1. Demonstrating the design and projected performance reflected in the Definitive Feasibility Study (“DFS”)
- Fast and high average leach extraction were achieved in all phases of the demonstration plant. As expected the leach extractions were higher during the open circuit leaching than during the closed circuit leaching. The latter which ultimately informs the project metallurgical assumptions.
- These open circuit extraction values ranged from 92.2% to 94.5% whilst the closed circuit extraction values ranged from 92.5% to 93.0%. The average extraction for the three cribs operated in closed circuit was 92.8% (compared to the DFS projection for a scaled up heap of 86.9%). The associated six columns recorded an average uranium extraction of 93.0% also operating in closed circuit.
- Average sulphuric acid consumption for the closed circuit trials was 13.6 kg/tonne (compared to the DFS projection of 17.6kg/tonne) for the cribs whilst the columns recorded an acid consumption of 14.2 kg/tonne.
- The leach solution collected was clear and clean and subsequent visual observations during the unloading of the cribs confirmed uniform percolation through the material and integrity of the agglomerate.
- The clarity of the solution raises an issue regarding the need for the two pinbed clarifiers currently included in the DFS processing flowsheet.
- Analysis indicated no evidence of build-up of deleterious elements occurring during the recycling of leach solution.
2. Further enhancing project knowledge
- Operation of the demonstration plant fostered a more complete understanding of the leach kinetics associated the Etango ore, in particular the management of oxidant demand during the leaching process.
- The total sample tested in Phase 1 through to Phase 5 amounted to approximately 275 tonnes significantly increasing the confidence in the metallurgical assumptions.
3. Pursuing value engineering
- The ore continues to leach quickly and uniformly even with coarser High Pressure Grinding Roll (HPGR) crushed ore and conventionally cone crushed ore, providing evidence for potential capital and operational cost savings.
- Uniform percolation coupled with rapid and high leach extraction once again points towards the potential to further optimise the heap leach configuration.
- The uranium extraction achieved for Phase 5 averaged 93.6%, in-line with preceding phases. Consistently fast leach kinetics have been observed from the Etango ore, achieving over 90% uranium extraction within 20 to 22 days. This compares to the DFS projection for a scaled-up heap of 86.9%.
- Average sulphuric acid consumption maintained the linear relationship with time as previously observed and averaged 14.4 kg/tonne (compared to the DFS projection for a scaled-up heap of 17.6kg/tonne).
- The columns treated with reduced or no binder addition (relative to DFS parameters) all showed similar percolation and leaching results. Although further geotechnical work would be required before binder addition is changed, the results are highly encouraging and point to the potential for further reagent optimisation.
The layout design of the Demonstration Plant took advantage of a historical exploration camp and the topography of the area. The flowsheet resembles the front end of the processing plant up to the heap leaching stage.
The photograph below shows the plant site from the north east. The plant is self-sufficient with respect to electricity and operates on a continuous cycle. The cribs are able to be operated in open (i.e. individually) or closed loop (i.e. in series) in order to fully simulate full scale operation of the heap leach pads.
Acid leaching of agglomerated ore stacked to 5m occurs in four 2m x 2m x 6m leach cribs. In addition to the cribs, eight 5m high columns with an internal diameter of 0.185m enable parallel leaching. This arrangement enables direct comparison of the leaching performance of the respective 200kg and 30 tonne samples and hence an assessment of the scale–up factors as well as the opportunity to conduct optimisation studies on smaller volumes. The setup of the cribs and the columns is shown in the photographs below.
The series of gates on the front of the cribs allow for the progressive stacking from the bottom up, instead of dropping the material in at the top. This practice simulates the vertical drop of the stacking procedure envisaged during full scale operation.
Bulk Sample
The 3,000 tonne bulk sample was sourced from the northern end of the Etango ore body as shown in the photograph below. This area was selected because the ore grade in this location was representative of the Etango ore body and the cost of excavation was low due to it outcropping. The clean ore allows for controlled blending with waste material to achieve the dilution assumed in the DFS.
Comminution
Blasting
The 12m by 21m area was blasted on a 2m x 2m pattern with 5m deep drill holes. The blasted material shown below highlights the clean nature of the granite (alaskite) hosted ore.
Crushing & Grinding
The blasted material was transported 2 km to the plant site. The bulk sample was crushed using conventional primary and secondary crushing to generate ore with P100 of <22mm. This product was then further process for tertiary crushing using a High Pressure Grinding Roll (HPGR) unit to generate the DFS product with a target P80 of 5.3mm.
Over and above generating material similar to that specified in the DFS, two other samples were generated to enable future value engineering studies. These were coarser HPGR and conventional crushed material. Waste material (Gneiss rock) to be used for ore dilution purposes was crushed in a similar manner. The ore and waste stockpiles generated are shown in the photograph below.
Ore Sample Preparation
The ore sample was diluted by 5% to 10% with Gneiss rock sourced from the planned pit area to generate a blended sample representative of the projected run of mine ore feed. Ore blending was conducted with the aid of a Quarter Splitter and a Front End Loader as shown below. The Quarter Splitter was used to prepare a uniform sample for each of the four cribs and eight columns.
The material in the four cribs had a P80 of between 3 and 4 mm by comparison to the DFS specification of P80 < 5.3 mm.
Ore Agglomeration
The ore was fed into the agglomeration drum via conveyor at a controlled feed rate, and agglomerated through adding sulphuric acid, a polymer binder and water to produce agglomerated ore with properties as per DFS specifications. Agglomerate samples were taken from the stacking conveyor on an hourly basis and analysed for moisture content.
The photographs below show the agglomerated ore at approximately 10 – 12% moisture content.
Ore Stacking and Column Loading
Crib stacking and column loading occurred concurrently with the agglomeration process using a conveyor. To minimise compaction and segregation in the cribs, a drop height of one meter from the head of the conveyor was maintained throughout.
The agglomerated ore fed to the representative columns was collected on an hourly basis from the conveyor during each crib stacking and then gradually introduced into each column using a sample bag. Care was taken to also maintain a one meter drop height for the columns stacking.
The cribs and columns were pre-loaded with a 350 mm layer of drainage material to serve as a filter for the solution collection during irrigation.
The cribs and columns were left to cure (three days for Phase 1, one day for Phase 2 and two days for Phase 3) after stacking before initiating the leach irrigation phase.
Solution Irrigation and Drain
Phase 1 & 2 entailed an open circuit heap leach operation of the Cribs and the respective Columns. Each crib was fed from a separate Irrigation Feed Tank located in the reagent mixing area as shown in the photograph below. The leach solution comprising of 16 – 17 g/l of Sulphuric acid and 3g/l Ferric was prepared in the Reagent Mixing Tank and then transferred to the respective feed tanks.
Leach solution was introduced at the top of each crib via dripper lines at a constant irrigation rate of 15 L/m2/hr. The photographs below show the even solution distribution as observed at the start of and during irrigation of a crib.
Leach irrigation was conducted for 20 days followed by a post leach drain of 2 days and then a rinse and post rinse phase of 3 days and 5 days respectively. A weak sulphuric acid solution (2g/l) was used as rinse solution.
Phase 3 of the Demonstration Plant work program entailed the closed circuit heap leach operation of three cribs. The configuration was designed to mirror the configuration of the full scale heap operations. Two columns were operated in parallel with each crib also configured in a closed circuit. The configuration of the Phase 3 layout is depicted schematically below.
The agglomeration and stacking procedure was similar to that in the preceding Phases of the work program. After stacking, the cribs and columns were left to cure for 2 days, before initiating the leach irrigation phase. The curing time was one day shorter than during Phase 1 and one day longer than during Phase 2. The curing time was selected based on the performance of the cribs during previous test work and taking into consideration the likely practical operational constraints that would be encountered during commercial operation.
Leach solution was introduced at the top of each crib via dripper lines. Leach irrigation was conducted for a total of 22 days in two separate stages in order to simulate the conditions of a commercial heap leach operation:
- Stage one extended from day 1 to day 15 where the freshly stacked ore was irrigated with intermediate leach solution (ILS),
- Stage two extended from day 16 to day 22 when irrigation was conducted with a simulated raffinate[1] solution of 20ppm.
The ILS utilised for Crib 7 was collected during the Phase 2 work program conducted in 2015, whilst ILS for Crib 8 and Crib 9 was collected from Crib 7 and Crib 8 respectively. In both cases the solution was collected from day 7 onwards.
The raffinate solution was made up by diluting the stored solution from previous test work to obtain a uranium tenor consistent with that of the DFS design.
The leach solution collected from Crib 7, Crib 8 and Crib 9, during the first six days of operation of each crib, was designated as the pregnant leach solution (PLS). The PLS was stored separately and will be utilised for the solvent extraction (SX) work which is part of the Phase 4 program.
The leach phase was followed by a post leach drain of 2 days and then a rinse and post rinse drain phase of 6 days and 5 days respectively. A weak sulphuric acid solution (2g/l) was used as rinse solution.
[1] The return solution from solvent extraction from which the U3O8 has been removed.
Redox Potential of Irrigate Solution
In the DFS, the operation of the full scale heap ferric to ferrous ratio will be controlled by addition of Hydrogen Peroxide (an oxidant which converts ferrous back to ferric). In order to reduce the number of hazardous chemicals during the phase 3 program it was decided to simulate the latter reaction by continuing to add additional ferric, thereby eliminating the need to add Hydrogen Peroxide. It was found that it was more difficult to manage the ferric/ferrous ratio due to the recirculation of the ferrous in solution. However, despite the difficulties experienced in managing the ferric/ferrous ratio the final leach extraction were in line with previous results, demonstrating the robustness of the process.
Ripios Uploading from Cribs and Columns
At the completion of the post rinse phase, all four cribs and their respective columns were carefully unloaded in a manner which enabled taking samples for assaying of uranium, moisture and size distribution. The location of these samples was accurately recorded to enable developing three dimensional profiles of the leach performance. The information was also used to determine the final extraction in each crib.
A total of nine core samples were obtained for each meter segment using a sampling grid designed to sample across the crib profile.
A core rod with an internal diameter of 110 mm and 1 m in length was used to obtain the core samples for each meter. The photographs below illustrate the unloading process and sampling methodology.
Ore from each meter interval was also carefully removed by hand and shovels and weighed to confirm the stacked tonnes. The samples were analysed for moisture to determine the actual dry stacked tonnage for each crib.
Heap stability and the integrity of the agglomerates were clearly evident during unloading process. As expected slight segregation of ore due to the stacking was observed at each meter segment, with the finer content located in the centre as shown in the photograph below. However no percolation issues were observed.
Head grade samples were taken during Ore Blending and Quartering Phase, agglomerate samples taken during the Agglomeration Phase and Ripios samples during the Crib unloading phase.
Monitoring of the solution inventory was conducted on a daily basis. Analytical services were provided by the Bureau Veritas laboratory in Swakopmund.
Uranium Extraction
On average, approximately 85% uranium extraction was achieved by day 16 and over 90% by day 20 (refer graph below). Leach irrigation was stopped at day 22 and the overall uranium extraction achieved after the drain, rinse and post rinse drain phase was approximately 92.8% (compared with DFS projections for a scaled up heap of 86.9%). The extraction curves for Crib 7, 8 & 9 are similar in profile but differ from that recorded during Phase 1 and Phase 2 between day 7 and day 18. This is due to the adverse ferric/ferrous ratios during this time as discussed previously. However as can be seen the terminal extractions are in line with those recorded during Phase 1 and Phase 2.
Acid Consumption
The rate of acid consumption was similar with previous testing and averaged 13.6 kg/tonne for the four cribs (compared with DFS projections of 17.6kg/t).
The area from where the ore sample was sourced has been rehabilitated.
Tarpaulins placed on all the stockpiles to both protect the samples and minimise dust generation.
Double lined solution evaporation ponds with leak detection.
