The Meliadine gold project is located in the Archean Rankin Inlet Greenstone Belt, in the Churchill Structural Province of the northern Canadian Shield. This belt consists of mafic volcanic rocks, felsic pyroclastic rocks, sedimentary rocks and gabbro sills of about 2.7 billion years old that have been polydeformed and metamorphosed. The Meliadine trend is defined by northwest-trending rock units as well as a major fault zone known as the Pyke Fault: a high-strain zone characterized by multiple foliations and regionally important shear zones.

Seven gold deposits have been identified at the Meliadine project, of which Tiriganiaq is the most significant. The other deposits are the Discovery, F zone, Wesmeg, Normeg, Wolf and Pump. The Normeg deposit, discovered in 2012, is a southern expansion of the Tiriganiaq deposit, and extends as far as the Wesmeg deposit. Gold mineralization in these seven deposits is mostly mesothermal quartz-vein-dominated gold systems in strongly sheared and complexly folded host rocks of Archean turbidites, iron formation and volcanic rocks. Within each deposit are many gold-bearing lodes of quartz vein stockwork, laminated veins and sulphidized iron formation with strike lengths of up to 3 km. The stratigraphic units generally strike northwest-southeast, dip steeply to the north and are overturned, with the oldest units to the north. All seven deposits remain open at depth.

Mineralization in the Tiriganiaq gold deposit is strongly associated with shearing and quartz veining, which likely developed during the Proterozoic third deformation event. The most intense and consistent gold mineralization is present within both the Upper Oxide Iron Formation and proximal to the Lower Fault in the siltstones of the Tiriganiaq Formation, but minor amounts of mineralization have been reported in all rock types. Areas of intense shearing are ‘healed’ by quartz veining and there is generally little to no brittle deformation.

Gold mineralization is directly associated with syn- to post-deformational quartz veining. Vein appearance is highly variable from a several-metres-thick laminated vein along portions of the Lower Fault to wispy, apparently erratic quartz stringers and stockwork in the Tiriganiaq Formation. With some exceptions, quartz-only and quartz-ankerite veins tend to be mineralized with gold, while quartz-calcite veins are commonly barren in all lodes. This observation implies to multiple veining events.

Sulphide minerals in the Tiriganiaq gold deposit are of two distinct generations: primary and late sulphides.

Primary sulphides consist of wispy, discontinuous, bedding-parallel laminations of pyrrhotite, pyrite and chalcopyrite in oxide iron formations and black argillite. Fine-grained, web-textured sulphides in oxide iron formations close to quartz veins may represent remobilization of primary sulphides, but may alternatively signify sulphur introduced during mineralization. Gold values are generally low to absent in this sulphide assemblage, in the absence of quartz veining.

Late sulphides consist of coarse-grained (sub-centimetre-size) clots of arsenopyrite crystals and pyrrhotite associated with highly gold-bearing zones located in and close to quartz veining. The coarse arsenopyrite crystals rarely show signs of strain, having suffered little post-depositional deformation. Arsenopyrite of this type, in association with quartz veining, is a good visual indicator of elevated gold values.

Visible gold is common in all lodes in the deposit and is present in quartz veins, pyrrhotite, and along the margins and late fractures of arsenopyrite crystals.

A conventional truck/shovel operation for both pits is planned with different mining approaches for ore and waste. The mining method must be suited to Arctic conditions and provide a good daily production in waste zones. To maximize the recovery of economic material and minimize dilution in the ore zones, the method must be very selective.

A long-hole mining method was selected to mine the Tiriganiaq deposit due to the shape, thickness and orientation of the orebody. The stope height was set at 25 m to minimize drill deviation and waste rock dilution, and to maximize mine recoveries. Two variations of long-hole mining (transverse or longitudinal mining) will be utilized; the variant selected for a particular stope will depend on the thickness of the ore zone.

A long-hole mining method will also be considered for the eventual exploitation of the underground indicated and inferred mineral resources located in other deposits (F zone, Pump and Discovery) not currently include in the LOM plan but with some variations because the lodes in these deposits are generally shallower-dipping and are narrower than the Tiriganiaq lodes. Longitudinal mining would be used for these deposits. Permanent rib pillars would be left between adjacent stopes, and uncemented rockfill would be used to fill the stopes. Levels would be 20 m apart (instead of 25 m) to minimize drill-hole deviation and dilution due to the shallower dip.

The proposed process plant will treat gold ore at a nominal process rate of 3,000 tpd from Year 1 to Year 3. A process plant expansion of 2,000 tpd will increase capacity to 5,000 tpd at Year 4 due to the addition of ore from the Tiriganiaq open pit operation. The Tiriganiaq deposit will be the dominant source of process plant feed for the plant during the early years. A combination of conventional gravity concentration and cyanidation processes has been selected for the project. The process plant will consist of three separate facilities:
- a crushing facility, including primary and related material-handling facilities that can crush ore from either underground or open pit operations;
- a main processing facility including a coarse ore surge bin and related feeding and reclaim systems, single-stage grinding, gravity concentration and an intensive leaching reactor, leaching feed thickening, CIL cyanide circuit, loaded carbon acid wash, elution and reactivation, gold electrowinning, smelting, cyanide destruction of the leach residue and filtration of final tails;
- a paste backfill plant.

The ore is trucked from the underground mine to a surface crusher located in the vicinity of the process plant. The trucks dump into a hopper equipped with a static grizzly and a hydraulic rock breaker that can handle any oversize rocks. An apron feeder is located under the dump hopper. The ore is fed to a reciprocating grizzly feeder via the apron feeder. Oversize rock from the grizzly feeder is fed to a jaw crusher. The jaw crusher product and the undersize from the reciprocating grizzly feeder are both transferred by gravity to the jaw crusher discharge conveyor, which then carries the material to a 2,500-tonnes-capacity ore bin adjacent to the process plant. The ore is reclaimed from the bin using an apron feeder that discharges onto the SAG mill feed conveyor at an average feed rate of 136 t/h.

A portion of the SAG mill discharge is diverted to the gravity circuit that is comprised of two centrifugal batch concentrators operating in parallel. Ahead of each concentrator is a vibrating screen that removes material coarser than 2 mm and returns it to the grinding circuit. The undersize from both screens feeds the centrifugal concentrators. The tailings from the concentrators are combined and returned to the cyclone feed pump box, while the concentrates are combined and sent to the intensive leaching reactor where they are leached prior to the pregnant solution being pumped to a dedicated electrowinning circuit. Space is also allocated, in the layout, for a third gravity concentrator unit.

The continuous gravity system is located outside of the grinding circuit. The concentrator is fed by the cyclone overflow after the removal of coarse material (+2 mm) by screening. All of the cyclone overflow is fed to the continuous gravity circuit.

The tailings report to the thickener while the concentrate is reground. The regrind circuit consists of a regrind mill in closed circuit with cyclones. The gravity concentrate would be reground at 80% passing 38 µm prior to being pre-aerated and leached through the leaching circuit.

The grinding thickener underflow is sent to a pre-aeration tank into which process air is sparged to achieve sulphide oxidation, and lime is added for pH adjustment. The pre-aeration tank is followed by five agitation CIL tanks in series. Leaching is performed using sodium cyanide, and pH is controlled with lime addition. All CIL tanks are also sparged with process air in order to keep the oxygen concentration sufficient for gold leaching.

Slurry travels through the CIL circuit via inter-stage pumping screens, while gold-loaded carbon is pumped counter-current to the slurry flow by carbon transfer pumps, to the previous CIL tank and finally to the loaded carbon screen. This step is performed regularly, based on the carbon gold loading. Space has been allocated to the west of the CIL circuit for future expansion consisting of adding three tanks of larger volumes. This expansion will maintain an appropriate residence time for the leaching process; the expanded cyanidation circuit will consist of two preaeration tanks followed by seven agitation CIL tanks in series.

Loaded carbon from CIL tank 1 is pumped to the loaded carbon screen. The screen undersize is returned to the same CIL tank while screen oversize feeds the stripping circuit. Loaded carbon flows by gravity from the loaded carbon screen to an acid wash vessel. A dilute solution of hydrochloric acid is used to remove calcium carbonate and other contaminants from the carbon. The hydrochloric acid solution is circulated in an up-flow manner through the carbon bed in closed circuit with the acid solution pump and acid wash tank. Once the washing step is complete, the solution is neutralized and discarded. The carbon is then rinsed with fresh water before it is transferred to a carbon stripping vessel. A pressurized Zadra desorption system is used to recover gold from the carbon. It is designed to run at approximately 140ºC and a maximum operational pressure of 380 kPa.

In the strip vessel, carbon is combined with hot strip solution. The overflow from the stripping vessel consists of pregnant (gold-bearing) strip solution that is pumped to electrowinning cells where gold is deposited onto stainless steel cathodes. The stripping cycle is operated for a predetermined period of time (typically 12 to 16 hours) and ends when the carbon loading reaches a low value. Barren solution is reused from one cycle to the other, but a solution bleed is periodically sent to the CIL circuit to control impurity buildups. Once the stripping cycle is completed, the carbon is rinsed and transferred to the reactivation kiln. The kiln thermally reactivates the carbon at ~700°C to remove organic compounds. Reactivated carbon is quenched with fresh water upon exiting the kiln. From there it is pumped to the reactivated carbon sizing screen to remove fines. As required, new carbon is added to this stream after having been attrited to remove fine particles. The screen oversize is returned by gravity to CIL tank 5, while screen undersize reports to the carbon fines collection tank.

Cyanide destruction is done in two tanks in series, using the SO2/Air process. The tails from the last CIL tank are directed to a carbon safety screen to recover any carbon that could have escaped the CIL circuit. The screen oversize is recovered into a bin, while the undersize is sampled before flowing to the cyanide destruction tank feed box that receives CuSO4, sodium metabisulphite (Na2S2O5) and lime. The slurry then flows by gravity to cyanide destruction tank 1 where it is agitated and sparged with process air. The tank overflows into cyanide destruction tank 2, where it undergoes further agitation and sparging with process air. The cyanide destruction tails, from the cyanide destruction tank 2, flow by gravity to a pumpbox where they are pumped to the filtration circuit.