The 2014 technical report for the Norasa Uranium Project describes a proposed processing facility based on a two-stage crush plus single-stage SAG mill circuit with acidic leaching, followed by filtration, ion exchange, solvent extraction, and ADU precipitation, with radiometric sorting deferred to a later stage.
Report context
The Norasa Uranium Project technical report, dated March 2014, was prepared for Valencia Uranium (Pty) Ltd / Dunefield Mining Company (Pty) Ltd. The report updates earlier engineering work, including the 2008 Definitive Feasibility Study completed by AMEC South Africa and a January 2010 Engineering Cost Study. Recent evaluation of additional mineral deposits close to the existing Valencia Main deposit increased reserves, prompting a review of the proposed plant design. During 2009, additional testwork was performed with the aim of reducing reagent consumptions and confirming the process flowsheet. This testwork was incorporated into an Engineering Cost Study that used the 2008 DFS as its starting point, which had included a staged crush-rod mill circuit without radiometric sorting. A two-stage crush plus single-stage semi-autogenous grinding mill using acidic filtrate at a grind of P80 ~600 μm delivered the highest overall value.
Processing route
Unit operations
The proposed processing facility comprises the following unit operations:
- crushing, sorting, screening and stockpiles
- milling
- leaching
- belt filtration
- continuous ion exchange (CIX)
- solvent extraction (SX) and ammonium diuranate (ADU) recovery
- filtration
- calcination
Radiometric sorting would be introduced at a later stage, and the system will have to be designed based on real crusher product rock size distribution. Radiometric sorting was not included as part of the process flowsheet in the study and would be treated as a modular retrofit to the plant at a later stage. The final process design will make provision for the addition of radiometric sorting following an initial evaluation period.
Crushing and comminution circuit
The crushing circuit comprises primary crushing, followed by scalp screening and open circuit secondary crushing of the oversize. The crushed ore reports to the coarse ore stockpile from where it is reclaimed and fed to the SAG mill.
Primary crushing. The design tonnage is based on 14.9 Mtpa to be crushed via the primary crusher, subject to radiometric sorting for the elimination of a portion of barren or low grade material, thereby increasing the head grade and decreasing the volume for further treatment. The feed size to the primary crusher is limited to a top size of 1,200 mm. The selected crusher is a gyratory crusher with a feed capacity of 3,600 tph. While this is in excess of the annual crushing tonnage requirements, it has been sized at this capacity due to mining requirements. The mining schedule requires that two truckloads of 150 t be able to be crushed in five minutes. The ROM bin will have two-sided truck tipping.
Sorting plant. Once radiometric sorting is implemented, the primary crusher product is fed via a conical stockpile to the sorting plant with a maximum rock size of 260 mm. The radiometric sorting plant is designed to sort barren or very low grade material from uranium-containing material. For the purposes of the study, it was assumed that all rock of +32 mm would be sorted, accounting for 69% of the crusher product. The plant is designed for a 36% reject rate of sort feed, with an associated 13% loss of metal. Hence, with a 31% sort plant bypass, the overall reject rate would be 24.8% and the metal loss 8.6%.
The sorter plant consists of four fine ore and five medium ore sorters in parallel, fed from the underflow of a vibrating screen at size -150 mm. The screen oversize (+150 mm) reports to two coarse ore sorters in parallel. The rejects from all sorters could report to either waste dumps or the tailings disposal facility. The accepted ore portion reports to the secondary crushing plant. Due to the improved grade in the ore and the difficulty of designing a radiometric sorter for a greenfields plant, the radiometric sorting plant would not be installed earlier than 2018. A cost benefit analysis to determine a grade cut-point, offsetting ore rejection with metal loss, would be conducted during the initial production stage to assess design parameters using real production data.
The sort plant mass balance provided in Table 17.4 of the report shows the following key parameters:
| Parameter | Value |
|---|---|
| ROM throughput | 14,901,543 tpa |
| ROM head grade | 200 ppm |
| ROM U3O8 content | 2,980,309 kg |
| UltraSort Feed | 10,282,065 tpa at 191 ppm, 1,964,023 kg U3O8, 69.0% mass % |
| UltraSort HG Conc | 6,580,522 tpa at 260 ppm, 1,708,617 kg U3O8, 64.0% mass % |
| UltraSort Discard | 3,701,543 tpa at 69 ppm, 255,406 kg U3O8, 36.0% mass % |
| Fines & Other | 4,619,478 tpa at 220 ppm, 1,016,285 kg U3O8, 31.0% mass % |
| Concentrator Feed | 11,200,000 tpa at 243 ppm, 2,724,902 kg U3O8 |
| Overall sort plant metal recovery | 91.43% |
| Overall sort plant rock reject | 24.84% |
| Overall sort plant reject grade | 69 ppm |
| Overall sort plant upgrade ratio | 1.22 |
Project website: https://www.forsysmetals.com/projects/
The sorting testwork indicated that the fines fraction showed an upgrade of about 10% by grade.
Secondary crushing. The secondary crushing plant consists of a single secondary cone crusher in open circuit with three tertiary cone crushers in closed circuit, as well as a sizing screen for classification. The F80 of the feed to this plant is 120 mm. The final crusher product (classification screen underflow) is minus 32 mm. The secondary crusher is sized to handle the design tonnage plus build an empty stockpile to full capacity within a specified time.
Stockpiles. Current designs are based on conical stockpile configurations. With a height of 40 m and base width of 85 m, the live capacity of the stockpile is approximately 33,600 t with an associated dead capacity of 95,700 t. Additional area is available for dozing capacity to cater for times of extended crusher downtime for crusher liner change. The stockpile base consists of three vibrating feeders onto a single conveyor.
Milling. The SAG mill is fed from the coarse ore stockpile. The mill feed conveyor has a straight approach to the SAG mill, to avoid maintenance issues associated with having feed conveyors over one of the twin pinion motor drives. The circuit incorporates grinding 1,400 t/h solids, at 70% w/w density, in acid liquor in a SAG mill with an 8.4 m diameter and 5.64 m grinding length. The mill is intended to be manufactured of LDX 2101 stainless steel due to the return of acidic water. As most of the acid in the liquor used for grinding will quickly be consumed due to the acid-consuming nature of the solids, the expected mill discharge slurry pH is between 3 and 5. By returning acid liquor to the mill and not neutralising the liquor with lime or limestone, the added cost of these reagents is avoided, offering a significant operating cost saving. Grinding in acid liquor has been commercially applied at other operations.
Slurry discharges via the trommel screen with the oversize material conveyed back to the mill feed conveyor. Undersize discharges into the mill discharge sump. The slurry is pumped to a classification screen near the ground-based section of the SAG mill feed conveyor. Screen oversize material is returned to the SAG mill feed conveyor and screen undersize reports to the leach circuit.
Leaching circuit
The design of the leach circuit includes an ambient sulphuric acid leach at a density of 48% solids using MnO2 (as pyrolusite) as the oxidant to oxidise ferrous iron to ferric iron. Alternative oxidants were evaluated and the density of the leach solution could vary from 48% to 65% based on the oxidant selected. A final decision on which oxidant would be used may still change.
The current layout caters for eleven leach tanks with a feed volumetric flow rate of 2,030 m³/h from tanks of 2,049 m³ capacity, measuring 13.23 m diameter and a 14.9 m operating height. Total leach time is 10.1 hours.
Filtration and clarification
Belt filtration was considered as an alternative to conventional sand/slime separation and CCD washing. A high-level comparison of these options was completed, using only belt filters for the leach tails washing and dewatering. Although slightly more expensive than the CCD options, the belt filters came out as the best technical solution.
The belt filters offer a much lower solution loss to tails than CCD, with the belt filter tails discharged as filter cake with a moisture content of approximately 20% (80% solids). The tails is then discharged and stored on a dry tails storage facility. The filters also allow for a smaller volume of solution to the IX circuit as a result of a smaller wash ratio required for washing. The uranium tenor in the solution to the IX will also be higher.
The belt filter installation consists of nine 149 m² units, fed via gravity from the leach tail distribution box. Filtration rate is 1,040 kg/m²h. Overall filter efficiency is expected to be 99.6%. The final filtrate solution is fed to a single hopper clarifier, with solids content reduced to approximately 20 ppm, which is considered tolerable for the ion exchange process.
Ion exchange circuit
Because Norasa has a fairly low head grade, resulting in a fairly low uranium tenor in the pregnant leach solution, an ion exchange circuit was selected for the initial upgrade of the PLS to a level that can be extracted by solvent extraction. A NIMCIX continuous ion exchange design was previously selected, and was compared to an up-flow moving packed bed (MPBIX) configuration with resin-retaining screens. The MPBIX technology is widely used in the uranium industry and has approximately three times the up-flow velocity, so smaller columns are required, although this option also requires an additional clarification stage upstream. MPBIX has the advantage of reduced resin inventory over the NIMCIX circuit. Five loading columns and five elution columns are required.
Solvent extraction and calcining
Conventional tertiary amine extraction with ammonia/ammonium sulphate stripping was selected for the Norasa flowsheet. Four extraction, three scrub and four strip stages were considered for the SX plant design. Conventional yellow cake precipitation with ammonia followed by centrifuge washing and calcination was selected for the base case Norasa flowsheet. Industry standard values were used for the design of the ADU precipitation plant.
Key reported parameters
| Parameter | Value | Basis |
|---|---|---|
| ROM throughput | 14,901,543 tpa | Design |
| ROM head grade | 200 ppm | Design |
| ROM U3O8 content | 2,980,309 kg | Design |
| Primary crusher feed top size | 1,200 mm | Design |
| Primary crusher capacity | 3,600 tph | Design |
| Secondary crusher feed F80 | 120 mm | Design |
| Final crusher product | −32 mm | Design |
| Coarse ore stockpile live capacity | 33,600 t | Design |
| Coarse ore stockpile dead capacity | 95,700 t | Design |
| SAG mill throughput | 1,400 t/h | Design |
| SAG mill dimensions | 8.4 m dia. × 5.64 m | Design |
| Grind size | P80 ~600 μm | Testwork |
| Slurry density to leach | 48% solids (65% possible) | Design |
| Leach tanks | 11 | Design |
| Leach tank volume | 2,049 m³ each | Design |
| Leach tank dimensions | 13.23 m dia. × 14.9 m operating height | Design |
| Leach residence time | 10.1 h | Design |
| Leach feed flow | 2,030 m³/h | Design |
| Belt filter area | 149 m² each, nine units | Design |
| Filtration rate | 1,040 kg/m²h | Testwork |
| Filter efficiency | 99.6% | Testwork |
| Filtrate clarity | ~20 ppm solids | Testwork |
| Sorter feed top size | 260 mm (after primary crushing) | Design |
| Sorter feed size range | −150 mm | Design |
| Sort plant reject rate | 36% of sort feed (24.8% overall) | Design |
| Sort plant metal loss | 13% of sort feed (8.6% overall) | Design |
| Sort plant mass reject | 3,701,543 tpa | Design |
| Sort plant reject grade | 69 ppm | Design |
| Sort plant reject U3O8 | 255,406 kg | Design |
| Sort plant concentrate | 6,580,522 tpa at 260 ppm, 1,708,617 kg U3O8 | Design |
| Sort plant fines | 4,619,478 tpa at 220 ppm, 1,016,285 kg U3O8 | Design |
| Overall sort plant metal recovery | 91.43% | Design |
| Overall sort plant rock reject | 24.84% | Design |
| Overall sort plant upgrade ratio | 1.22 | Design |
| Concentrator feed | 11,200,000 tpa at 243 ppm, 2,724,902 kg U3O8 | Design |
| Ion exchange loading columns | 5 | Design |
| Ion exchange elution columns | 5 | Design |
| SX stages | 4 extraction, 3 scrub, 4 strip | Design |
Technical qualifications
The report identifies several limitations and items requiring further evaluation:
- The radiometric sorting plant was not included in the base case flowsheet and would be retrofitted at a later stage. The system will have to be designed based on real crusher product rock size distribution, and a cost benefit analysis to determine the grade cut-point, offsetting ore rejection with metal loss, will be conducted during the initial production stage using real production data. The radiometric sorting plant would not be installed earlier than 2018.
- A final decision on which oxidant would be used in the leach circuit may still change, with the density of the leach solution varying from 48% to 65% depending on the oxidant selected.
- The sort plant mass balance assumes no radiometric sorting is applied to the fines fraction, which reports directly to the concentrator feed at the ROM grade.
- Industry standard values were used for the design of the ADU precipitation plant, rather than site-specific testwork.
Source: Norasa Uranium Project, NI 43-101 Technical Report, March 2014, Sections 17 and 17.1–17.8.
