This technical report describes mineral processing and metallurgical testing for the Cauchari-Olaroz project, based on proposed design and historical operating data.
Report context
This technical report is dated April 2, 2026, and concerns the Cauchari-Olaroz project. The report presents mineral processing information based on original technical-report evidence that includes proposed plant design, historical operating data, and metallurgical testwork results. The project is a gold mine operation with associated processing infrastructure.
Processing route
Oxide and Free-Milling Circuit
The proposed processing plant comprises two largely independent circuits, each designed to accommodate distinct ore types based on mineralogical and metallurgical characteristics. The Oxide and Free-Milling Circuit is designed to process oxide, transition, and free-milling ore. The flowsheet includes standard crushing, ball milling, gravity recovery via Knelson concentrators, and a conventional carbon-in-leach (CIL) circuit.
Sulphide Refractory Circuit
The Sulphide Refractory Circuit is purpose-built for the treatment of sulphide refractory ore. The flowsheet consists of primary crushing, milling, flash and conventional flotation, ultrafine grinding (UFG), and cyanidation via a Pumpcell carbon-in-pulp (CIP) circuit. The flotation concentrate is subjected to a gravity flow pre-oxidation stage, followed by a leaching and CIP circuit.
Plant Performance and Flexibility
The processing plant averages 89% life-of-mine gold recovery (excluding optional flotation tails leaching), with a range of 78.4% to 96.4%. This variability stems from different free-milling and refractory ore types, necessitating blend control to optimise material routing through either the CIL circuit for oxide/free-milling ores or the flotation-UFG-leach circuit for sulphide refractory ores. This dual-circuit design provides operational flexibility as oxide and transitional ores deplete, allowing for partial retreatment of tailings to enhance overall recovery.
Metallurgical Testwork History
Since 2006, metallurgical test work has been conducted to characterise ore variability, define geometallurgical domains, and establish recovery parameters aligned with the current plant flowsheet. Key test work components have included Bottle Roll Leach Tests (BRT) to evaluate cyanide-soluble gold variability across different lithologies and domains, informing direct leachability and recovery modelling; Laboratory-Scale Plant Simulation Tests to assess comminution, gravity recovery, flotation (for sulphide association), and cyanidation leach to estimate overall recoveries; and Gold Deportment and Diagnostic Mineralogy to identify gold carriers and quantify refractory, free-milling, and encapsulated gold proportions, revealing metallurgical constraints.
Domain-Specific Testwork Results
Recovery results are domain-specific (oxide, transition, fresh) and show head grade and mineralogical variations. Results of recent test work completed to characterise new deposits or new domains are summarised as follows:
- Rhino: Oxide samples achieved over 90% extraction, while transitional and fresh samples showed lower BRT extractions of approximately 83% and 72%, respectively. Subsequent plant simulation with finer grind yielded improved recoveries to 85% for both material types.
- Agbarabo: Oxide and transition domains demonstrated strong direct-leach responses (greater than 90% and approximately 85%, respectively). High-grade fresh material was consistent, while low-grade fresh averaged approximately 75% (due to finer gold association), with further recovery improvements anticipated via UFG.
- Airbo: Preliminary BRT on fresh material yielded 66% direct-cyanidation extraction. Additional flowsheet simulation test work (flotation, UFG, pre-oxidation/intensive leach) is ongoing.
- Ndala: Oxide and transition domains achieved over 90% extraction. Fresh samples reported lower extraction (less than 75%) due to lower liberation, but additional extraction was achieved at a finer grind size.
- Sessenge–Gorumbwa: Upper-lens samples averaged a low 61.45% recovery (linked to high arsenic content of approximately 2,500 ppm), whereas bottom lenses responded with higher gold extraction (greater than 70%).
- Oere: Oxide domains achieved over 90% extraction, transition domains approximately 85%, and fresh samples around 82% (with recovery of 88% at finer grind).
- KCD 11000 Lode and KCD Deep: Averages of approximately 82% and 79%, respectively, showing limited variability. However, significant recovery improvement with finer grind was achieved with both deposits reporting over 90%.
These outcomes are consistent with identified geometallurgical controls including preg-robbing and submicroscopic or occluded gold within sulphides. Where direct-cyanidation extractions are lower, test work indicates significant recovery uplift through flotation, ultrafine grinding, elevated dissolved oxygen, and adequate residence time. This supports a dual-route strategy: CIL for free-milling domains, and flotation-UFG-concentrate cyanidation for gold attached or enclosed within sulphides, especially in high-arsenic, refractory domains.
Key reported parameters
| Parameter | Value | Basis |
|---|---|---|
| Plant design capacity | 7.2 million tonnes per annum | Design |
| Average LOM gold recovery | 89% | Design (excluding optional flotation tails leaching) |
| Gold recovery range | 78.4% to 96.4% | Design range |
| Rhino oxide BRT extraction | Over 90% | Testwork |
| Rhino transitional BRT extraction | Approximately 83% | Testwork |
| Rhino fresh BRT extraction | Approximately 72% | Testwork |
| Rhino transitional improved recovery (finer grind) | 85% | Testwork |
| Rhino fresh improved recovery (finer grind) | 85% | Testwork |
| Agbarabo oxide direct-leach | Greater than 90% | Testwork |
| Agbarabo transition direct-leach | Approximately 85% | Testwork |
| Agbarabo high-grade fresh | Consistent | Testwork |
| Agbarabo low-grade fresh average | Approximately 75% | Testwork |
| Airbo fresh BRT direct-cyanidation extraction | 66% | Testwork |
| Ndala oxide extraction | Over 90% | Testwork |
| Ndala transition extraction | Over 90% | Testwork |
| Ndala fresh extraction | Less than 75% | Testwork |
| Sessenge–Gorumbwa upper-lens average recovery | 61.45% | Testwork (high arsenic content approximately 2,500 ppm) |
| Sessenge–Gorumbwa bottom lenses extraction | Greater than 70% | Testwork |
| Oere oxide extraction | Over 90% | Testwork |
| Oere transition extraction | Approximately 85% | Testwork |
| Oere fresh extraction | Approximately 82% | Testwork |
| Oere fresh recovery at finer grind | 88% | Testwork |
| KCD 11000 Lode average recovery | Approximately 82% | Testwork |
| KCD Deep average recovery | Approximately 79% | Testwork |
| KCD deposits improved recovery with finer grind | Over 90% | Testwork |
| Total mine production 2025 | 8.322 million tonnes | Historical operating data |
| 2025 head grade | 2.79 g/t Au | Historical operating data |
| 2025 gold production | 673 thousand ounces Au | Historical operating data |
| 2025 recovery | 90.31% | Historical operating data |
Project website: https://www.lithium-argentina.com/projects/cauchari-olaroz
Technical qualifications
The processing information in this report is based on proposed design, historical operating data, and metallurgical testwork results as described in the source technical report. No independent verification of metallurgical testwork data has been performed by the author of this article. The report does not contain economic analysis for material expansion of current annual production. The Qualified Persons note that mineral processing and metallurgical testing fundamentals are well established at the operation, with ore characterisation insights gained through ongoing test work and actual operations contributing to relatively high, consistent, and predictable gold recoveries.
Source: AngloGold Ashanti NI 43-101 Technical Report for Kibali Gold Mine, DRC – effective date 31 December 2025, dated April 2, 2026, Sections 1.7 (Mineral processing), 1.13.3 (Mineral processing conclusions), and related sections.
