Building Colombia’s first caving mine A McCauley AngloGold Ashanti, Australia G Dunstan

Caveman Consulting Pty Ltd, Australia L Nienaber AngloGold Ashanti, Australia Abstract The
Nuevo Chaquiro (NC) deposit is part of the Minera de Cobre Quebradona Project (MCQ)
located in the southwest of Antioquia, Colombia, approximately 104 km south-west of
Medellin. The nearest municipality is Jerico, which is located 7 km north-west of the ore
deposit. AngloGold Ashanti completed a pre-feasibility study in early 2019 that supported the
utilization of sublevel caving (SLC) mining method on the maiden ore reserve at an annual
production rate of 6.2 Mtpa. The life of mine (LoM) is forecasted to be a 20-year period with
extraction of over 120 Mt of ore and 7 Mt of waste. The MCQ deposit is a large blind copper,
gold, silver porphyry style with a ground surface elevation of 2,200 masl (on mountain) and
approximately 400 m of caprock above the economic mineralisation. Due to the caving
constraints of the deposit the first production level to initiate caving (undercut or UC) was
located approximately 100 m below the top of the mineralisation at 1675 masl (~525 m below
the top of the mountain) and the mining block extended approximately 550 m in depth (20
production levels at 27.5 m interlevel spacings). The main ore transfer horizon is located 75 m
higher in elevation than the mine access portals at 1080 masl and the proposed valley
infrastructure (in valley). The initial mining block will be accessed by twin tunnels developed in
parallel for 2 km at which point a single access ramp will branch up towards the undercut; the
twin tunnels will continue another 3.7 km to the base of the SLC where the crushing and
conveying facilities will be located. This paper will discuss the challenges associated with
designing Colombia’s first large scale underground mine in a unique geographical setting with
various operating constraints from mining regulations, local government legislation, and
operating logistics. It will explain how these constraints were overcome through the design
process, and thus enabled an accelerated cave establishment plan to be developed that
challenges the standard SLC mine construction logic associated with geological and
geotechnical conditions, production ramp-up, materials handling, ventilation, services and
egress. 1 Introduction The Minera de Cobre Quebradona Project (Quebradona or MCQ) is the
first large scale hard rock underground mine planned for development in Colombia. The
Colombian mining industry consists mostly of open pit operations and small underground
precious metal and aggregate mines. The Quebradona project has been undergoing drilling
and study since 2004. In recent years the AGA has focussed the exploitation methodology
toward a large scale (>5Mtpa) caving operation. Previous studies have focused on employing
either sublevel caving (SLC) or block caving (BC) or a combination of the two in order to exploit
the deposit with a bulk floatation process to produce a gold rich copper concentrate. The
district in which the project is located is dominated by coffee plantations and lifestyle
properties of wealthy Medellin families. P300 MassMin2020 718 2 Technical Parameters 2.1
Geology Nuevo Chaquiro is a copper-gold porphyry deposit located in the Middle Cauca Belt of
Colombia. The geology is comprised of a cluster of Miocene stocks and dykes of diorite and
quartz diorite which intrude into flat-lying coeval andesitic tuffs and volcaniclastics. None of
the intrusive rocks, excluding some late barren porphyries, breach the present-day surface
(Pablo & Anglo Gold Ashanti 2018). Two intrusive centres at depth, an early highly mineralised
quartz-diorite and a later intermineral diorite are the focus of alteration and mineralisation at
Nuevo Chaquiro. There is a well-zoned alteration system comprised of early propylitic, potassic
and calcic-potassic assemblages, which have been overprinted by a lower-temperature
sericite-chlorite alteration, and a broad phyllic (quartz-sericite) blanket at surface to 400 m
depth. The Nuevo Chaquiro porphyry copper-gold deposit has an estimated mineral resource
(indicated + inferred) of 602 Mt containing 0.73% Cu, 0.67 g/t Au and 4.08 g/t Ag (constrained
by the underground mining envelope) based on the 2019 AGA resource model (AngloGold
Ashanti 2019). Figure 1 shows the spatial arrangement of the quartz diorite dykes at Nuevo
Chaquiro. The pre-mineral body is in the upper middle part of the mineralised set, and is a
cylindrical body with a width of 80 to 120 m. The body itself is weakly mineralised, averaging
around 605 nppm Cu with a minor Au mineralisation. Early quartz diorite hosts the bulk of the
>0.6% Cu copper mineralisation. The mineralisation is associated with intense quartz veins and
stockwork comprising >10% of the rock volume. In the logs it was locally identified as a cupola
zone. Mineralisation extends into the volcanoclastic host rocks adjacent to early quartz diorite,
especially near the apex. Figure 1 Isometric view of dyke arrangement (AngloGold Ashanti
2018) MassMin2020 719 The Nuevo Chaquiro deposit is centred on a composite stock of
quartz diorites. At least four kinds of intrusive bodies are present. The strongest mineralisation
(>0.6% Cu) is associated with the early quartz diorite. The other intrusive rocks show moderate
copper mineralisation and contribute to the geometry and define the western extents of
weaker mineralisation (>0.45% Cu). The alteration and mineralisation styles at Nuevo Chaquiro
are typical of a telescoped porphyry copper system, where early potassic alteration and quartz
vein stockwork development has been subsequently overprinted by sericite-chlorite alteration
within much of the diorite. Additionally, an extensive quartz-sericite-pyrite blanket developed
at a shallower depth above the diorites in the volcanoclastic host rocks. The sediment
succession in the study area is mainly comprised of claystone; siltstones of reddish-brown
colorations, with the presence of calcareous concretions, and of continuous parallel flat
lamination; and fine sandstones of reddish-brown colour, well calibrated and with a clay matrix
and calcareous cement. Occasionally sediments of the conglomeratic type, such as
sedimentary breccias, are present in the south and northeast sectors of the study area. The
above description corresponds to the tertiary Amagá Formation which is in discordant contact
with the Combia formation; these two units overlay the colluvial quaternary deposit named
the Cauca Viejo Formation. See Figure 2 for integrated geological information (Figure 2A in
blue are places where drilling was executed) and to the right geological contact between the
Amagá and Cauca Viejo formation (Figure 2B). The Cauca Viejo Formation corresponds to
recent colluvial deposits, formed by the deposition of rock fragments from the Combia
formation. The Combia formation in the depression of the Cauca river basin is a result of the
neo-tectonic contribution of the Arma fault system and its satellites (Figure 3). According to
the geotechnical exploration studies, the Cauca Viejo Formation has a thickness that varies
from approximately 110 m, corresponding to the sector of greater accumulation near one of
its dejection cones in the upper part closest to the escarpment of the Combia Formation, to a
null thickness in the flat areas. In general, thickness averages of 28 m. It is composed of sub-
angular to sub-rounded fragments of volcanic breccias and tuffs of generally fresh crystals of
grey, greenish-grey and oxidised patches, supported on a heterogeneous matrix, grading from
brown claystone to fine brown sandstone. Figure 2 Amagá and Cauca Viejo Formation –
Contact Relationship, (AngloGold Ashanti 2018) MassMin2020 720 Figure 3 Regional Geology
Quebradona district showing: Nad- intrusive bodies, Nmc- volcanoclastic sequence, late
Miocene (Combia Formation); Nos – sediments (with coal layers), medium Oligocene (Amagá
Formation); Kvo-Kv igneous and sedimentary rocks, early Cretaceous (Quebrada Grande
Formation); Kies – metamorphic sequence (Arquia Complex). Modified from the Geologic map
of Colombia 2015, SGC. Source: (AngloGold Ashanti Geology 2020) 2.2 Geotechnical 2.2.1
Domains Geotechnical domain boundaries and rock mass properties were assessed based on
diamond drilling and local knowledge gained from surface mapping. The rock mass model has
taken the following key elements into account, viz., geological structures, lithology, alteration,
RQD, strength, joint condition, spacing and frequency, alteration, and mineralization based on
the Q-system of rock mass classification. A comparison of implicit and explicit modelling logic
was undertaken to assess the accuracy of each method in regards to the drillhole data. The
explicit modelling returned a higher accuracy (SRK Consulting Inc. 2020). The domains within
the proposed mining area are numbered and lettered according to their combination of
structural and alteration units, as shown in Figure 4. Domain 1: Surface “blanket” of saprolite
material (10m to 40m thick) Domain 2: Sericite supergene “-S” – Sericite Domain 3A: Above
the fractured or relay zone “-CS” – Chlorite- Sericite “-P” – Potassic Domain 3B: Below the
fractured or relay zone “-CS: - Chlorite-Sericite “-PA” – Potassic-Actinolite MassMin2020 721
Domain 4: Volume associated with the Set 4 faulting and fractured zone (relay zone) “-CS”-
Chlorite-Sericite “-P” – Potassic Domain 5-SCS and 6-PA are north of the mining area with
sparse drillhole data, with the upper area mostly in Domain 2-S, hence defined separately as
Domain 5-SCS Figure 4 Top: View looking north (left) and (right) northwest showing the ten
geotechnical domains. Bottom: Cut-out view looking north (left) and northwest (right) showing
the chlorite sericite domains (SRK Consulting Inc. 2020) 2.2.2 Caving and Stability As
Quebradona is “perched” above sea-level the regional in-situ stress values are relatively low
above the orebody, but will become more pronounced with depth as confinement increases.
Without reliable insitu stress measurements, the base case principal stress orientation and
magnitude are as follows based on (Sandy 2020): σ1 ≈ 1.7 ρgD oriented horizontal EW σ2 ≈ 1.3
ρgD oriented horizontal NS σ3 ≈ σV = ρgD MassMin2020 722 The SLC will be initiated in the
North West corner and progress to the southeast in accordance with the lead-lag rules to
manage stress changes and maximise the promotion of caving in the tuff domain. This
orientation of the cave front also ensures its future stability as it crosses the various structures
through the orebody. The critical span requirements can be seen in Figure 5. The span shown
here is approximately 50% of the final designed span. To further enable consistent caving, the
upper 3 levels of the SLC have been designed with a consistency of shape, both vertically and
horizontally. Initiating the cave in the northwest corner places it below a low point on the
surface (the surface topography varies by ~150 m across the mining footprint) and allows the
Fea Creek structure to be used to assist cave initiation (Figure 6). Cave initiation will be largely
dictated by the more competent rock mass directly above the undercut level (Domain 3A-CS
and 3A-P). Once caving propagates through this rock mass, sustaining caving in Domain 2-S
above is unlikely to be a concern due to its lower quality rock mass. Figure 5 shows Domain 3A-
CS and 3A-P, shaded green and Domain 2-S shaded yellow in (SRK Consulting Inc. 2020). Figure
5 Hydraulic radii of the cap rock geotechnical domains: Domain 2 (yellow), Domain 3A-CS (pale
green), and Domain 3A-P (dark green) (Laubscher 2000) MassMin2020 723 Figure 6 Level Plan
showing Geotechnical Domains (SRK 2020) The key long-term infrastructure including the
internal access decline, primary crusher, ventilation fans and shafts have been located outside
the areas influenced by the changes in stress due to the application of SLC. Despite the easterly
plunge of the orebody toward the decline, the competent nature of the rock mass at
Quebradona is expected to remain stable, limiting the influence of the relaxation zone. 2.2.3
Subsidence Estimate As Quebradona will be Colombia’s first caving mine, subsidence (ground
disturbance) has been a key focus. Caving is a dynamic process with complex inter-
relationships, and estimating cave propagation and subsidence has inherent uncertainty,
especially at the greenfields stage. Figure 7 shows a plan and section estimate of the final
subsidence crater and cave column. The cave is expected to breakthrough to the surface
approximately 3 years after the first production slot firing on the undercut level. In the
following years as production continues the subsidence cone will continue to widen and
deepen in line with the tonnes extracted. A comprehensive monitoring system has been
allowed for to ensure redundancy in tracking the cave’s progression to the surface. This
monitoring will continue for the life of mine to ensure there are no hang-ups or discrepancies
between drawn material and the growth of the crater (Figure 7). A series of surface diversion
channels will be installed to redirect the rainfall runoff water so that it does not enter the cave
column, and subsequently the mine works. A security fence will also be erected to ensure
there is no unauthorized entry to the active subsidence area. Due to the sensitivity of the
regulator and the local community to the caving method proposed for Quebradona and the
resulting subsidence crater, AGA engaged SRK (Vancouver) to assist in the prediction of the
impacts of mining. With the limited data (drillholes only, no underground exposure) SRK used
the geotechnical domains as inputs to empirical methods (Laubscher 1990; Jakubec &
Laubscher 2000) to determine the subsidence estimate (Figure 8, Figure 9 and Figure 10).
MassMin2020 724 Figure 7 Geotechnical monitoring of the cave & subsidence zone (plan and
section views), (SRK Consulting Inc. 2020) Figure 8 Views looking north (left) and northwest
(right) of the SRK 2020 geotechnical domain model showing the seven geotechnical domains
(Domain 1, not shown in the legend, is in orange) MassMin2020 725 Figure 9 East-west section
showing the geotechnical domains, major structures and design development (SRK Consulting
Inc. 2020) Figure 10 North-south section showing the geotechnical domains, major structures
and design development (SRK Consulting Inc. 2020) MassMin2020 726 The depth extent
details and rock mass rating values are in Table 1. The assessment was done for the 35th and
65th percentiles, representative of the range of rock mass quality within each domain. Table 1
Depths and rock mass properties used in the subsidence estimate (SRK Consulting Inc. 2020)
Domain Depth From m (RL) Depth To m (RL) IRMR (P35) IRMR (P85) Weathering Orientation
Stress MRMR (P35) MRMR (P65) 1 0 (2,150) 50 (2,100) - - - - - 20 20 2 50 (2,100) 300 (1,850) 44
51 90 90 100 36 41 3A 300 (1,850) 550 (1,600) 57 62 96 90 100 49 54 4 550 (1,600) 700 (1,450)
54 61 90 85 100 41 47 3B 700 (1,450) 1,000 (1,150) 59 65 100 100 90 53 59 2.2.4 Subsidence
Angles Laubscher’s empirical method was used to estimate the extent of the final subsidence
zone. For Laubscher’s subsidence method, two charts are available (Figure 11). The first chart
(Fracture Zone) is typically used for the purposes of draw control and water inflow estimation.
The second chart (Infrastructure Zone) is more appropriate for locating the boundary of the
stable zone. It is important to note that the empirical subsidence angle estimation process is
an iterative one since theoretical backfill of the crater must be estimated, which in turn
influences the subsidence angle. An extract from the SRK Laubscher subsidence calculation
spreadsheet showing the input values and results are in Table 2 (P35) and Table 3 (P65). Table
2 Caving calculation spreadsheet for P35 rock mass values (SRK Consulting Inc. 2020) Table 3
Caving calculation spreadsheet for P65 rock mass values (SRK Consulting Inc. 2020) The overall
angles calculated for the two empirical charts are 74 degrees and 69 degrees respectively for
the 35th percentile case, and 76 degrees and 71 degrees for the 65th case. The results are
illustrated in Figure 11. MassMin2020 727 Figure 11 SRK empirical subsidence estimates (SRK
Consulting Inc. 2020) Several iterations of subsidence crater depth and angles were done to
calculate volumes. It appears that a crater approximately 150 m deep and 800 m in diameter
roughly matches the caved material minus the mined ore volume when considering the 65th
percentile rock mass values. For the 35th percentile the crater is expected to be wider and
hence shallower, with a crater approximately 100 m deep and 950 m in diameter. SRK
developed 3-D fractured zone volumes at several stages of the mine life. This was done using
the subsidence estimate profiles from SRK (July 2019) because the SRK geotechnical model had
not been finalized then. The fractured zone subsidence profile in SRK (July 2019) is close to
that of the 65th percentile rock mass values. The generated volumes and comments on the
possible cave behaviour are in Figure 12. It is acknowledged that the approach does not
explicitly account for the effect of major structures. Also, for simplicity and due to the amount
of information available, the volumetric study assumed that overhangs associated with the
sublevel stepouts will not form. The topography over the orebody is undulating and rises most
steeply to the east. The authors are not aware of any established empirical method for
quantifying the caving response to irregular topography, however a benchmarking study by
Woo et al. (2013) found that the greater the range in topography, the higher the degree of
asymmetry in the caving angle (Figure 13). MassMin2020 728 Figure 12 Images from the SRK
caving volume study showing the possible progression of the fractured zone and crater (SRK
Consulting Inc. 2020) MassMin2020 729 Figure 13 Undercut depth versus caving angle for
block and panel caving operations showing higher ranges in caving angle in areas of irregular
topography (blue) than flat (red) (Woo 2013) 2.3 Mining 2.3.1 Method During the pre-
feasibility study some key outcomes were locked in, including the use of sublevel caving at a
mill throughput rate of 6.2 Mtpa (ore and waste) and a maximum TSF capacity of 119.2 Mt.
During the feasibility study a review was undertaken to identify the most economic ore
extraction strategy, considering geotechnical and caving constraints to ensure a high reliability
mine plan. The SLC draw plan was constrained in the early levels to ensure the presence of an
ore blanket and also to ensure that the risk of an airblast was mitigated. Further, from the post
caving breakthrough to surface the extraction rate was increased to a maximum of 150% on
any level. As the higher grades move towards the top of the SLC the draw percentage can be
increased, and a higher value of the material can be extracted. 2.3.2 Design and Sequencing
The SLC production level design utilizes the “panel” layout to improve safety and production by
eliminating high speed interactions through the separation of the production loading activities
from the other supporting activities occurring on each level. This is done by placing the ore
passes on the orebody side of the perimeter drive (Figure 14), thus ensuring that the
production LHDs are captive to their panel when undertaking their primary task. Each panel
has a primary and back-up pass available to the LHD. Access in and out of the other panels on
the level is possible via the perimeter drive, which is used for person and material movement
only, and no primary production tasks. MassMin2020 730 Figure 14 Typical SLC Production
Level (AngloGold Ashanti 2020) The internal decline and all critical infrastructure are located to
the east of the production footprints. The levels are accessed from the internal decline slightly
to the north of the level centre, as the first production rings fired on a level are in the
northwest. The lead lag rules for the cave generate a straight line at 30 degrees across the
level. The cave front follows the orebody trend. Due to the orientation of the ore levels
relative to the initial mine access ramp and ventilation locations there was an opportunity to
develop a bypass drive to the north of the footprint. This provides early access to the slot
drives on the upper 3 SLC levels. Early slot access enables the early establishment of a
“stacked” mining front across 3 vertical levels and allows an accelerated ramp up of production
once the full materials handling system (MHS) is commissioned. See Figure 15 for a
comparison against other similar SLC operations. Figure 15 SLC Ramp Up comparison chart
(AngloGold Ashanti Mining 2020) MassMin2020 731 Each of the production levels, both on the
production levels the individual panels, are independent ventilation districts. Fresh air is fed
directly from the surface to the secondary fans located in a wall on the northern end of the
perimeter drive and is then ducted directly to the individual panels. This circuit design
eliminates recirculation and enables activities to continue, with minimal disruption, when
blasting occurs on other levels. 2.3.3 Equipment Selection Due to the ventilation requirements
in Colombia the mining fleet selected for Quebradona is predominantly electric at 1,000V
because of the use of electric cabled loaders. Fourteen tonne loaders were selected for the
production levels based on their speed, bucket size (enables side to side loading in the crosscut
and identification of oversize material) and cable length. Twenty-five tonne loaders were
specified for the transfer level to accommodate the shorter tramming lengths and limited
operating areas (there are a maximum of 2 loaders per side of the crusher due to the layout).
As battery technology improves in the coming years the selection of loader sizes may change
as additional options become available. The selection of the present Sandvik fleet was
predominantly based on the electric loaders and their ability to provide other front-line
development and production machines required to undertake SLC mining. This decision also
took into account the use of automation for the majority of production activities, so utilising a
common platform is seen as the most pragmatic at this stage. It has been proposed that the
maintenance of the machines be done by Sandvik under a maintenance and repair style
contract (MARC) since there is a heavy reliance on their equipment and systems. This will also
reduce the need for AGA to identify highly skilled maintenance technicians to keep the fleet
operational, as they will then have the backing of the factory and worldwide access to trained
resources. 2.3.4 Underground Fixed Infrastructure 2.3.4.1 Key Infrastructure An integrated
materials handling system (MHS) for the SLC was designed from the ore pass grizzlies, located
on the production levels, to the process plant. Due to the length of the ore passes (up to 500
m), and the predicted comminution expected by the time the rock appears on the transfer
level, larger than industry standard grizzly apertures of 1,500 mm were selected. The design
criteria for the underground crusher was that it needed to reduce the ore to a size suitable for
placement on the conveyor belt and delivery to the surface coarse ore stockpile (COS), after
which a secondary crushing prior to delivery at the process plant will be undertaken. Assuming
the maximum size reduction ratio for the crusher of ~6:1 at a throughput rate of 6.2 Mtpa, a
51” Gyratory crusher was selected. This crusher is also suitable to support block cave (BC)
mining should the conversion of mining method occur. Figure 16 shows the proximity of mine
development and the final subsidence area. MassMin2020 732 Figure 16 SLC Long Section
Showing Crusher Location and Final Subsidence Zone (AngloGold Ashanti Mining 2020) A 3-
stage tramp removal system was located after the crushed ore bin and prior to the ore feeding
onto the primary conveyor belt to the surface. Due to the location of the surface COS the
portal conveyor belt is approximately 6 km long at a very slight decline, and two belt turning
stations were included on the return side of the belt to minimize spillage, and idler and belt
wear. Figure 17 shows a schematic view and data sheet for the system and Figure 18 shows
the cross section through the conveyor tunnel. Figure 17 Materials Handling Flowsheet
MassMin2020 733 Figure 18 Conveyor Tunnel Cross section (showing dewatering
configuration) The primary dewatering system is located near the tramp removal station and is
comprised of a dual stream, 60l/s and 120l/s, dirty water pumping system. The dewatering
pumps are fed from the production levels via a dewatering raise which directs the water to the
vertical dam, located between the transfer level and pump station. An agitator is located in the
vertical dam to keep the solids suspended. The dual stream system is to ensure that the
velocity of water flow in the pipes is sufficient to maintain the suspension of the solids in the
relatively level pipelines. The primary ventilation fans are located underground to minimize the
noise and visual impacts associated with locating them on the top of the mountain. The
underground installation allows the supporting electrical and communications networks for
the entire mine to be streamlined. There will be 2 main workshops located inside the mine,
one will be located toward the top of the mine to service the development and production
equipment, the other will be located on the transfer level and will service both the transfer
loaders and the crusher. Satellite service bays will be located on the perimeter drive of each
SLC level for the production loaders. The workshop strategy is based on maintaining all primary
mining equipment underground to limit the amount of traffic in the access tunnels. A surface
workshop and warehouse will be required for the maintenance of fixed plant equipment.
2.3.4.2 Basis of Engineering The fixed infrastructure has been designed using the data and
observations from at least 6 preceding international caving mines and projects. This design
therefore benefits greatly from them in both the functionality and constructability areas. Front
end loading of key areas enables greater levels of certainty to be applied to those activities
which have the potential to cause timeline and cost increases later in the project. 2.4
Metallurgical The process plant will include high pressure grinding rolls (HPGR) as the main
crushing unit on the surface; it will be supported by a secondary crusher to deal with oversize
material. The ore then feeds to a ball mill before being discharged to the floatation circuit. The
gold enriched copper concentrate will be piped to the filter plant for drying and the removal of
water down to a moisture content of 10%. The tailings will be segregated to pyrite and non-
pyrite streams before being distributed to one of two filter presses. Dry stacking of the tailings
will be utilised, with the pyrite bearing tailings being encapsulated within the larger inert
tailings footprint (AngloGold Ashanti 2018). MassMin2020 734 The concentrate will be trucked
400 km to the pacific coast port for loading on to ships for sale to the international smelter
market. Due to the volume of concentrate generated there will be an estimated 20 truckloads
of concentrate per day at the peak of production. 2.5 People AGA will base its human
resources program around the company’s 5 principles of “How we work” – Strata: the right
people in the right roles, effective labour relations, 3 level management structure and
responsibility. As the mine is a multigenerational operation its ability to act as a catalyst to
improve the education and economic levels of the local community cannot be understated.
There is currently a concern from the local agribusiness providers that many of their workers
will abandon their current employment to join the mine. AGA is aware that to be successful in
the community it needs to balance the expectations of everyone in the community and has
been conducting focus groups to understand the community issues. Many of the skills required
to support the mine site are non-existent in the local community. To overcome this AGA has
engaged with local training providers to identify potential future employees and train them in
the specific mining and processing tasks that will be required during the construction and
operation of the mine. The timeframe for this training and identification to form a preferred
talent pool is approximately 3 years. Upon completion AGA will have a list of personnel in the
area who have the basic skills for working on the site, this list will also be provided to the
contractors engaged to work on the site to accelerate their recruitment processes. The
Colombian mining and industrial relations regulations (Colombia Ministry of Mines and Energy
2015) provide some challenges when designing a shift and roster structure for an underground
workforce. Until environmental conditions can be confirmed an 8-hour shift is the longest
possible for underground work. To overcome this constraint as many crucial production tasks
as possible were relocated to the surface and aided by technology. Primarily these roles were
production and transfer loading, production drilling and shift monitoring. When located on the
surface a 12 hour rotating roster is permissible for the workforce. 2.6 Government
engagement As part of the mining lease application process a detailed, and structured,
Environmental Impact Assessment (EIA) process must be followed. This government
engagement process is crucial for AGA to show not only the professionalism of the company
and its team members but also the technical and commercial benefits of the project for
Colombia. To support this a very detailed EIA submission was made in late 2019 and the
evaluation process is underway at the time of writing this paper. As part of the process the
relevant authorities and their nominated consultants are to be hosted on the proposed mine
site for in excess of a week to visualise and understand the impacts of the proposed mining
process and the location of critical infrastructure. In addition, part of the approval process is
having the mining method endorsed by the AIMC (Colombian Association of Mining Engineers).
The AIMC is made up of retired mining personnel who act as quasitechnical specialists for the
government. One of the challenges faced by the project is that there is no caving experience in
Colombia and the majority of the mining reviewers are former open stope miners who
extracted aggregates for the construction industry. As such, many of the initial questions were
related to why open stoping was not selected. Following a number of engagements, the AIMC
now endorses SLC and has offered to assist in socialising the project as required. MassMin2020
735 2.7 Execution and Path Forward The feasibility study for the project is due for completion
in Q4 2020. Pending a successful outcome the proposed site execution works program will
commence in Q3 2021, resulting in less than 12 months for AGA Colombia to transition from a
relatively small study and community/government relations team to a moderately large
project execution team. The initial underground works are relatively straight forward: a couple
of portals and parallel tunnels. It is the initial site establishment work which requires close
attention. The early works program includes, but is not limited to: the establishment of site
access roads, laydown areas and workshops, material stockpile areas, offices and change
houses with the associated ablutions and services, a medium voltage power supply and a raw
water supply. Acknowledgements The authors would like to thank AGA for their permission to
publish this paper. Additional thanks to all those involved with the feasibility study and EIA
processes for the Quebradona. Without your hard work and dedication this project would not
be where it is.