The 2012 technical report describes the proposed process plant design for the Rosia Montana gold-silver deposit, a large tonnage low grade epithermal deposit with partially refractory ore characteristics.
Report context
This technical report, dated October 2012, documents the recovery methods and process plant design for the Rosia Montana Project. The project is a large tonnage low grade epithermal gold-silver deposit with gold:silver ratios generally averaging around 1:5 and as high as 1:10. The material has been shown to be partially refractory with precious metals associated with, and partially locked in, sulphide minerals, mainly pyrite. The process plant has been investigated through several studies and design phases including a detailed feasibility study undertaken by Minproc in 2001 and subsequent stages of preliminary engineering design undertaken by SNC, Bechtel Corporation (for whom the process plant design was undertaken by Ausenco Ltd.), the Washington Group, and most recently directly by RMGC. The current design presented by RMGC is based on the design developed by SNC during its period as EPCM contractor between 2006 and the end of 2007, further developed since 2007 as part of a Cost Optimisation Program (COP) and documented in the Front End Engineering Design (FEED) report prepared by RMGC dated September 2009.
Processing route
Proposed process flowsheet
The proposed process route comprises open circuit primary crushing, stockpiling of primary crushed ore with reclaim using apron feeders, and two stage milling (SAG and ball) to 80% passing 150 μm with provision for removal and crushing of critical size material in a pebble crusher (SABC circuit). Closed circuit classification uses hydrocyclones with the underflow returning to the ball mills. Gravity concentration is incorporated using high speed centrifugal concentrators on a portion of the hydrocyclone underflow, with intensive leaching of the concentrate through to production of electrowon gold and silver metals.
The milled product reports to a Carbon in Leach (CIL) circuit consisting of one stage of pre-leach and six CIL adsorption stages. Tailings dewatering occurs in a high rate thickener with recycle of overflow solution to the milling circuit. The tailings thickener underflow is treated in a cyanide detoxification circuit using copper sulphate and sodium metabisulphite (SO₂/air process). Loaded carbon is treated through carbon acid wash, continuous Zadra elution, and regeneration of stripped carbon. Gold recovery from the eluate occurs via electrowinning, with retorting for removal of mercury and smelting of the gold to doré.
The plant will operate as a single stream up to the discharge of the SAG mill, after which the two ball mills, centrifugal concentrators, and classifying hydrocyclones will operate as two separate independent streams. The streams will then be recombined for transfer to leach/CIL where they are again separated into two parallel streams for leach and CIL. Following acid wash, the loaded carbon will be stripped in three parallel elution circuits. A single tailings thickener will be installed.
Design development history
The first detailed feasibility study for the project was undertaken by Minproc in 2001 and included metallurgical testwork undertaken by Ammtec to assess grind versus recovery, detailed cyanidation testwork (including the effect of cyanide strength, leach time, oxygen addition, and leach additives), gravity separation, flotation on gravity tails, heap leach amenability, carbon adsorption kinetics and equilibrium, cyanide destruction, intensive treatment of sulphide concentrate by fine grinding and pressure oxidation, and diagnostic leach. Additional testwork was undertaken into equipment sizing parameters including Bond Work Index (crushing, rod and ball), Unconfined Compressive Strength (UCS), Abrasion Index (Ai), Media Competency Tests/JK Pendulum tests for SAG milling, settling rate tests, and viscosity measurements. Pilot testing simulated the proposed milling circuit of SAG plus ball mills in series with pebble crushing of the SAG mill discharge oversize. The study presented a treatment flowsheet with a treatment rate of 20 Mtpa.
The basic engineering design phase was initially awarded to SNC in 2002 and subsequently to Bechtel in 2003, with further testwork commissioned by SNC including investigations into cyanide detoxification undertaken by SGS in Canada. The plant throughput was reduced to nominally 13 Mtpa with a view to reducing initial project capital cost. Bechtel subcontracted the process plant design to Ausenco, who commissioned further metallurgical testwork mainly undertaken by Ammtec, including investigations into gravity concentration, leach kinetics (including the effect of carbon addition), the effect of cyanide concentration, oxygen addition, carbon adsorption kinetics, carbon equilibrium, and pulp viscosity measurements. Further tests were undertaken on behalf of Ausenco into cyanide detoxification of leach tails by CyPlus.
Additional investigations were undertaken by Newmont, including testwork on grind size, gravity concentration, cyanide strength, leaching conditions (addition of lead nitrate and increased lime addition), and potential preg-robbing tendencies. Newmont also investigated ore variability on samples from the different pits, with results reviewed by Aurifex Pty Ltd. to develop revised recovery algorithms for gold and silver for the different ore sources. Two further phases of metallurgical investigations were undertaken under the direction of John Goode and Associates using testwork facilities at SGS in Canada, investigating the effect of gravity recovery after milling, leaching at coarser grind sizes, extended leach time, recirculation of cyanide solution from the tailings, and potential reuse of residual cyanide including the effect of iron chemistry on cyanide stability.
Mineralogy and ore types
The deposit comprises dacite and mixed breccia as the majority of ore, with dacite being the dominant lithology. Three alteration types have been identified: argillic (highly altered, not competent, sometimes friable), potassic (competency ranging from medium to hard), and silicic (hard and abrasive). Gangue minerals consist of quartz, feldspar, and muscovite.
Mineralogical and diagnostic investigations concluded that up to 80% of the gold is free or amenable to cyanidation, 20-30% of the gold is occluded in the sulphides as fine grains (“blebs”) or as solid solution, only a small portion of the gold (less than 5%) is locked in silica, gold is almost always associated with silver, and impurities include copper (generally up to 100 ppm, around 59.68 ppm average) and arsenic (around 89.91 ppm average). Other investigations indicated that a certain amount of coarse gold could be expected, supported by significant variations between assayed head grade and back-calculated head grade for some samples in the variability testwork.
Summary of metallurgical testwork findings
Testwork conclusions from the investigations undertaken to date indicate that all ore types are amenable to conventional SAG/ball milling with pebble crushing of critical size material (SABC circuit). Relatively high recoveries can be achieved by milling to a relatively coarse grind followed by cyanidation. Recoveries of gold and silver at the same grind size are related to the feed grade, sulphur content, and the ore source (Cetate, Carnic, Jig, and Orlea pits). Liberated coarse gold and silver can be recovered by gravity after milling, resulting in marginally improved overall gold recoveries of 1-2%. Primary milling to below 150 microns does not improve leach recovery, with higher gold grades generally observed in the finer fractions, highlighting the partially refractory nature of the ore. The simultaneous addition of carbon with cyanide was noted to improve the leaching kinetics, particularly for gold. Lime consumption is relatively high due to the sulphide level in the ore.
Design basis
The design has been further developed since 2007 by RMGC as part of a Cost Optimisation Program (COP) aimed at reducing project capital and operating costs. In 2012, the capital and operating costs were updated by RMGC to show a base date of Q3 2012. Minor changes were made to the basis of the costs (consumptions of liners, choice of reagents, etc.) although these are not considered significant.
Key reported parameters
| Parameter | Value | Basis |
|---|---|---|
| Ore type | Partially refractory epithermal gold-silver, gold:silver ratio averaging 1:5 up to 1:10 | Testwork and mineralogical investigations |
| Crushing | Open circuit primary crushing | Proposed design |
| Grinding | SAG and ball mill (SABC circuit) to 80% passing 150 μm | Proposed design |
| Gravity recovery | High speed centrifugal concentrators on hydrocyclone underflow | Proposed design |
| Leaching/CIL | One stage pre-leach and six CIL adsorption stages (two parallel streams) | Proposed design |
| Elution | Continuous Zadra elution, three parallel circuits | Proposed design |
| Cyanide detoxification | Copper sulphate and sodium metabisulphite (SO₂/air process) | Proposed design |
| Gold recovery (forecast over LoM) | Around 80% (range 70-83% annually) | Design forecast, varying by ore source, feed grade, and sulphide sulphur level |
| Silver recovery (forecast over LoM) | Around 60% (range 55-66% annually) | Design forecast, varying by ore source, feed grade, and sulphide sulphur level |
| Nominal plant design throughput | 13 Mtpa (reduced from 20 Mtpa in Minproc DFS) | Proposed design |
| Maximum scheduled throughput | Up to 15.4 Mtpa in years 10 and 11 | Proposed design schedule |
| Gold occluded in sulphides | 20-30% | Testwork |
| Gold locked in silica | Less than 5% | Testwork |
| Average copper impurity | Approximately 59.68 ppm | Testwork |
| Average arsenic impurity | Approximately 89.91 ppm | Testwork |
Project website: https://en.wikipedia.org/wiki/Ro%C8%99ia_Montan%C4%83_Project
Technical qualifications
The report notes that the overall recoveries forecast vary significantly dependent on the ore source (Carnic, Cetate, Jig, or Orlea pits), the feed grade for gold and silver, and the sulphide sulphur level in the feed. The ore feed to the plant is classified into soft, medium, and hard material, and the throughput achievable varies from the nominal plant design figure depending on the blend of ore types. SRK considers the assumed recoveries to be reasonable but notes that these may vary from those estimated on a month-to-month basis.
Source: Rosia Montana Project , 2012 Technical Report, October 2012, sections 17 Recovery Methods.
