minerals Article Alternative Process Flow for Underground Mining Operations: Analys Conceptual Transport Methods Using Dcrete Event Simulation Jenny Greberg 1, Abubakary Salama 2, Anna Gustafson 1 Bartłomiej Skawina 1, * 1 Department Civil, Environment Natural Resources Engineering, Divion Mining Rock Engineering, Luleå University Technology, Luleå SE-971 87, Sweden; jenny.greberg@ltu.se (J.G.); anna.gustafson@ltu.se (A.G.) 2 Department Chemical Mining, College Engineering Technology, The University Dar es Salaam, Dar es Salaam P.O. Box 35901, Tanzania; asalama@udsm.ac.tz * Correspondence: bart.skawina@ltu.se; Tel.: +46-920-492-949 Academic Editor: Michael Hitch Received: 25 December 2015; Accepted: 13 June 2016; Publhed: 30 June 2016 Abstract: As near surface deposits are being mined out, underground mines will increasingly operate at greater depths. Th will increase challenges related to transporting materials from deeper s to surface. For many years, ore waste transportation from most deep underground mines has depended on some or all following: truck haulage, conveyor belts, shafts, rails, ore pass systems. In sub- caving, where ore passes are used, trains operating on main lower transport ore from ore passes to a crusher, for subsequent hoting to surface through shaft system. In many mines, use ore pass system has led to several problems related to ore pass availability, causing dturbances incurred cost time for ore pass rehabilitation. These dturbances have an impact on mining activities since y increase operational costs, lower mine throughput. A continued dependency on rock mass transportation using ore passes will generate high capital costs for various supporting structures such as rail tracks, shaft extensions, crushers for every new main. Th study was conducted at an exting underground mine analyzed transport ore from loading areas at lower s up to exting shaft points using trucks without employing ore passes. The results show that, when costs extending ore passes to lower s become too great or ore passes cannot be used for, haul trucks can be a feasible alternative method for transport ore waste up ramp to exting crusher located at previous main. The use trucks will avoid installing infrastructure at next main extending ore passes to lower s, hence reducing costs. Keywords: rock mass transportation; haulage system; trucks; ore pass; dcrete event simulation; sub- caving 1. Introduction As resources near surface are being mined, underground mines worldwide are operating at increasing depths. The cost time required for rock mass transportation increases with increased mining depths, with aim to increase mined volumes reduce costs, rock mass transportation from deeper s to mine surface presents challenge for underground mines. Despite extence availability several haulage options, in most deep underground mines depends on safe continuous operation ore pass system. Ore passes are used for material transportation, can also serve as a means to store ore in underground Minerals 2016, 6, 65; doi:10.3390/min6030065 www.mdpi.com/journal/minerals
Minerals 2016, 6, 65 2 14 mines [1]. However, in order to design a well-functioning ore pass, it necessary to examine entire ore hling system mine, from areas to shaft points [2]. In sub- caving, ore mined on each sub- from hanging wall to forefront footwall, starting with overlying sub-s proceeding downwards. As ore mined from a sub-, hanging wall collapses covers mining area with broken waste rock [3]. The ore transported to desired destination by eir use ore passes, or by or hauling methods. When ore passes are used, trains operating on main transport ore from ore passes to crusher, for subsequent hoting to surface through shaft system. Th incurs a high capital cost for various structures such as rail tracks, shaft extensions, crushers for every new main. Despite extence design guidelines, ore pass problems, including hang-ups wall failures still perst in underground mines. These problems result in threats to safety personnel, loss, an increase in costs repairs [4,5]. Even when ore passes are carefully designed, ore pass system susceptible to several problems over its design life, including stability, wear, material flow [6]. These problems increase delays in mining operations [7], delays have an impact on mining activities since y increase operational costs reduce mine throughput. A study more than 200 ore passes conducted in South African mines observed that more than 50 percent se ore passes had stability problems, 16 percent m had been aboned [8]. Several methods can be employed to restore ore passes, such as use long-hole drilling blasting, flushing a blocked ore pass with water, pushing rods, use appropriate block size infrastructure to prevent passage oversized boulders [9]. In most cases, actual cost restoring ore passes extremely high compared with initial cost development [8]. If ore pass restoration not possible, ore pass refore will not be used for, alternative transportation methods must be considered. The aim th study to evaluate a conceptual future haulage method at an exting underground sub- caving mine by comparing different haulage equipment various sizes operating from drifts to crushers, enabling rock mass transportation without using ore passes. Dcrete event simulation was used for analys, study based on data from an operating mine. 2. Haulage Systems for Deeper Levels Exting transport methods that are feasible for deep mines include truck haulage, shaft systems, conveyor belts. Trucks are highly flexible in travel routes fleet size, provide high productivity [10]. Shaft systems may be inflexible because ir limited number fixed feed points, but once installed y fer low operational costs [11 13]. Conveyor belts typically provide most economic method for material transportation due to ir high carrying capacity [13]. The choice conveyor parameters influenced by nature material to be conveyed, available tunnel space, overall economics system. Normally, all se various haulage systems work in combination with or ore-hling components, such as Load-Haul-Dump (LHD) machines, ore passes crushers. The haulage process one most cost-intensive activities in a mining operation, thus, one main contributors to operational costs [10,14]. The expected increase in transportation costs for deeper mines makes choice hauling equipment essential when seeking cost reductions for deeper operations. The high cost developing a new main when a mine operates at great depth necessitates technological innovation an alternative haulage system. Frequent problems with ore pass failures, significant costs involved in managing se ore passes, make continuation current ore hling approach questionable. While se challenges continue, mining operations seek to optimize haulage system so that desired rates are achieved on time, at minimum cost. Th study evaluates use trucks to transport ore from loading areas directly to dumping point at shaft without using ore passes. The use trucks will help avoid new main development infrastructure costs, hence reduce capital costs.
Minerals 2016, 6, 65 3 14 Minerals 2016, 6, 65 3 14 3. Simulation Haulage Systems Underground haulage systems are complex, involve high investment costs. For many years, 3. Simulation analytical methods Haulagehave Systems been used to evaluate performance system. These methods produce Underground useful information haulage systems on are system complex, being analyzed, involve high but investment ir application costs. For many limited; years, for example, analytical methods when have systems been used involve to evaluate non-steady performance state conditions, system. or when These operations methods produce involve uncertainties useful information [14 16]. on Under system se being conditions, analyzed, an but alternative ir application method for limited; analyzing for example, a system when dcrete systems event involve simulation. non-steady The rom state conditions, dynamic or when nature operations involve haulage uncertainties systems [14 16]. mining operations Under se makes conditions, m very an alternative difficult to method model using for analyzing analytical amodels. system For dcrete that reason, eventin simulation. th work The we have rom chosen to dynamic use dcrete nature event simulation haulageas systems technique mining to evaluate operations makes mine operations. m very When difficult simulation to model using employed, analytical model models. input Forcan thatbe reason, based inon th appropriate work we have probability chosen todtributions use dcrete that event characterize simulation as input technique variables. tothe evaluate process mine designing operations. a model When a real simulation system usually employed, takes into model account input can a set be based assumptions appropriate for probability operating system dtributions [17]. These that characterize assumptions are input expressed variables. in The mamatical, process logical, designing a model symbolic arelationships real system usually between takes objects into account asystem set assumptions interest, for y operating can be solved system analytically [17]. These or assumptions by using simulation. are expressed With in mamatical, aim optimizing, logical, improving, symbolic analyzing relationships between planning objects exting system future systems, interest, various y mining can be operations solved analytically can be analyzed or by using simulation. These With include: aim fleet optimizing, requirements, improving, mine analyzing scheduling, planning mine planning exting [14,17]. There futureare systems, a number various simulation mining operations tools available can bein analyzed market, using such simulation. as SimMine, These SLAM, include: SIMAN, fleet ARENA, requirements, AUTOMOD. mine scheduling, In th study, mine planning SimMine [14,17]. simulation There arestware, a numberwhich simulation based tools on dcrete availableevent in market, simulation suchprinciples as SimMine, SLAM, has been SIMAN, developed ARENA, specifically AUTOMOD. for Insimulation th study, mining SimMine operations simulation [18,19], stware, used. which based on dcrete event simulation principles has been developed specifically for simulation mining operations [18,19], used. 4. Case Study 4. Case Study 4.1. Mine Description 4.1. Mine Description Th study was carried out in one largest underground mines in Sweden. The mine consts Th study a high-grade was carriedmagnetite out in onedeposit largest approximately underground four mines kilometers in Sweden. long The running mine consts in norasterly a high-grade direction, magnetite with deposit an average approximately thickness four kilometers between 80 long running 100 m. in The mine norasterly uses sub- direction, caving with anmethod average (see thickness Figure 1). between In th 80method, 100 m. development The mine uses drifts are sub- opened caving first, followed method (see by Figure drilling 1). In th method, ore passes. development The ore passes drifts are extend opened vertically first, followed from bycurrent drilling mining area ore down passes. to Thebottom ore passes a extend new mining vertically area, from where a current transportation mining area down located. to bottom Horizontal a sub-s new miningare area, created, where aincluding transportation crosscuts that located. provide Horizontal access sub-s to are created, drifts. including The self-supported crosscuts that provide crosscuts access are to drilled through drifts. orebody, The self-supported perpendicular crosscuts to are access drilled routes. through The spaces orebody, between perpendicular sub-s are to about access 28.5 routes. m, while The spaces between sub-s crosscuts are 25 about m. 28.5 At m, while crosscuts, spaces near-vertical between crosscuts rings are holes 25are m. At drilled crosscuts, in a fan-shaped near-vertical pattern. rings Each ring holes contains are drilled around in a 10,000 fan-shaped t ore pattern. waste. EachThe ringore contains mined around on each 10,000 sub-, t orestarting waste. with The overlying ore mined sub-s on each proceeding sub-, starting downwards; with overlying each sub-, sub-s ore proceeding removed downwards; from hanging eachwall sub-, to forefront removed footwall. from As hanging ore wall mined to from forefront a sub-, footwall. hanging As wall ore collapses mined from by design, a sub-, covers hanging mining wall collapses area with bybroken design, waste covers rock [3]. mining area with broken waste rock [3]. Figure 1. Sub- caving mining method (Courtesy Atlas Copco). Figure 1. Sub- caving mining method (Courtesy Atlas Copco).
Minerals 2016, 6, 65 4 14 4.2. The Current Haulage System in Mine The mine divided into 10 main areas, called blocks, which extend from uppermost mining down to current main. Each mine block currently consts 10 sub-s. Each block 400 to 500 m in length, has its own group ore passes located at center area extending down to main haulage. Currently, main haulage at 1365 m. Mining continues in each block using electric LHDs with a capacity 25 t that operate from 6 a.m. until 10 p.m., while semi-automated diesel LHD machines with a capacity 21 t operate from 10 p.m. to 6 a.m. Blasting normally occurs at around 00:00 every day. The LHDs load ore from draw points within each drift, transport ore to ore passes. Large trains, operating on main, transport ore from ore passes to a crusher. The crushed material stored in ore bins, after that transported on a small conveyor belt to hoting system. To conform to mining restrictions, once mining begins in a block, it must be maintained until all available ore removed before starting to mine next one. The current loading operation from each block results in an average daily 6035 t. The future plan to increase to 37 Mt crude ore per year from all ten blocks, which means a daily 10,000 t from each block. 4.3. Conceptual Haulage System The next main at mine will be at depth 1685 m, followed by anor 2005 m deep. With increasing depth, stresses increase, creating higher rk for geomechanical problems such as ore pass failures. If ore pass restoration not possible, ore passes cannot be used for, alternative transportation methods have to be applied. In th study, a conceptual truck haulage system for transport material from areas to crusher on previous main located at 1365 m was modeled simulated. By using a truck haulage system, installation a rail-mounted track system can be avoided hoting system will not have to be extended, thus reducing infrastructure costs. 4.4. Simulation Model In order to identify number size trucks needed to reach future target, a simulation model was developed. The model formulation was done using SimMine simulation stware (SimMine AB, Malå, Sweden), which uses a full graphical user interface for model set up. It utilizes stattical dtribution functions to model variations in process times. For verification purposes, to increase understing, tool has a three-dimensional environment that fers animated vual feedback model, allowing viewing dynamic system as it operates [19,20]. 4.5. Model Settings The conceptual mine layout modeled in simulation shown in Figure 2. The modeled area consts current main (1365 m) where crusher located, a series ramps connecting areas with main. In th study, two simulations run, can be described as follows: 1. Two sub-s (located closest to next main at 1685 m) with two areas in each was considered. These two s selected for modeling since y are s located furst away from crusher on previous main, hence require largest number trucks most time spent in queuing. These can be seen in Figure 3 as sub-s 9 10. 2. Two or sub-s (s 14 15) situated below main at 1685 m also simulated. Th was done to analyze effects furr increasing dtance to exting crusher at main 1365 m.
Minerals 2016, 2016, 6, 65 65 Minerals Minerals 2016, 6, 6, 65 14 555 14 14 Figure 2. 2. The conceptual conceptual mine layout. layout. Figure Figure 2. The The conceptual mine mine layout. Figure 3. The area considered in th study. Figure Figure 3. 3. The The area area considered considered in in th th study. study. For For sub-s sub-s above above main main at at 1685 1685 m, m, modeled modeled areas areas (indicated (indicated as as For sub-s above2) main for at 1685 m, modeled areas (indicated as areas 1, 2, 3, 4 in Figure chosen purpose truck modeling simulation. areas 1, 2, 3, 4 in Figure 2) chosen for purpose truck modeling simulation. areasto1,mining 2, 3, 4 in Figure 2) chosen for number purpose truck modeling used simulation. Due Due to mining restrictions, restrictions, th th maximum maximum number areas areas that that can can be be used at at Due to mining restrictions, th maximum number areas that can be used at same same time. time. Areas Areas 11 33 are are in in same same,, which which for for th th study study called called 9, 9, same time. Areas 1 4 3 on are in same, which for th 10. study also called 9, while while areas 2 are below, which here called We simulated four while areas 2 4 are on below, which here called 10. We also simulated four areas 2 4areas are on below, which here called 10. We also simulated four areas (two (two at at each each ) ) for for s s 14 14 15. 15. The The vertical vertical dtance dtance between between two two areas (two at each ) for s 14 15. The vertical dtance between two s s s 28.5 28.5 m. m. 28.5 m. Each Each modeled modeled areas areas in in simulation simulation consts consts 17 17 s s drifts drifts (see (see Each modeled areas in simulation consts 17 s drifts (see Figure Figure 3), 3), total total expected expected amount amount ore ore to to be be mined mined being being 6.17 6.17 Mt. Mt. According According to to mine s mine s Figure 3), total expected amount ore to be mined being 6.17 Mt. According to mine s regulations, regulations, at at 10 10 starts starts when when 99 75 75 percent percent completed. completed. Th Th means means that that regulations, at 10 starts when 9 75 percent completed. Th means that 1.54 Mt 1.54 Mt remains to be hauled from 9, while 6.17 Mt will be hauled from 10. There 1.54 Mt remains to be hauled from 9, while 6.17 Mt will be hauled from 10. There no no remains to be hauled 10 fromcommence 9, while 6.17 Mt will be hauled from 10. Therecomplete, no sub- sub- sub- below below 10 to to commence when when lower lower 75 75 percent percent complete, so so below 10 to commence when lower 75 percent dturbance. complete, so reason results results results that that originate originate from from last last 25 25 percent percent 10 10 will will show show less less traffic traffic dturbance. The The reason that that next next below below it it new new main main at at 1685 1685 m. m.
Minerals 2016, 6, 65 6 14 that originate from last 25 percent 10 will show less traffic dturbance. The reason that next below it new main at 1685 m. Haulage done by trucks from all areas to a crusher at main (1365 m). Trucks are loaded at loading chamber use a ramp to transport material to crusher. The simulation model also includes time loss when trucks meet to give way to each or, when y meet at intersection points between ramp crosscuts, when y meet at corner points. The model logic applies rule that when trucks meet, empty truck gives way to loaded truck at nearest waiting zone. The haulage dtances from each ramp to entry point at main (1365 m) are 2320, 2610, 4060 4350 m for s 9, 10, 14 15 respectively. The length main 2150 m, measured from entry ramp to crusher. Th makes total haulage dtances traveled by trucks 4470, 4760, 6210 6500 m for s 9, 10, 14 15 respectively. The crusher can be fed from two dumping points. During simulation, it was assumed that crusher would never break down. The simulation was run until machines in operation finhed removing all available ore from simulated areas. All ramps have an inclination 1:10. 4.6. Scenarios Seven scenarios evaluated analyzed, all aiming at meeting stated target 10,000 t per day per area (Table 1). Different combinations machines used for each scenario. Two sizes LHDs with capacities 10 21 t five different sizes trucks, with capacities 20, 21, 40, 42 63 t used. Table 1. Scenarios. Scenario LHD (Load-Haul-Dump) Bucket Capacity (Tonnes) Truck Capacity (Tonnes) Sub-Levels 1 21 21 9 10 2 21 42 9 10 3 21 63 9 10 4 21 21 14 15 5 21 42 14 15 6 10 20 9 10 7 10 40 9 10 For all scenarios, both electric diesel trucks used, number electric LHDs was limited to two at each area. When two LHDs employed, each one m worked on a different side area, meaning that re was no interaction between LHD machines during. The LHDs used to load trucks at each working area. The rationales behind selection different scenarios are as follows: Scenario 1 was selected to analyze haulage system when two LHDs serve smaller sized trucks (with capacities 21 t). Scenarios 2 3 selected to analyze effect traffic queuing on loading dumping points for each area when bigger units (trucks with capacities 42 63 t) are employed. Scenarios 4 5 selected to simulate effects increases in depth when having to transport ore from deeper s (14 15) to exting dump station at current main at 1365 m. Scenarios 6 7 selected to allow for smaller drift sizes, such that larger trucks cannot fit. Therefore, two electric LHDs, each with a bucket capacity 10 t, used to load trucks with box capacities 20 40 t. In all scenarios, aim was to determine number trucks needed to reach desired goals. In scenarios 1 to 5, operations with large drifts/openings (5 m wide) modeled
Minerals 2016, 6, 65 7 14 analyzed, whereas in scenarios 6 7, operations with small drifts/openings (less than 5 m wide) modeled analyzed. Note that in scenarios 1 4, one bucket assumed to fill truck. Th would in real operation create a rk overloading truck. Still, se scenarios are presented here since y are used to illustrate what happens when a smaller truck used. 4.7. Input Data The cases analyzed considering variations in availability trucks areas, also dturbance from or mining vehicles on ir way to from haul areas. The truck availability was assumed to be 90 percent. The area availability was set at 100 percent for areas with fewer interactions, 80 percent for areas with several or mining activities. The availability LHD machines was set at 100 percent for two cases, since LHD was immediately replaced after breakdown with anor during course, 90 percent when down time was considered. Keeping a back-up unit can be a costly strategy especially for a small mining operation. However, for a large mining operation like one presented in th paper, th a feasible strategy refore used in th study. According to data obtained from mine, dturbance from or mine vehicles to trucks was modeled as somewhere between zero percent two percent. Th means that up to two percent total available time lost due to traffic dturbance, hence delay for trucks increases when mine vehicles are driving up or down ramp along main. The term traffic dturbance refers to percentage time assumed to be lost when trucks meet in haul ways ( main drift ramp). No traffic dturbance occurs inside areas because within drifts, only LHDs are travelling to transport material from faces to a hauling point. Input data for simulation collected from mine, from equipment manufacturers. Five different cases studied for each scenario (Table 2). Table 2. Input parameters for each case. Cases LHD Availability (%) Truck Availability (%) Area Availability (%) Traffic Dturbance (%) 1 100 90 100 0 2 90 90 100 0 3 100 90 80 0 4 90 90 80 0 5 90 90 80 2 The assumed truck parameters (Table 3) used in simulation are based on data from mine from manufacturers. The data for 21 t truck obtained from a different mine since 21 t truck was not used in studied mine. There a great variability in loading time hauling units. To determine proper dtribution function, a stattical analys was performed a triangular dtribution was selected to model loading times. The size drifts does not accommodate truck size refore, during loading, trucks wait at access point that connects main drifts. The loading time includes time taken by LHDs to travel from drifts to truck loading point. The loaded truck moves up ramp, which has an incline 1:10, while empty truck moves down ramp decline to loading points.
Minerals 2016, 6, 65 8 14 Table 3. Truck parameters. Machine Truck Capacity (t) Loading Time (s) Dumping Time (s) Speed when Empty (km/h) Average Speed when Loaded (km/h) Volvo 21 t Truck 21 Tri (10,12.5,15) 10 30 18 incline 30 horizontal decline Atlas Copco MT42 42 Tri (145,162,179) 10 30 11.3 incline 30 horizontal decline Atlas Copco MT6020 63 Tri (450,505,567) 10 30 12 incline 30 horizontal decline Volvo 20 t Truck 20 Tri (143,167,180) 10 30 18 incline 30 horizontal decline Atlas Copco MT42 40 Tri (403,452,507) 10 30 Tri = triangular dtribution Tri (a,m,b) where m most likely value. 11.3 incline 30 horizontal decline 4.8. Verification Validation Simulation model verification process ensuring that model design has been transformed with sufficient accuracy into a computer model [21]. Validation process ensuring that model sufficiently accurate for purpose intended [21]. The stages for verification validation processes used in th study are based on one proposed by Robinson [22]. As seen in Figure 4, stages included in process are: Conceptual model validation. Th stage determines if scope detail proposed model are sufficient for purpose intended, that assumptions are correct. Data validation. Th stage determines that data used for building model, for validation for experimentation, are accurate. White box validation. Th stage determines that components computer model accurately represent corresponding real world elements. Black box validation. Th stage determines that overall model represents real world with sufficient accuracy. The model considered valid when assumptions underlying conceptual model are correct, when it has been determined that model represents real system [15,21,23]. The data validation was performed according to steps described below: 1. Production data from operating mine toger with machine specifications used to derive calculate output parameters interest. The data used actual t/h, shift schedule, equipment availability. The calculations resulted in specific values for number working hours required to load ore from a area, number required days required to load ore from a area number hours equipment unavailable. 2. The next step was to run a simulation using same input data boundary conditions as for calculations in step 1. The simulation model uses more detailed equipment data (speed, time to load, time to dump, as presented in Table 3) based on mine operating data data from manufacturer.
Minerals 2016, 6, 65 9 14 3. The results from simulation n compared with results from calculation in step 1, ensuring that simulation model gave similar output results. A mamatical analys was also done when calculating verifying oretical cycle time for trucks (drive to dump location, dump, drive back again), to verify upper limit trucks. Black box validation was n carried out by comparing output from simulation model with output from real system. Th was done by comparing actual hourly mine with simulation output for one mine block (shown in Figure 3). The average actual hourly for th block 4147 t/h with 21 t loader. The output from simulation model resulted in an hourly 41,474 t/h with 21 t loader, which validated model proved that it behaved like real system. White box validation was carried out by using debugging techniques, animations, model inspections by specialts, by running model under varying conditions. Minerals 2016, 6, 65 9 14 Figure 4. Verification validation in modeling process (modified from source [22]). 5. 5. Results Results Dcussion Dcussion The The simulation simulation was was run run based based on on mining mining in in 9th, 9th, 10th, 10th, 14th, 14th, 15th 15th s s in in mine mine (see (see Figure Figure 2). At 2). each At, each, trucks trucks used to used transport to transport material material from loading from loading areas to areas crusher, to crusher, which which located at located current at main current main at 1365 m. at Each 1365 m. Each consts consts two two areas. Two areas. electric Two LHDs, electric with LHDs, bucket with capacities bucket capacities 21 10 t, 21 used 10 t, to load used trucks to load at each trucks at each area. The 21 area. t LHDs The 21 t LHDs used to provide used in to provide case in large case openings, large while openings, 10 t LHDs while 10 t LHDs used for smaller used openings. for smaller The openings. aim The simulation aim was simulation to determine was to determine number trucks number needed trucks to meet needed to meet target 10,000 target t when 10,000 ore t pass when system ore pass not used. system not used. 5.1. Trucks Needed The simulation was first conducted for case 21 t LHDs loading 21 t trucks (scenario 1). After first simulation, model was adjusted, or scenarios shown in Table 1 simulated. The results for scenario 1 show that, when using smaller sized LHDs to load onto smaller sized trucks, between 13 16 trucks are needed, depending on case being analyzed (see Figure 5). 5). In In case case one, one, fewer fewer trucks trucks (13) (13) are are required required to meet to meet target target compared compared with with or cases. or Th cases. Th because, because, for thfor case, th we case, assumed we assumed that LHDs that LHDs immediately immediately replaced replaced with anor with anor after after breakdown during, we also assumed that re could be no dturbances from or mine operations in area or with or mine vehicles on ramp. The results for scenarios 2 3 show that, when larger trucks are used, fewer machines (between eight ten for scenario 2, between six seven for scenario 3) are required to achieve goal. For se scenarios, truck sizes used 42 63 t, respectively. The reason for requiring fewer trucks that, although bigger trucks have a slightly slower
Minerals 2016, 6, 65 10 14 breakdown during, we also assumed that re could be no dturbances from or mine Minerals operations 2016, 6, 65 in area or with or mine vehicles on ramp. 10 14 Minerals 2016, 6, 65 10 14 Figure 5. Truck requirements for large openings. The results for scenarios 26 37 show are shown that, when in Figure larger6. trucks In se arescenarios, used, fewer electric machines LHDs (between with a eight capacity ten 10 for t scenario used 2, to load between trucks six with seven capacities for scenario 20 3) 40 aret. required It shows tothat achieve a higher number goal. 20 t trucks For se scenarios, needed (between truck sizes 17 used24) compared 42 63with t, respectively. 40 t trucks The (between reason for 10 requiring 16). fewer trucks that, although bigger trucks have a slightly slower speed, over same time period y dump more material than smaller sized trucks. The results for scenarios 4 5 show that more trucks are needed when mining deeper: between 17 21 trucks for scenario 4, between 11 14 trucks for scenario 5. The reason for th that, under se scenarios, trucks have longer cycle times since y have a longer dtance to travel to dump station (6210 m for 14 6500 m for 15). Compared with scenarios 1 2, more trucks required to achieve daily under scenarios 4 5, despite fact that compared scenarios use Figure 5. same Truck truck requirements sizes. Cases for large 3 openings. 4 result in similar number trucks for all scenarios despite difference in truck loader sizes. Th because for se cases, area availability The results was reduced for scenarios to 80%, 6 which 7 are leads shown to more in Figure idle time 6. In for se loading scenarios, haul electric units. LHDs with a capacity The results 10 t for scenarios used to 6load 7 are trucks shown with in Figure capacities 6. In se 20 scenarios, 40 t. It shows electricthat LHDs a higher with anumber capacity 2010t ttrucks used toneeded load (between trucks with 17 capacities 24) compared 20 with 40 t. 40 Itt shows trucks that (between a higher 10 number 16). 20 t trucks needed (between 17 24) compared with 40 t trucks (between 10 16). Figure 6. Truck requirements for small openings. For example, in scenario 6, case 5, more than 20 trucks required to attain target due to presence higher traffic dturbance when a large number trucks are in operation. For scenario 7, although case 5 also allows for two percent time loss due to traffic, fewer trucks needed to meet target. The reason for th that, when goes ahead in th case, no trucks are available for in or areas. Th, refore, reduces congestion trucks on ramp. Comparing all scenarios, one requiring fewest trucks to achieve daily target was scenario 3. Th because trucks with highest capacity used for th scenario. For cases 3, 4 5, since area availability lower, a large number trucks needed to meet daily target. Furrmore, based on results from scenarios 4 5, it can be shown that, deeper mine Figure gets, 6. Truck more requirements trucks are for small needed. openings. 5.2. Truck For example, Traffic on in scenario Ramps 6, case 5, more than 20 trucks required to attain target due to presence higher traffic dturbance when a large number trucks are in operation. During simulation, different haulage options evaluated by determining truck For scenario 7, although case 5 also allows for two percent time loss due to traffic, fewer trucks traffic at each area depending on which trucks working. The term traffic needed to meet target. The reason for th that, when goes ahead in th case, refers to percentage time lost when trucks meet in haul ways ( main drift ramp). no trucks are available for in or areas. Th, refore, reduces congestion trucks
Minerals 2016, 6, 65 11 14 For example, in scenario 6, case 5, more than 20 trucks required to attain target due to presence higher traffic dturbance when a large number trucks are in operation. For scenario 7, although case 5 also allows for two percent time loss due to traffic, fewer trucks needed to meet target. The reason for th that, when goes ahead in th case, no trucks are available for in or areas. Th, refore, reduces congestion trucks on ramp. Comparing all scenarios, one requiring fewest trucks to achieve daily target was scenario 3. Th because trucks with highest capacity used for th scenario. For cases 3, 4 5, since area availability lower, a large number trucks needed to meet daily target. Furrmore, based on results from scenarios 4 5, it can be shown that, Minerals deeper 2016, 6, 65 mine gets, more trucks are needed. 11 14 5.2. When Truck trucks Traffic meet on at Ramps intersection points between ramp crosscuts, or at corner points, empty During truck simulation, gives way to different loaded haulage truck at options nearest wait evaluated zone. by As determining seen Figure 7 truck for s traffic9 at each 10 (scenarios 1, area 2, 3, depending 6 7), onaverage which time loss trucks due to queuing working. waiting The term for trucks traffic on refers to ramps percentage 8.17 percent. time Scenario lost when 6 generates trucks meet higher intruck haul traffic ways on ( ramps main drift due to depth ramp). When s trucks 9 meet 10, at also intersection due to points large between number ramp trucks crosscuts, in operation or atcompared corner points, with or empty scenarios. truck The gives third way scenario to loaded generates truck less at traffic nearest on wait ramps zone. since As seen it requires in Figure 7fewest for s trucks 9 to 10 meet (scenarios required 1, 2, 3, 6 7), target. average The time average loss due transportation to queuing time waiting rose to for over trucks 23 percent on ramps total 8.17working percent. time Scenario when 6 generates smaller higher sized truck LHDs traffic used on to ramps load duetrucks to (s depth 9 s 10). 9 10, Since alsomore due to trucks large are needed number at s trucks 14 in operation 15 due to compared increased with depth, or scenarios. truck traffic Theon third scenario ramps increases, generates lesshence trafficresults on in ramps increased sinceidle it requires time for each fewest truck trucks due to toqueuing meet problems. required Th shows target. that, when The average deeper s transportation are in operation, time rose a lot to over time 23 percent lost due to queuing total working waiting time when time for trucks smaller on sized ramps. LHDs used to load trucks (s 9 10). Figure 7. Truck traffic for all scenarios. Since An analys more trucks are results needed indicates sthat 14 15 most duefavorable to increased option depth, for large truck openings traffic on ramps haulage increases, system using hence trucks results with a capacity in increased 63 idle t. The time results for each also truck show due that to queuing favored problems. option for smaller openings to use trucks with a capacity 20 t. To achieve planned target 10,000 t per day from each mine block, option with large openings requires about seven trucks to be in operation, whereas more than 20 trucks would be required for smaller openings. Th shows that, for smaller drifts, hauling using trucks will not be an economical option in order to attain target 10,000 t per day. Diesel truck haulage systems are highly flexible in terms travel routes fleet size,
Minerals 2016, 6, 65 12 14 Th shows that, when deeper s are in operation, a lot time lost due to queuing waiting time for trucks on ramps. An analys results indicates that most favorable option for large openings haulage system using trucks with a capacity 63 t. The results also show that favored option for smaller openings to use trucks with a capacity 20 t. To achieve planned target 10,000 t per day from each mine block, option with large openings requires about seven trucks to be in operation, whereas more than 20 trucks would be required for smaller openings. Th shows that, for smaller drifts, hauling using trucks will not be an economical option in order to attain target 10,000 t per day. Diesel truck haulage systems are highly flexible in terms travel routes fleet size, present no electrical hazards. However, use a large number trucks underground increases rks from flammable fuel, results in higher heat emsions noe s, emits toxic gases into mine environment. Th results in higher costs because additional energy needed for longer haul dtances, need for additional ventilation to mitigate geormal heat exhaust gas emsions. Based on ratio emsions amount fuel used, diesel machines emit 2.68 kg CO 2 gas for every liter diesel fuel used [24]. The need for ventilation to mitigate engine heat emsions will be higher, th will lead to increased operating costs. One possible alternatives when using smaller trucks to develop two ramps: one ramp will be used by loaded trucks (incline) or by empty trucks (decline). For large openings, analys shows that seven trucks can be employed to attain planned target. More than seven trucks will increase traffic congestion, result in longer idle times, higher operating costs, lower rate. Due to increased depth, hence increased rock stresses, problems with ore passes failure are expected, which will lead to furr delays in material transportation. In sub- caving, when ore passes trains are used to transport ore to a crusher, ore pass management necessary in order to avoid interruptions to material flow. When costs managing ore passes become too high, or when ore passes cannot be used for, haul trucks can be employed as an alternative method to transport ore waste up ramps to exting crusher located at previous main. If truck haulage adopted, a crusher a rail-mounted track system at lower main will no longer be required, re will be no need to extend a shaft from previous s. Th will reduce reinvestment costs for new main infrastructures. 5.3. General Dcussion Results Method The choice using simulation to solve problems sometimes dputable in many cases analys using simulation too complex time consuming given nature study. For th specific study, using simulation was preferred option, results presented above could not have been generated with same accuracy through manual calculations alone. Analyses given problem include more input information than could be hled by manual calculation still generate results with same degree detail accuracy as possible with a simulator. In simulations concerning 9 10, output results also showed important information on how traffic dturbances affected overall performance variance in total time each truck was working in one area until it started to drive on ramp ( trucks driving time loading time) which affects frequency at which trucks entered ramp. The number trucks also affected queuing, both inside each area at dump location, availability loaders anor important factor that affects how trucks work. All se factors, which are hard to describe calculate since y are not determintic values or constants, require a dynamic model that behaves like reality to obtain a reliable result. For purpose investigating a possible future scenario, most appropriate tool possible should be used, which in th case a simulator.
Minerals 2016, 6, 65 13 14 6. Conclusions Th study was conducted in an exting deep underground mine that uses sub- caving. Dcrete event simulation was used to evaluate options for future haulage from lower mine s without use ore passes by comparing different haulage equipment with various sizes operating from drifts to crushers. The results indicated following: Haul trucks can be employed as an alternative haulage method when costs managing ore passes become too high, or ore passes cannot be used for. For large openings, favored option for th case a haulage system using trucks with a capacity 63 t. For smaller drifts, hauling using trucks will not be economical in order to attain desired target. The study showed that dcrete event simulation toger with verification validation processes to develop a model for analys a suitable tool for investigating analyzing mine operations prior to new investments or implementation new systems. There are, however, many additional considerations that analys by simulation may not be able to resolve. To evaluate additional costs considerations when using trucks, dcrete event simulation can be combined with economic analys models to improve understing reduce rk related to selection operational systems. Future research in th area will focus on combining dcrete event simulation mixed integer programming to study following: ore pass management costs, additional truck operating costs, new haulage systems for deeper s, capital costs required for installation necessary infrastructures in new main. Future work on comparon capital operational cost electric versus diesel trucks also suggested. Acknowledgments: Th study part I2 Mine Project (Innovative Technologies Concepts for Intelligent Deep Mine Future), WP 2 subtask 2.1.1, has been carried out at Divion Mining Geotechnical Engineering at Luleå University Technology. The project funded by EU 7th framework programme. Author Contributions: Jenny Greberg Abubakary Salama conceived designed experiments/simulation. Abubakary Salama, Jenny Greberg Bartlomiej Skawina performed experiments by simulation. Jenny Greberg, Abubakary Salama, Bartlomiej Skawina Anna Gustafson analyzed data. Jenny Greberg Abubakary Salama main authors paper, assted by Anna Gustafson Bartlomiej Skawina. Conflicts Interest: The authors declare no conflict interest. References 1. Stacey, T.R.; Wesseloo, J.; Bell, G. Predicting stability rock passes from geological structure. J. S. Afr. Inst. Min. Metall. 2005, 105, 803 808. 2. Stacey, T.R.; Swart, A.H. Investigation into Draw Points, Tips Orepasses Chutes; Technical Report Report to Safety in Mines Research Advory Committee, Project OTH 303; Mine Health Safety Council: Johannesburg, South Africa, 1997; Volume 1. 3. Kuchta, M.; Newman, A.; Martinez, M. Long- short-term scheduling at LKAB s Kiruna mine. In Hbook Operations Research in Natural Resources; Weintraub, A., Romero, C., Bjørndal, T., Epstein, R., Eds.; Springer: New York, NY, USA, 2007; pp. 579 593. 4. Iverson, S.R.; Jung, S.J.; Bwas, K. Comparon orepass computer simulations for design against dynamic load. In Proceedings SME Annual meeting, Cincinnati, OH, USA, 26 28 February 2003. 5. Brummer, R. Design Ore Passes Methods for Determining Useful Life Ore-Passes Based on Previous Experience Case Studies; Technical Report CAMIRO; Mining Divion Limited: Sudbury, ON, Canada, 1998. 6. Hadjigeorgiou, J.; Esmaieli, K.; Harrson, R. Observation ore pass system performance at Brunswick Mine. CIM Bull. 2008, 101, 1 13. 7. Hadjigeorgiou, J.; Stacey, T.R. The absence strategy in ore pass planning, design management. J. S. Afr. Inst. Min. Metall. 2013, 133, 795 801.
Minerals 2016, 6, 65 14 14 8. Joughin, W.C.; Stacey, T.R. Rks associated with rock passes in deep- tabular mines based on htorical pass performance. J. S. Afr. Inst. Min. Metall. 2005, 105, 795 802. 9. Hadjigeorgiou, J.; Lessard, J.F. Strategies for restoring material flow in ore waste pass systems. Int. J. Min. Reclam. Environ. 2010, 24, 267 282. [CrossRef] 10. Kecojevic, V.; Komljenovic, D. Haul truck fuel consumption CO 2 emsion under various engine load conditions. Miner. Eng. 2010, 62, 44 48. 11. Atkinson, T. Selection sizing excavating equipment. In SME Mining Engineering Hbook, 2nd ed.; Hartman, H.L., Ed.; SME: Littleton, CO, USA, 1992; pp. 1311 1333. 12. McCarthy, P.L. Selection shaft hoting or decline trucking for underground mines. In Driving Down Haulage Costs; AMC Reference Library: Kalgoorlie, Australia, 1999. 13. Pratt, A.G.L.; Ellen, P.J. Selection an ore haulage system for Telfer deeps. In Proceedings Hot Haul Conference, Melbourne, Australia, 5 7 September 2005; Australasian Institute Mining Metallurgy: Perth, Australia, 2005; pp. 131 140. 14. Salama, A.; Greberg, J.; Schunnesson, H. The use dcrete event simulation for underground haulage mining equipment selection. Int. J. Min. Miner. Eng. 2014, 5, 256 271. [CrossRef] 15. Banks, J.; Carson, J.S.; Nelson, B.L.; Nicol, D.M. Dcrete Event System Simulation; Pearson Education: Upper Saddle River, NJ, USA, 2010. 16. Raj, M.G.; Vardhan, H.; Rao, Y.V. Production optimization using simulation models in mines: A critical review. Int. J. Oper. Res. 2009, 6, 330 359. [CrossRef] 17. Sturgul, J.R. Dcrete mine system simulation in United States. Int. J. Min. Reclam. Environ. 1999, 13, 37 41. [CrossRef] 18. Salama, A. Haulage Systems Optimization for Underground Mines: A Dcrete Event Simulation Mixed Integer Programming Approach. Ph.D. Thes, Luleå University Technology, Luleå, Sweden, 2014. pp. 1402 1544. 19. Greberg, J.; Sundqvt, F. Simulation as a tool for mine planning. In Proceedings 2nd International Future Mining, Sydney, Australia, 22 23 November 2011; pp. 273 278. 20. SimMine, Version 1.19. Available online: http://simmine.com/ (accessed on 29 June 2016). 21. Sargent, R.G. Verification validation simulation models. In Proceedings 37th Winter Simulation, Syracuse, NY, USA, 4 7 December 2005; pp. 130 143. 22. Robinson, S. Simulation model verification validation: Increasing user s confidence. In Proceedings 29th Winter Simulation Conference, Atlanta, GA, USA, 7 10 December 1997; IEEE Computer Society: Washington, DC, USA, 1997. 23. Balci, O. Validation, verification, testing techniques throughout life cycle a simulation study. Ann. Oper. Res. 1994, 53, 121 173. [CrossRef] 24. Salama, A.; Greberg, J.; Skawina, B.; Gustafson, A. Analys energy consumption gas emsions for loading equipment in underground mining. CIM J. 2015, 6, 179 188. [CrossRef] 2016 by authors; licensee MDPI, Basel, Switzerl. Th article an open access article dtributed under terms conditions Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).