConstraints. 2004. Available from www.eligoldratt.com. [Accessed 10 June 2010].

Michael AndrewsTrans4mine Project Manager, Sandvik Mining and Construction.

I have worked for Sandvik Mining and Construction for 6 years during which time I worked as aTechnical Project Manager, an Account Manager and a Sandvik trans4mine Systems Engineer andtrans4mine Project Manager. I formed part of the team responsible for developing andimplementing the Sandvik trans4mine department and offering, a global consulting departmentassisting mines in optimising productivity and reducing operational costs. I have run trans4mineprojects locally in South Africa, in Africa (Zambia), and in Europe (Bulgaria).

a systematic approach to the optimization of extra low

Documents