Central Sulawesi — Geology & Mineral Resources CENTRAL SULAWESI — GEOLOGY & MINERAL RESOURCES Central Sulawesi represents one of Indonesia’s most geologically diverse mineral provinces, where ophiolitic–ultramafic complexes, metamorphic belts, intrusive systems, volcanic rocks, and major structural zones create a variety of mineralization environments. Its resource potential extends beyond nickel. 🟢 Ni–Co Laterite Associated with the East Sulawesi Ophiolite Belt, where prolonged weathering of ultramafic rocks produces distinct limonite, saprolite, and serpentinized bedrock profiles. 🟡 Gold The Palu–Poboya area hosts gold mineralization associated with quartz veins and intrusion-related hydrothermal systems. 🟠 Cu–Au Porphyry The Sassak area represents a different mineralization style, associated with intrusive rocks and hydrothermal alteration. ⚙️ Fe–Cr, Sc & REE Potential Ultramafic and weathered granitoid systems also provide geological settings for chromium–iron mineralization and potential critical elements such as scandium and rare earth elements. What makes Central Sulawesi particularly interesting is the relationship between geology, tectonics, alteration, weathering, and mineralization. From a resource perspective, understanding these geological controls is essential before moving into: Exploration → Geological Domaining → Sampling → Estimation → Resource Classification → Mining & Processing Central Sulawesi therefore offers an excellent example of how one regional geological framework can host multiple mineral systems with very different resource-estimation challenges. The next step is not simply asking “How much mineral is there?” It is asking: “Why is the mineral there, where does it occur, and how confidently can we estimate it?” #CentralSulawesi #IndonesiaMining #Geology #MineralResources #Nickel #Cobalt #Gold #Copper #REE #Scandium #MineralExploration #ResourceEstimation #Geometallurgy #Mining #ZVENIA

"Why Southeast Sulawesi?" When people hear the name "Southeast Sulawesi", most immediately think of "nickel". Yet, for a geologist, nickel is not the story—it is merely the outcome of a much longer geological history. The true story began tens of millions of years ago, when fragments of ancient oceanic lithosphere were emplaced onto the continental margin during complex tectonic interactions. Today, this ophiolitic belt represents one of the most remarkable geological archives in Southeast Asia. What makes Southeast Sulawesi exceptional is not simply the presence of ultramafic rocks, but the combination of several geological processes: • Obduction of oceanic lithosphere • Extensive serpentinization of mantle peridotites • Long-term tropical chemical weathering • Stable geomorphological evolution • Hydrological and structural controls on laterite development Together, these processes transformed mantle-derived rocks into world-class nickel laterite deposits. However, not every ultramafic body becomes an economic nickel deposit. Ore formation depends on a delicate interaction between lithology, structural architecture, weathering intensity, drainage conditions, topography, groundwater dynamics, and the duration of lateritization. Understanding these relationships remains one of the greatest challenges in economic geology and mineral resource evaluation. For exploration geologists, Southeast Sulawesi provides a natural laboratory to investigate laterite genesis. For resource geologists, it offers invaluable insights into geological domaining, grade continuity, weathering profiles, density variability, and uncertainty in resource estimation. Perhaps the most fascinating perspective is this: -) "The world sees Southeast Sulawesi as a nickel province. A geologist sees an ancient fragment of Earth's oceanic lithosphere, reshaped by millions of years of tropical weathering into one of the planet's most significant lateritic nickel systems." Understanding the geology behind the resource is what transforms data into knowledge—and knowledge into better exploration and resource decisions.

REE and Nickel Ore Series #2 “Domaining Geological” Geological domaining is where resource estimation begins. Before compositing, variography, or grade estimation, we need to understand one fundamental question: What actually controls the mineralization? In Nickel Laterite, geological domains are commonly controlled by the weathering profile. Limonite, saprolite, transition zones, and bedrock may show distinct geological and grade characteristics, requiring the estimation model to respect these boundaries. In REE deposits, the controls can be considerably more variable and deposit-type dependent. Host lithology, alteration, mineral assemblage, structural features, weathering, and enrichment processes may all influence the distribution of REE mineralization. Therefore, the same domaining philosophy cannot simply be transferred from Nickel to REE. A reliable geological domain should: -. Represent the actual geological control of mineralization -. Separate populations with different grade behaviour -. Preserve geological continuity -. Support appropriate compositing and geostatistical analysis -. Provide a defensible framework for resource estimation The objective is not to create more domains — it is to create meaningful domains. Different commodities may require different geological interpretations, but the principle remains the same: Geology should control the estimation — not the estimation method control the geology. #ResourceEstimation #GeologicalDomaining #GeologicalModelling #Geostatistics #Nickel #REE #RareEarthElements #MineralResources #MiningGeology #ResourceGeologist #ZVENIA

Composite is not simply an averaging process — it is a critical step in ensuring representative data for Resource & Reserve Estimation. Nickel laterite and REE deposits may require different compositing strategies due to differences in mineralization style, geological domains, grade variability, and sample support. Therefore, composite length should be defined based on the geological characteristics of the deposit rather than applying a uniform approach across commodities. Despite these differences, the fundamental principles remain consistent: validated assay data, robust geological and grade domaining, appropriate sample support, statistical and geostatistical analysis, and consistency with the block model and estimation methodology. A properly designed composite should preserve the geological character of the mineralization while providing a reliable and representative input for resource estimation. Geology should drive the composite strategy — not the commodity name alone. #ResourceEstimation #Geostatistics #Nickel #REE #MineralResources #MiningGeology #GeologicalModelling #ZVENIA

Voici le résultat central du Module 2 de notre formation Data Analyse & Géostatistique avec R et Python. Le cas d'étude : une latérite nickélifère à quatre horizons oxyde, limonite, saprolite, bedrock. 1 135 analyses de 1 m à regrouper en composites de 2 m. Méthode A — grille fixe Découpage tous les 2 m depuis la surface, lithologie majoritaire attribuée à chaque composite. → 569 composites, dont 46 à cheval sur un contact. Méthode B par domaine Le découpage redémarre à zéro à chaque contact. → 580 composites, longueurs de 1 à 2 m. Le résultat sur l'oxyde | Méthode | Teneur moyenne | | Grille fixe | 0,401 % Ni | | Par domaine | 0,229 % Ni | | Écart | +75 % | La figure ci-dessus rend le mécanisme évident. Les quatre histogrammes portent chacun leur moyenne : oxyde 0,229 %, limonite 1,160 %, saprolite 1,526 %, bedrock 0,124 %. Notez surtout les échelles de l'axe horizontal l'oxyde se lit entre 0,10 et 0,45 %, la saprolite entre 0 et 3,5 %. L'oxyde est l'horizon le plus mince du profil. C'est donc celui que les débordements de composites affectent le plus : la limonite sous-jacente titre cinq fois plus. Chaque composite qui franchit le contact tire la moyenne de l'oxyde vers le haut. Pourquoi ce n'est pas un détail de méthode Un composite à cheval sur un contact mélange deux populations statistiques. C'est exactement le problème que le domainage sert à résoudre et que le Module 3 démontrera formellement, avec un test de Kruskal-Wallis à p = 1,1 × 10⁻⁸³. Composer à travers un contact revient donc à fabriquer soi-même le problème qu'on cherchera ensuite à corriger. La contrepartie, à assumer La méthode par domaine produit des composites de longueur inégale. Tous les calculs ultérieurs moyennes, variogrammes, krigeage doivent être pondérés par la longueur. Ce n'est pas optionnel. Le contrôle qui valide le code Un compositage pondéré correctement conserve le métal contenu. Si la teneur moyenne varie de plus de 1 % entre les analyses brutes et les composites, la pondération est fausse. Sur notre jeu de données, l'écart mesuré est de 0,000 % sur les six éléments.

Iron Ore Reserve Calculation: More Than Just Tonnage A geological resource does not automatically become a mineable reserve. For iron ore, reserve estimation depends on the integration of: • Geological interpretation & domain modelling • Block model and grade estimation • Density and cut-off grade • Mining recovery & dilution • Pit optimization and slope constraints • Economic and processing parameters • Reconciliation between model and actual production A small change in one parameter can significantly affect the final reserve. The key is not simply to calculate more tonnes — but to define tonnes that are realistically mineable, economically viable, and supported by geological confidence. Better Geology → Better Model → Better Reserve → Better Mine. #IronOre #ReserveCalculation #ResourceEstimation #GeologicalModelling #Mining #Geology #Geostatistics #MinePlanning #MineralResources #MiningEngineering

The Illusion of Precision in Mineral Resource Estimation One of the biggest misconceptions in mineral resource estimation is believing that more decimal places mean greater accuracy. A block model may report grades such as 1.234 g/t or 1.237 g/t, but those numbers should never be interpreted as absolute truth. They are the result of geological interpretation, sampling quality, estimation methods, and statistical assumptions—not direct measurements of reality. Every resource estimate carries uncertainty, regardless of how sophisticated the software or workflow may be. Understanding the difference between precision and accuracy is essential for every geologist and mining professional. True confidence in a resource estimate comes from : 1. Robust geological interpretation 2. High-quality sampling and QA/QC 3. Appropriate domaining and estimation strategy 4. Comprehensive validation 5. Transparent communication of uncertainty As geologists, our responsibility is not to create models that appear perfect, but to build models that are geologically defensible, statistically sound, and fit for decision-making. A resource model is a decision-support tool—not a perfect representation of reality. I'm curious to hear your perspective: In your experience, what contributes more to misleading resource estimates: poor geological interpretation or misplaced confidence in software outputs? #Mining #Geology #MineralResources #ResourceEstimation #Geostatistics #MiningEngineering #CompetentPerson #GeologicalModeling #MiningIndustry #EarthScience #ZveniaMining

WHY SOME HIGH-GRADE DEPOSITS FAIL One of the most common misconceptions in mining is that a high-grade deposit automatically leads to a successful mining operation. In reality, many projects with attractive grades fail to deliver value—not because the grade was low, but because critical geological, mining, metallurgical, and economic factors were not fully understood or incorporated into the decision-making process. A resource model may look impressive on paper, but several factors can significantly impact the outcome: • Poor geological continuity • Incorrect geological domaining • Non-representative density data • Over-extrapolation beyond data support • Low metallurgical recovery • Excessive mining dilution • Geotechnical constraints • Hydrogeological challenges The journey from an exploration target to recovered metal is a continuous process of value conversion. At every stage, uncertainty must be managed, and assumptions must be tested against reality. This is why resource estimation should never be viewed as a standalone exercise. It is the foundation upon which mine planning, reserve conversion, production performance, and ultimately project economics are built. A valuable reminder for all geoscientists and mining professionals: > "The best resource is not the one with the highest grade, but the one that can be mined, processed, and sold profitably." What, in your experience, is the most underestimated factor that causes the gap between a resource model and actual mine performance? Discussion is always welcome. #Mining #Geology #ResourceEstimation #MineralResources #OreReserve #MiningEngineering #Geostatistics #ResourceGeology #MinePlanning #JORC #Exploration #MiningIndustry #Geoscience #Reconciliation

STOP BLAMING THE CUT-OFF GRADE In many technical reviews, discussions often focus on cut-off grades, pit optimization, reserve changes, or project economics. However, these outcomes are merely the final products of a much longer chain of geological assumptions. A cut-off grade does not create value. It only filters value from the model that already exists. Before reserve statements are published, before pit shells are generated, and before economic scenarios are evaluated, fundamental decisions have already been made regarding: • Geological domaining • Spatial continuity assumptions • Variogram interpretation- • Search strategy and estimation parameters • Grade distribution modeling Two competent professionals can start with the same drillhole dataset, use the same software, apply the same cut-off grade, and still arrive at different reserve outcomes. Why? Because geological interpretation drives the model long before economic optimization begins. The greatest risk in resource estimation is often not the estimation algorithm itself, but the assumptions used to represent geological reality. As resource professionals, our responsibility is not simply to generate models—it is to ensure that the geological understanding behind those models is robust, transparent, and defensible. Data is what we have. Geology is how we understand it. The model is how we quantify it. Value is what we create with it. #Mining #Geology #ResourceEstimation #Geostatistics #Variogram #CutOffGrade #MineralResources #OreReserve #MiningEngineering #ResourceGeology #GeologicalModeling #Kriging #MinePlanning #MiningValueChain #Zvenia

In many laterite nickel projects, especially during early exploration stages, we often face one uncomfortable reality: Our drilling density is still too sparse, but we already want to build a resource model. This is where the choice of interpolation method becomes extremely important. Recently, I compared RBF Interpolant (Leapfrog) and IDW on a nickel laterite dataset with relatively wide drill spacing. The results were very interesting and strongly relate to current discussions about whether we sometimes “over-force” geostatistics on immature data. What Happened in the Comparison? Using the same dataset: RBF Interpolant produced: smoother continuity, more geological-looking grade trends,more natural transitions between grade zones and minimal bull-eye effect. Meanwhile: IDW produced: stronger local influence around drillholes, more spotty distributions, clearer bull-eye patterns, higher local variability. This difference becomes very obvious visually. Why Does This Matter? Because in sparse drilling conditions: continuity is still uncertain, variograms are often unstable, anisotropy is not fully understood, and Ordinary Kriging may create a false sense of confidence. This is exactly why RBF becomes attractive. Instead of forcing a variogram model from limited data, RBF uses implicit mathematical continuity to generate smoother and more geologically coherent surfaces. The Interesting Part: Histograms The comparison also showed: RBF : lower variance, smoother distribution, more stable grade population. IDW : slightly higher variance, more local fluctuation, sharper spikes in distribution. This reflects the nature of each method: RBF prioritizes continuity, IDW prioritizes local distance weighting. Neither is automatically “better” they simply answer different geological questions. Should We Replace Kriging with RBF? Not necessarily. For formal resource reporting such as JORC and KCMI. Ordinary Kriging is still widely preferred because it provides: estimation variance, kriging efficiency, slope of Regression, defensible geostatistical validation. However using Ordinary Kriging on data that is still too sparse can sometimes be more dangerous than using a simpler method honestly aligned with the data quality. In Nickel Laterite, RBF can be very useful for: 1. Early-stage exploration 2. Wide drill spacing 3. Geological domaining 4. Implicit saprolite/limonite boundaries 5. Preliminary grade shells 6. Fast model updates The “best” interpolation method is not the most sophisticated one. It is the method that best matches: the drilling density, geological understanding, and confidence level of the dataset. CMIIW..