Optimization study of a polymetalic sulfurous narrowvein mining operation

Diogo Jose das Fontes Estevensdiogofestevens@tecnico.ulisboa.pt

Instituto Superior Tecnico, Lisboa, Portugal

November 2019

Abstract

Vein mining operations are an important source of high economic value metallic ore containing ele-ments like gold and silver. The extraction of this type of orebody have a significant complexity due to thesometimes unpredictable geometry and complex geologic setting.

Saucito mine from Fresnillo plc faces a progressive decrease in vein thickness with the profundizationof its operations. The continuation of the use of the present extraction drift dimensions in the new areaswith smaller vein thickness results in a considerable increase in dilution turning some areas of the minesub-economic and reducing the profitability of the operation.

To evaluate from the mine planning perspective the impact of the implementation of operative optimiza-tion practices in the Jarillas vein, four extraction models were generated using the Datamine software.The models were generated to allow a comparative analysis between mining methods and extraction driftdimensions, varying between models the extraction parameters that influence dilution.

The output numerical data from the generated models support the thesis that to the selected studyarea the optimal extraction strategy is the optimization of the Overhand cut & fill mining method, altoughthe Longhole Stoping showed a bigger improvement. The methodology used and the results from thisstudy provide a framework to the refinement of the comparative analysis and for the operative optimizationstudies in the Saucito mine.Keywords: Vein Mining; Dilution; Mine planning; Operative optimization; Extraction models.

1. Introduction

Mining operations, booth underground and sur-face, are one of the economic activities most cap-ital intensive, not only in a initial stage, whenthe economic feasibility of the project is studiedthrough a series of surface drilling campaigns, butalso during the operational stage, when there is al-ready the production of some kind of mineral ore.One of the main problems that mining companiesface resides in the quantification and qualificationof minerals volumes [2]. This process will impactdecisively the decision of whether or not extract agiven orebody.

The importance of mining companies, and thefact that most of them have some significant por-tion of their capital on open negociation on thestock market, leads to the necessity of correctlyquantify the mineral volumes that can be econom-ically extracted, specially considering the impactthat an increase of reported reserves can have onthe stock price of a mining company. This pro-cess known as resource and reserves estimationis executed based on a series of bidimensional or

tridimensional samplings, applied to a certain vol-ume of mineralized rockmass, whose dimensionsare considerably superior to the volume sampled[3].

To the process of resource and reserves estima-tions, a series of systems were developed, usuallyapplied by regions of the world. Despite the difer-ent systems, the procedures, purpose and charac-teristics are relatively similar. The three most usedsystems are the Canadian Norm National Instru-ment 43-101, developed by the Canadian Instituteof Mining, Metallurgy and Petroleum, the PERCCode - Pan-European Reserves and ResourcesReporting CCommitee, and the Australian norm,know as the JORC CODE , acronym of Joint OreReserves Code, also know as Australasian Codefor Reporting of Exploration Results, Mineral Re-sources and Ore Reserves, which is the norm usedfor estimation of resources and reserves in the min-ing companies operating in Mexico. Somethingthat is comum in all the codes is the legal figureof the ”competent person”. This legal figure is re-sponsable for the quantification of resources and

1

reserves, a process that must take into accountother factors beside the sampling but also consid-erations about the geomechanical conditions of thesurounding ore mass, the characteristics of the orethat might afect the metalurgical process, as alsoeconomic, social, environmental and political fac-tors that might affect the economic potential of theoperation.

These factors, known as ”modifying factors”, playa essential role in the process and a consider-able higher level of imprecision on the quantifica-tion of these variables might have significant impli-cations on the mining process and affect decisivelythe profitability of the extraction. The relation be-tween the exploration results, resources, reservesand the modifying facts are shown in the figure 1.

Figure 1: Relation between the exploration results, resourcesand reserves. Taken from the JORC Code 2012 Edition [1].

One class of modifying factors are the oneswhose magnitude is dependent on the extractionprocess. These factors include all the parametersable to influence the dilution, the recovery and theore grade. A change on the modifying factors mightlead to a increase or decrease on the mineral vol-umes reported as reserves [5]. Due to this influ-ence, mining companies try to, not only predict thevalues of these parameters but also optimize theiroperations to reduce the influence of these param-eters and increase the profitability of the operation.

The actual mine planing softwares are a conse-quence of the development of the informatic toolsin the last 40 years. The mine planing softwaresallow, with the knowledge of some input parame-ters, the generation of geological models to ascer-tain if a given area of the orebody can be minedprofitably. To correctly use these tools two areasare important, the Geostatistics, know for the useof mathematical methods in the earth sciences toquantify and qualify the ore reserves and mineralresources, and the Mine Planning that producesstudies to evaluate the more optimized methodsand plans to extract the reserves quantified.

1.1. Case StudyThe Saucito mine, owned and operated by Fres-nillo plc, is a mining industrial unit located in theFresnillo district, known for the existence and ex-traction of sulfurous polimetalic veins with a signif-icant grade of silver, in his area territory. During aseries of surface drilling campaigns near the Fres-nillo mine, a series of veins were discovered south-west of the Fresnillo city. In 2011, a new industrialunit was opened to extract these veins, some ofthem with a significant vertical and horizontal ex-tension, which is the case of the Jarillas vein. In2018, 2.79 milion tonnes with an average grade of257 grams per tonne of silver were extracted fromthis mine, data that make Saucito mine the biggestprimary silver source in the world.

Two mining methods are used in the Saucitomine. The most used, altough a decrease in its usewas noted in recent years, is the Overhand Cut &Fill. This method uses a typical drift and then theextraction is executed upwards, using the extractedwaste, after the mucking of the ore, as a platformto the next line of extraction. This method is visiblein the figure 2.

Figure 2: Use of the Overhand Cut & Fill in the Santa Nataliasgroup of veins in the Saucito mine

The other mining method used is the LongholeStopping. This method uses two drifts, one supe-rior and one inferior, on which the extraction is exe-cuted. The superior drift is used to the drilling andfilling activities and the inferior drift is used for themucking of the ore. An example of the use of theLonghole Stoping can be seen in the figure 3.

The veins in the Saucito mine have diferent thick-ness and characteristics, in some areas of the Jar-illas vein, the thickness can reach 20 to 25 meterswhile in other areas, like the Santa Natalias groupof veins, the average thickness is around 1.8 me-ters. In the next years the extraction is going toadvance to the more profound areas of the Jarillasvein, where, according to the most up to date geo-logical model, there are considerable areas wherethe average vein thickness is smaller then 2.5 me-ters. The use of the current extraction dimensions

2

Figure 3: Use of the Longhole Stoping in the San Diego vein inthe Fresnillo mine

with the present equipments used, on areas wherethe average vein thickness is smaller can lead to asignificant increase in dilution and turn some areasof the vein non economical.

1.1.1 Purpose of the study

This study was included in a series of studies indevelopment in the Fresnillo plc group that try tosearch and evaluate alternatives to optimize theextraction in the vein mining operations in the com-pany. The study tried to evaluate the impact thata reduction on the extraction drift dimensions, aswell the use of optimized equipments, can have onthe extracion parameters and on the economic pa-rameters of the extraction. To do that, the studyused the Saucito dilution database to evaluate howthe implementation of the operative optimizationprocedures on the Santa Natalias group of veins,known for the smaller vein thickness , affectedthe extraction parameters linked to dilution. Withthose values, four extraction models were gener-ated, each one with diferent characteristics, to al-low a comparative analysis between mining meth-ods and between optimized and non-optimized ex-traction drifts. To complete the study, a series ofoperational and economic parameters were ana-lyzed. To allow the generation of the extractionmodels a computarized mine planning system wasessential. The study used three diferent versions ofthe mine planing software Datamine, the Datamine2 for the generation of the geological reserve mod-els and the selection of the extraction areas, theDatamine studio 3 for the mine design of the in-frastructures and the Datamine 5DPlanner for thesequencing of the extraction.

2. BackgroundThe extraction and processing of minerals is anessential part of the way diferent civilizations andeconomic blocks trade and interacted with eachother. It is considered that modern mining hadits begining in 1556, date of the publication of themanuscript De Re Metalica by Georgius Agricola,a monography that detailed with precision the prac-tices of extraction and processing of ore in themines of Erzgebirge, in the Saxony during the XVIcentury. Modern mining nowadays is the result ofa complex chain, on which the modern equipmentsand informatic tools play a significant role.

2.1. Mine PlaningMine Planing can be defined as the process of ex-ctraction of the correct ore from the mine, in theright time, in order to achieve the lowest possiblecost per final unit of commodity in order to fulfill thestrategic and operational targets of the operatingcompany

. Mine planing is an essential part in every stageof the mining process, from the generation of thegeological models through the geostatistic tools tothe conception of a realistic extraction plan. Themine planing, when executed strategicly, is a pro-cess on which the extraction plan is integrated andalligned with the strategic targets of the operatingcompany and it needs a constant and cyclical feed-back to adjust the operations and the ore reservesto accomodate changes in the modifying factors.

.Mine plans are elaborate with a given financial

objective. The Net Present Value (NPV) is a fi-nancial and economic formula able to calculatethe present value of future payments, discountedwith a given interest rate, minus the present cost,known. The most comum formula is the one on theequation 1.

NPV (i,N) =

n∑t=0

Rt

(1 + i)t(1)

Rt represents the cash-flows of a given time, Nrepresents the number of intervals of time and i theinterest rate for projects with a similar risk profile.Historically, mining companis have adopted difer-ent NPV strategies altough the most comon is thesearch for the maximization of the NPV.

The evolution of mine planning as a sciencehave been hand to hand with the evolution of thecomplexity of extraction operations. The increaseof the informatic processing tools allowed the de-velopment of a significant and diverse range ofmine planing softwares. This type of informatic pro-gram play a significant role in the planning of theextraction operations.

3

Due to the need for a constant and cyclical feed-back between the mine plan and changes in themodifying factors, a computarized mine planingsystem is a necessity for the mining companiesthat want to increase the resiliency of their oper-ations.. The mine planing software Datamine is one ofthe most used programs on the industry and allowsthe monitoring of the extraction process from thegeneration of the geological and reserves modelsto the mine design, sequencing of the extracitonand the final step of reconciliation of mining data.

The relatively user-friendly use and the speed onwhich a model can be generated, for the same areaof the orebody, allows the modeling and simulationof diferent extraction conditions as a tool to evalu-ate which alternative, or extraction model, providesthe best economic output for the extraction.

2.2. Narrow vein miningThe extraction of narrow veins, despite the his-toric relevance, is an activity that has been loos-ing importance in the mining sector globally. Thechanges in the mining methods, as a result fromthe introduction of new technological solutions,lead to a series of new ways to mine orebodiesbeyondairlegmining. It is generally considered theexistence of three significant mining methods forthe extraction of sulfurous polimetalic veins. TheJackleg drilling method, seen in the figure 4 is stillused in some operations were the the grade, sig-nificantly high is paired with a vein thickness signif-icantly reduces.

Figure 4: Use of Jackleg Drilling method beyondairlegmining.

Altough the use of this mining method, opera-tions were the vein thickness is bigger and the min-eralogical contents are less valuable do not allowthe use of this method, specially since the neces-sary tonnage for the profitability of the operation isconsiderably higher. the suitable mining methodsneed to provide higher tonnages to achieve prof-itability while allowing the extraction with safety andgeomechanical standards. The two other miningmethods used in vein mining are the Cut & Fill andthe Longhole Stoping.

The Cut & Fill is a mining method that requiresthe use of rock filling in the voids resulting from the

extraction. The variation of the Cut & Fill, known asOverhand Cut & Fill and represented in the figure 5is the most used variation of the Cut & Fill in narrowvein mining.

Figure 5: Overhand Cut & Fill mining method [4].

The other mining method being used currently ina large scale in vein mining is the Longhole Stop-ing. This method requires two extraction drifts andallows for the extraction of more tonnage altoughsignificantly impacting the dilution. A schematicfrom the Longhole Stoping can be seen in the fig-ure 6.

Figure 6: Schematic of the Longhole Stoping mining method[4].

2.3. Dilution in mine planingThe dilution (D) is the relation between the massof mined ore (mm) and the mass of waste rock(me), usually mined together. The calculation ofthe dilution varies base on the area of the MiningEngineering that is doing the calculation. For themetalurgical calculation of the dilution, the most co-mom formula is the one formulated in the equation2, where

D (%) =me × 100

mm +me(2)

This formula defines dilution as the percentageof material, of the total, that is waste rock. How-ever, in mine planing the dilution is usually definedacording to the equation 3, as an increment ofmass.

D (%) =me × 100

mm(3)

The second equation is the one used globally inthe calculation of the dilution. Altough the single

4

value usually given to the dilution, this parameteris usually the agregate of various diferent types ofdilution, a planned one, resulting from the usuallyineficiente mining processes, an internal dilutionusually comon in vein mining, and an unplannedone, resulting from non planned ineficiences in theprocess.

3. MethodologyThe methodology used in this project was basedon the processes used in the Saucito mine for thegeneration of models of mineral resources and orereserves. To the base process, a series of adjust-ments were implemented, to generate a series ofextraction models to allow a comparative analysisbetween them. The workflow used is summarizedin the figure 7

Figure 7: Workflow of the methodology used in the study.

To take advantage of the extraction data sum-marized in the Saucito dilution database, and con-sidering the implementation of operative optimiza-tion measures in the Santa Natalias vein, the studygenerated the extraction models using the datafrom this dilution databasw and aplying the data inthe area from the Saucito mine that is going to facein the future, a reduction on the vein thickness.

To fully understand the problem and correctly ap-ply the values, the study started with a research onalternatives of optimization and bibliographic ref-erences. The second stage of the process wasa numerical analysis on the extraction data fromthe Saucito dilution database. This analysis pro-vided the extraction parameters for the generationof the reserves models. The third step was theImplementation stage were the values obtained inthe numerical analysis were applied in a specificarea of study. This step included the generation ofthe reserves models, the selection of a adequatestudy area, the mine and stoping design and thesequencing of the extraction of the four extractionmodels.

Finally, a series of analysis were done to eval-uate the economic and operational output fromthe generated models, provided by the sequencingprocess.

4. DevelopmentThe development of the study was conducted togenerate four extraction models, acording to the ta-ble 1. The non optimized models were the model1, using the Overhand cut & Fill and non optimizedextraction drift dimensions and the model 2, usingthe Longhole stoping and non optimized extractiondrifts. The optimized models were the model 3,using the Overhand Cut & Fill and the optimizedextraction drift dimensions, and the model 4, usingthe Longhole Stoping the the optimized extractiondrift dimensions.

Table 1: Proposed extraction modelsModel Mining method Optimized

1 Overhand cut & fill No2 Longhole stoping No3 Overhand cut & fill Yes4 Longhole stoping Yes

The optimization comparative analysis must bemade, for the Overhand cut & fill mining method,between the model 1 and 3 and, for the Longholestoping, between the models 2 and 4.

4.1. Numerical Analysis and geometric parametersof extraction

This stage analysed the Saucito dilution databaseand, base on the the values on optimized extractionareas and non optimized, selected the geometricparameters of extraction most suited for the mod-els. The study considered three geometric param-eter, the extraction drift width ξ, represented in thefigure 8, the minimum width of extraction ς, and theplanned dilution, δ.

ε

Filão

Figure 8: Geometric representation of the extraction drift (ξ)

The extraction drift width parameter is the width,in meters of the extraction drifts where the drillingand the loading processes are executed. The min-imum width of extraction, when its value is smallerthan the projected vein thickness, represented inthe figure 9 is linked to the selectivity of the equip-ments used and the eficiency of the drill and blast-ing process.

When ς is smaller than the projected vein thick-ness, the process ignores its value and assumesthat the width of extraction is the projected veinthickness plus the planned dilution δ. On the other

5

𝛿

2

𝛿

2

ζ

Filão

Espessura aparente do filão

𝛿

2

𝛿

2

Figure 9: Geometric representation of the minimum width ofextraction when is smaller than the projected vein thickness

hand, when the minimum width of extraction is big-ger than the vein thickness, a situation representedin the figure 10, the process includes and aditionaldilution in the planned dilution.

ζ

Espessura aparente do filão

Filão

𝛿

2 𝛿

2

Figure 10: Geometric representation of the minimum width ofextraction when is bigger than the projected vein thickness

The final values used in the study were; extrac-tion drift width, ξ, of 4.5 for the non optimized mod-els, 1 and 2, and 3.4 for the optimized models, 3and 4. The minimum width if extraction, ς, wasdimensioned at 2, for the non optimized models,1 and 2, and 1.5 for the optimized models, 3 and4, assuming a increasse in selectivity capability onequipments. The Planned dilution, δ, was set at0.54 for the model 1, 0.64 for the model 2, 0.43 forthe model 3 and 0.51 for the model 4.

4.2. Experimental StageAfter the numerical analysis, the study imple-mented the numerical values on the models withthe Datamine software.

4.2.1 Study area selection

To correctly execute the study, a suitable studyarea was chosen. The selection followed six rules.The study area must be in a suitable vein, theselected area must be a cluster of economic ex-tractable bloc, it must be a isolated cluster of eco-nomic extractable blocks to correctly evaluate thediferences in extractable areas, it must have a con-siderable horizontal and vertical extension to de-sign a realistic extraction model, it must be near

areas currently being extracted and, at last but themost important rule, it must have an average veinthickness smaller than 2.5 meters. The selectedarea is represented in the figure 11.

Figure 11: Area de estudo selecionada para a implementacaodos modelos de extraccao em Jarillas Central em ambiente deDatamine 2

4.2.2 Geological reserve models generation

Four diferent geological reserve models were gen-erated base on the values and considering twodiferent cut-off values, on dolars. One for the Over-hand Cut & Fill and another, lower, for the Long-hole Stoping. Due to confidentiality issues this stepcannot be covered as desired. Despite this issue,some considerations must be done. The diferentcut-off value, in dolars, are essentially the cost ofextraction per tonne. The geological reserve mod-els generated for each model have two voids ineach one of them. These voids, visible in the geo-logical reserve model generated for the model 1, inthe figure 12, are due to two diferent reasons. Thevoid at the left is linked to the fact that the extractionof this area from the Jarillas vein does not belongto the Saucito mine, due to a diferent area of explo-ration. The second void, at the center of the Jarillasvein, is an area near the currently extraction zone,and its geological model is generated with a difer-ent setting of data, that also takes into considera-tion a hand sample database to generate a moreprecise geological model.

Figure 12: Geological reserve model generated for the model1

4.2.3 Extraction polygons design

To select the extraction areas, based on the ge-ological reserves models generated, a series ofpolygons were designed on Datamine 2. Thesedesign followed the code, summarized in the ta-ble 2 and was executed considering the constraints

6

of each mining method assigned to each modelas well the geomechanical considerations, repre-sented by the inclusion of a crown pilar in the mid-dle of the study area selected.

Table 2: Color code for the design of the extraction polygonsColor codeDatamine Cor Code Description

2 DS Extraction drift/Development5 BL 1st level of Longhole stoping (DBS)6 BD 2nd level of Longhole stoping (DBS)7 BC 3rd level of Longhole stoping (DBS)8 CR Overhand cut & fill (CEA)

17 PL Crown pilar

After the design of the bidimensional polygons,through a process of scripting, the bidimensionalmodels turned into tridimensional models, with thevolumes calculated according to the values se-lected in the numerical analysis and in the geo-logical reserves models for each extraction model.The images of the tridimensional models can beseen in the figures 13 and 14 for the models usingOverhand sut & Fill and for the models using theLonghole stoping, respectively.

Figure 13: Bidimensional extraction polygons - models 1 (A)and 3 (B)

Figure 14: Bidimensional poligons of extraction on the models2 (A) and 4 (B)

4.2.4 Mine Design

With the extraction polygons designed, the studyproduced for each extraction model the mine de-sign acording to the code summarized in the table3.

Table 3: Color code for the mine designColor codeDatamine Cor Code

Infrastructure Description

3 ACCS Acesses4 CNFT Paralel extraction drifts9 FRTE Exctraction drifts

10 RAMP Ramp11 ROBB Ventilation shaft12 CRRO Water management drift39 PLZA Ventilation shaft acess41 CARG Ore loading area

Due to the similarity between models, only twofull mine designs were executed in Datamine Stu-dio 3, one for the Overhand cut & fill mining methodand other for the Longhole stoping. For the mod-els 3 and 4, small adjustments were made to eachmine design to accomodate the increase in areasand the existence of now economically extractableblocks in the extraction drifts areas, that were, inthe non optimized models, 1 and 3, non economictherefore generated through the mine design andnot through the extraction polygons.4.2.5 Extraction sequencing

The sequencing process is based on the adjust-ment of the time of extraction of each extractionblock, either from the mine infra-structures, gener-ated through the mine design, or the blocks witheconomic ore, generated through the extractionpolygons. This process faces, usually a series oferror like the duplication of areas, the severe an-gle changes or weak connections. This problemswere solved in the Datamine 5DPlanner, withoutthe need to return to the previous steps of the pro-cess. One of the ouputs from this process is avideo animation of the sequenced extraction pro-cess for each model. A frame from the videos ofthe Overhand cut & fill models is represented inthe figure 15.

Figure 15: Video frames from the sequencing of the models 1(A) and 3 (B) in Datamine 5D Planner

In the figure 16 is a frame from the videos gen-erated based on the models using the Longholestoping

Figure 16: Video frames from the sequencing of the models 2(A) and 4 (B) in Datamine 5D Planner

Agregated to the animation two other files weregenerated, a Datamine 5DPlanner file that can beused to further adjust the extraction and a se-quencing table with the extraction parameters permonth. This step concluded the implementationstep of this study.

7

5. Results & DiscussionTo analyze the results from the generated models,and the extraction parameters of each model, thesequencing tables generated to each model wereused and the results were divided in operationaland economic.

5.1. Operational analysis5.1.1 Ore mass and average grade

One of the main parameters analyzed were the oremass extracted (mm) and the average grade (tm),being the tp the weighted average grade of the totalmodel, calculated acording to the equation 5.1.1.

tp =

∑Nk=1mm × tm

Tm(4)

The data generated in the sequencing process wassummarized in the figures 17 for the Overhand cut& fill models, 1 and 3.Massa de minério extraída Modelo 1

Massa de minério extraída Modelo 3

0

1000

2000

3000

4000

5000

6000

7000

jan/

19

fev/

19

mar

/19

abr/

19

mai

/19

jun/

19

jul/1

9

ago/

19

set/

19

out/

19

nov/

19

dez/

19

jan/

20

fev/

20

mar

/20

abr/

20

mai

/20

jun/

20

jul/2

0

ago/

20

set/

20

out/

20

nov/

20

dez/

20

jan/

21

fev/

21

mar

/21

abr/

21

mai

/21

jun/

21

jul/2

1

ago/

21

set/

21

out/

21

nov/

21

0

20

40

60

80

100

120

140

Mas

sa d

e m

inér

io e

xtra

ída

[ton

]

Tempo

Teor

méd

io A

g [g

r/to

n]

Massa de minério extraída e Teor médio - CEA

Massa de minério extraída Modelo 1 Massa de minério extraída Modelo 3 Teor médio Modelo 1 Teor médio Modelo 3

Modelo 4

Figure 17: Mass of extracted ore and average grade on themodels 1 and 3

For the models using the Longhole stoping, thedata for the mass of extracted ore and averagegrade is summarized in the figure 18

0

2000

4000

6000

8000

10000

12000

14000

jan/

19

fev/

19

mar

/19

abr/

19

mai

/19

jun/

19

jul/

19

ago/

19

set/

19

out/

19

nov/

19

dez/

19

jan/

20

fev/

20

mar

/20

abr/

20

mai

/20

jun/

20

jul/

20

ago/

20

set/

20

out/

20

nov/

20

dez/

20

jan/

21

fev/

21

mar

/21

abr/

21

mai

/21

jun/

21

jul/

21

ago/

21

set/

21

out/

21

nov/

21

0

20

40

60

80

100

120

140

Mas

sa d

e m

inér

io e

xtra

ída

[ton

]

Tempo

Teor

méd

io A

g [g

r/to

n]

Massa de minério extraída e Teor médio - DBS

Massa de minério extraída Modelo 2 Massa de minério extraída Modelo 4 Teor médio Modelo 2 Teor médio Modelo 4

Figure 18: Mass of extracted ore and average grade on themodels 2 and 4

In the figures presented is visible a increase inextracted ore mass and in the averagde grade.This visual perpection is verified when the raw datais analyzed. The table 4 contains the data for theextracted ore mass in the four generated models.

Between the models 1 and 3, a increase of 3.95%in extracted ore mass, from 80 837 to 84 029 provesthe theorethical increase in extractable area withthe optimization. In the models 2 and 4, the in-crease was even more significant, from 70 127 to

Table 4: Variation in the mass of extracted ore in the four ex-traction models generated

Parameter/Model Model 1 Model 3 Diference [%]Total mass of extracted ore [ton] 80 837 84 029 3.95

Parameter/Model Model 2 Model 4 Diference [%]Total mass of extracted ore [ton] 70 127 77 345 10.29

77 345, a percentual variation of 10.29%. The datafrom the average grade is concised in the table 5.

Table 5: Variacao no teor medio de prata nos quatro modelosde extraccao gerados

Parametro/Modelo Modelo 1 Modelo 3 Diferenca [%]Teor medio Ag [gr/ton] 111.77 112.82 0.94

Parametro/Modelo Modelo 2 Modelo 4 Diferenca [%]Teor medio Ag [gr/ton] 108.20 109.80 1.48

The diference in this parameter is less signifi-cant, with the extraction models presenting an in-crease of 0.94% between the models 1 and 3 andan increase of 1.48% in the models 2 and 4.

5.1.2 Silver produced

The silver produced, or equivalent mass of silver(Ageq), is a parameter that calculates the amountof silver that can be produced considering notonly the mining recovery but also the metalurgicalprocess. This parameter is usually measured inounces. To analyze this parameter, the Ageq wasagregated to the extracted ore mass, in the figure19 for the models 1 and 3, and the figure 20 for themodels 2 and 4.

Massa AG equivalente Modelo 3Massa AG equivalente Modelo 4

0

1000

2000

3000

4000

5000

6000

7000

jan/

19

fev/

19

mar

/19

abr/

19

mai

/19

jun/

19

jul/

19

ago/

19

set/

19

out/

19

nov/

19

dez/

19

jan/

20

fev/

20

mar

/20

abr/

20

mai

/20

jun/

20

jul/

20

ago/

20

set/

20

out/

20

nov/

20

dez/

20

jan/

21

fev/

21

mar

/21

abr/

21

mai

/21

jun/

21

jul/

21

ago/

21

set/

21

out/

21

nov/

21

0

5000

10000

15000

20000

25000

30000

35000

40000

45000

50000

Mas

sa m

inér

io e

xtra

ída

[ton

]

Tempo

Mas

sa p

rata

equ

ival

ente

[oz]

Massa de prata equivalente e Massa de minério extraída - CEA

Massa de minério extraída Modelo 1 Massa de minério extraída Modelo 3 Massa AG equivalente Modelo 1 Massa AG equivalente Modelo 3

20000

40000

60000

80000

100000

120000

140000

Onç

as P

rata

Equ

ival

ente

sFigure 19: Mass of silver produced and mass of ore extractedin the models 1 and 3

0

2000

4000

6000

8000

10000

12000

14000

16000

jan/

19

fev/

19

mar

/19

abr/

19

mai

/19

jun/

19

jul/1

9

ago/

19

set/

19

out/

19

nov/

19

dez/

19

jan/

20

fev/

20

mar

/20

abr/

20

mai

/20

jun/

20

jul/2

0

ago/

20

set/

20

out/

20

nov/

20

dez/

20

jan/

21

fev/

21

mar

/21

abr/

21

mai

/21

jun/

21

jul/2

1

ago/

21

set/

21

out/

21

nov/

21

0

20000

40000

60000

80000

100000

120000

Mas

sa d

e m

inér

io e

xtra

ída

[ton

]

Tempo

Mas

sa d

e pr

ata

equi

vale

nte

[oz]

Massa de prata equivalente e Massa de minério extraída - DBS

Massa de minério extraída Modelo 2 Massa de minério extraída Modelo 4 Massa AG equivalente Modelo 2 Massa AG equivalente Modelo 4

Figure 20: Mass of silver produced and mass of ore extractedin the models 2 and 4

The raw data from the sequencing tables weresummarized in the table 6

8

Table 6: Mass of silver produced in the four extraction modelsParameter/Model Model 1 Model 3 Diference [%]

Mass of silver produced [oz] 688 977 717 395 4.13Parameter/Model Model 2 Model 4 Diference [%]

Mass of silver produced [oz] 573 540 734 223 26.02

This parameter saw an increase of 4.13% in themodels 1 and 3, in line with the increase in themass of ore extracted, and an increase of 26.02%in the models 2 and 4, significantly above the per-centual increase of the mass of ore extracted inthese models. This last value can be linked to anincrease in dilution in conjuntion with the increasein areas economically extractable.

5.1.3 Dilution

The main parameter analyzed in this study wasthe mine dilution. The dilution in the four extrac-tion models can be divided in three diferent types,the extraction dilution from the extraction of thevein, the development dilution, from the develop-ment of the mining method drifts in economicallyextractable blocks, and the crown pilar dilution, setat 20%. The Total dilution in the model aggregatesall the three dilution types. The raw data from thesequencing table is concised in the table 7

Table 7: Dilution values calculated in the four generated modelsParameter Model 1 Model 2 Model 3 Modelo 4

Extraction dilution 22.10 26.47 22.18 22.07Development dilution 86.20 80.08 39.06 68.50Crown pilar dilution 20.00 20.00 20.00 20.00

Total dilution in the model 24.80 35.21 22.84 29.87

The data goes in line with the predicted resultswith a deccrease in dilution between non-optimizedand optimized models. The models generated ac-cording to the Overhand cut & fill methods sawa percentual decrease of 7.90% from 22.07% to20.00% and the models generated using the Long-hole stoping, the percentual decrease was evenmore significant, set at 15.17%, from 68.50% to29.87%.

5.2. Economical analysis5.3. Operacional cost of extraction, net and gross

profitthree of the mains economic parameters are theoperational cost of extraction (CE), the net profit(RL) and the gross profit (RB), all provided by thesequencing tables. The relation between these pa-rameters is established by the equation 5.

RL = RB − CE (5)

The data from the aplication of this equation tothe sequencing tables is summarized in the table 8.In this table is visible the significant increase in allthe economic parameter with the Longhole stopingmodels presenting a higher level of increase, in linewith the previous discussed results. According to

this data, the prefered model that would allow themaximization of the net sales is the model 3.

Table 8: Values from the Operational cost of extraction, net andgross sales in the four generated models

Parameter/Model Model 1 Model 3 Diference [%]Operational cost of extraction 806 245 833 842 3.42

Net sales 4 959 863 5 172 008 4.28Gross sales 5 766 178 6 005 850 4.16

Parameter/Model Model 2 Model 4 Diference [%]Operational cost of extraction 723 990 837 595 15.69

Net sales 3 965 389 5 041 330 27.13Gross sales 4 689 378 5 878 925 25.37

5.3.1 Cash-flow and Net present value

To analyze the models in light of the NPV maxi-mization strategy, a series of cash-flows were gen-erated considering an interest rate of 8% anuallyand using the equation 6 where CF is the cash-flow, Rb is the gross sales, COP is the operationalcost of extraction and CIN is the investment in de-velopment to reach the ore.

CF = Rb − COP − CIN (6)

.The generated cumulated cash-flows, NPV are

presented in the figure 21

$-4000 000.00

$-3000 000.00

$-2000 000.00

$-1000 000.00

$-

$1000 000.00

$2000 000.00

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36

Valo

r [us

d]

Tempo [meses]

NPV acumulado

NPV acumulado Modelo 1 NPV acumulado Modelo 2 NPV acumulado Modelo 3 NPV acumulado Modelo 4

Figure 21: Accumulated cash-flows (NPV) for the four extrac-tion models

The results presented strengthen the alreadyverified level of optimization that can be achievedin the Longhole stoping optimization, seen whencomparing the models 2 and 4, specially when con-sidering that the model 2, due to the higher levelof dilution and higher cost in the invesment of in-frastructures to reach the vein, presents a nega-tive cash-flow, a situation not verified in the model4, already economically viable. Altough this opti-mization, the model that can provide a better op-tion for the extraction of the selected study area isthe model 3, the optimization of the Longhole stop-ing. This model allows the extraction of the area,with a known and already in use mining method,while reducing dilution, increasing the mass of sil-ver produced and maximization of the NPV througha higher level of possible net sales. All the param-eters displayed a better performance in the opti-mized models reinforce the idea that the optimiza-tion of extraction is a good aproach to increase theeconomic viability in orebodies of this nature.

9

6. Conclusions6.1. General conclusionsThe four generated models, and the endgame re-sults from the sequencing were in line with the the-orethical prediction. The magnitude of the anal-ysed extraction parameters are not far from whatwould be achieved in a regular mine plan to anarea with those characteristics which is proof thatthe used methodology was applied correctly. Thefour generated extraction models allowed the studyto fully understant and analyse the potential difer-ences between them not only from a o operationalstandpoint but also from the potential economicoutput that each one of them can generate.

The diference in height between the block modeldimension of 5 meters , and the proposed anddesigned optimized drift dimension, of 4 meters,means that a portion of the extraction area is cal-culated, through the process, as a extraction drift,where in reality, the extraction is executed throughone of the mining methods used. Due to the higherdilution value associated with the extraction drift,the extraction total dilution of the model is overes-timated, further increasing the conservadorism ofthe optimized models.

According to the economic analysis, the modelthat can maximize the profit in the selected studyarea is the model 3, altough the significant opti-mization that can be achieve with the optimizationof the longhole stoping, simulated in the model 4.

6.2. Limitations and achievementsThe applied methodology and the changes intro-duced to the usual reserves models generationproduced a set of realistic results and extractiondata, in line with the data usually produced in aarea similar to the selected study area. The pro-posed optimization plan, of reducing the dimensionof the extracion drifts, allow, if well executed, thedecrease of dilution, the increase of mineable ar-eas and the optimizaion of the NPV and economicparameters.

The selected study area was corrected chosenand the results from this study can be applied inother areas with similar vein thickness.

One of the main limitations in this study was thelearning curve of the Datamine software, limita-tion that increased the duration of the experimentalstage and lead to a decrease of the depth of theeconomic analysis, a stage that was planned to bemuch more focused and developed in the beginingof the study process.

6.3. Recomendations and future worksThe data from this study sugests that a changein mining method, altough inadvisable, can beachieved is well planned and executed. In thatsense a full study of the implementation of an al-

ternative and experimental mining method, like theAlimack method, can provide aditional data and apossible alternative with good results.

The inexistence of a drilling and blasting patternin the Saucit mine and the need for the adaptation,without previous plan, of the drift dimensions, in-creases ineficiencies and dilution. A future mineplan, with optimized drift dimensions, will need adetailed plan with a specific drill and blast patternfor the optimized drift dimensions.

The inability of the geologic block model to con-sider a 4 meter height block, in the optimized driftdimensions, provide further conservatism to themodels but also increases the diferences to a pos-sible real extraction of the area. To accommodatethis diference a adjustment to the code that gen-erate the sequencing can be made to increase thereliability of the generated data.

The main mining parameter analyzed, the dilu-tion, is measured in situ bidimensionally. In thatsense, altough the relatively realiability of the dilu-tion database, a more eficient and precise methodfor measuring the dilution is needed to further de-velop the dilution database from the Saucito mine.

At last, this methodology, and the models gener-ated, are theorethical representations of a extrac-tion. To further adjust the models and the method-ology a pilot area is needed, fully developed us-ing the dimensions of the optimized models, to in-crease the reliability of the models and identify theneeds for adjustment in the models.

References[1] JORC, 2012 Edition. Australasian code for

reporting of exploration results, mineral re-sources and ore reserves, (THE JORC CODE,2012 Edition). Prepared by the Joint Ore Re-serves Committee of the Australasian Instituteof Mining and Metallurgy, Australasian Instituteof Geoscientists and Minerals Council of Aus-tralia (JORC), 2012.

[2] Revuelta M.B. e Jimeno C.L. , Manual deevaluacion y diseno de explotaciones mineras,Madrid 1997.

[3] Sinclair, A. J. et Blackwell, G.H. . Applied Min-eral Inventory Estimation. Cambridge Univer-sity Press, New York, 2006.

[4] SME - Mining Engineering Handbook, ThirdEdition, 2011.

[5] Smith, C.L. e Appleyard, G.R. - Non-ResourceInputs to estimation of ore reserves - The Mod-ifying Factors, Mineral Resource and Ore Re-serve Estimation - The AusIMM Guide to GoodPractice - The Australasian Institute of Miningand Metallurgy, 2001.

10