VMS Deposit Modeling: From Grade-Based Modeling to a Strategic Geological Approach

In volcanogenic massive sulfide (VMS) deposits, metal distribution responds to a well-defined geochemical pattern. The solubility of copper in hydrothermal solutions decreases rapidly with a drop in temperature, ceasing to be soluble at relatively high temperatures. Therefore, copper, being less mobile in solution and more stable at high temperatures, tends to precipitate in zones proximal to the hydrothermal vent or feeder channels, frequently in quartz breccias, stockwork zones, or areas of intense alteration, typically located in the footwall of the volcano-sedimentary series and the massive sulfide bodies.

In contrast, zinc and lead remain soluble at lower temperatures, allowing them to migrate greater distances from the thermal source. This favors their concentration in more distal zones, generally as massive sulfide blankets or stratiform bodies, commonly intercalated within the volcano-sedimentary sequences.

Although this zonation is widely recognized in geological literature, mining practice often reveals resource models that fail to reflect this metallogenic architecture. Many of these models have been built using direct grade interpolation without a clear geological foundation. This can be attributed to time constraints, pressure to deliver results, or the economic context of the time, where low copper prices made its modeling unattractive, especially in deep or marginal zones.

One of the most frequent mistakes in the industry is building models based exclusively on grades without a prior geological model to guide the interpretation. This generates a series of problems:

1. Lack of real geological representation

Grade-based models tend to group zones with similar values without considering whether they belong to the same geological unit or mineralization type. This leads to:

• Geologically inconsistent interpretations.
• Wireframes that do not respect actual contacts or structures.
• Difficulty integrating mineralization with alterations, lithologies, or structures.
• Models with no exploratory value: Loss of the model's predictive value as a tool to define new drill targets.

2. Underrepresented or missing key zones

• Polymetallic layers intersected by most drillholes are often underrepresented by old, overly conservative wireframes, preventing their continuity throughout the deposit.
• Quartz breccias, stockwork zones, or intense alteration areas mineralized with copper—typical of proximal zones in VMS systems—are absent in many historical models. This is because these zones were not drilled in the past or were discarded for not meeting economic criteria at that time.

3. Lack of economic adaptation

When a model is based on grade values without considering updated economic factors, there is a risk of leaving out zones that are profitable today. A clear example is the evolution of copper:

• Old Zneq formula in a mine where Asturmine signs resources: Zneq = Zn + 0.4·Pb + 1.4·Cu
• Updated formula, based on real sale prices: Zneq = Zn + 0.614·Pb + 4.496·Cu + 0.038·Ag + 2.482·Au

With this formula, previously discarded copper zones now enter as economically viable potential, forcing a reconsideration of the resource model's architecture.

4. Models without exploratory value

When the architecture of the system is not represented (for example, its vertical zonation, lithological changes, structures...), the model loses its value as a predictive tool. In other words, it does not help define new drill targets, nor does it guide future exploration.

To overcome these limitations, it is necessary to reorder the workflow and restore geology to the central role it must play in resource modeling.

3.1. Construction of a litho-structural geological model

Before interpolating grades, a conceptual geological model must be generated to:

• Group the main lithologies (e.g., porphyries, barites, massive Zn-Pb layers, quartz breccias, footwall shales).
• Identify relevant hydrothermal alteration zones (sericitization, chloritization, silicification) if they are associated with mineralization.
• Incorporate structural data (faults, contacts, unconformities, folds).

This model can start basically—grouping lithostratigraphic units and mineralized zones—and then be refined based on deposit knowledge.

3.2. Definition of mineralized domains

Within the geological model, estimation domains are defined, separating:

• High-grade zones, where the highest tenors are concentrated.
• Low-grade zones, with possible marginal potential or lower interest.
• Possible transition or mixing zones that require specific criteria.

These domains can be established based on lithological, structural, metallurgical, alteration, or statistical criteria.

3.3. Evaluation of the modeling strategy: Zneq or separate elements?

Once the domains are defined, a decision must be made whether to estimate:

• Each element separately (Zn, Pb, Cu, Ag, Au), which is more rigorous if there is no correlation.
• A metal equivalent (Zneq) if there is a high correlation between metals and it is economically justified.

This decision must be backed by robust statistical analysis (Pearson, Spearman) and an understanding of metal distribution in each domain.

3.4. Targeted drilling campaigns

With the new geological model as a foundation, more efficient drilling campaigns can be defined:

• Targets focused on providing continuity to underrepresented polymetallic layers.
• Drillholes directed at confirming the presence of copper in proximal quartz zones.
• Deepening of areas with unevaluated economic potential under old models.

The information generated by these drillholes can feed back into the model iteratively, improving its accuracy and predictive value.

3.5. Proper resource classification

It is important to distinguish between:

• Geological interpretation, which allows extending wireframes and providing continuity to mineralized bodies.
• Resource classification, which must be based on data density, estimation quality, and interpretation confidence.

Thus, a body with good geological continuity but sparse information can be classified as an inferred resource, while well-sampled zones with reliable estimation can be measured resources. Outlying zones or those based on projections could be considered exploratory potential.

This allows aligning geological interpretation with international classification standards (such as JORC, NI 43-101, or PERC) without confusing extension with certainty.

VMS deposits are complex systems that require modeling approaches beyond mere grade. Building models solely from numerical values without incorporating the system's geological logic can lead to severe errors: underestimating the resource, excluding key metals, and an incomplete view of the deposit's potential.

The key is to return to the basics: geology first, estimation second. A well-constructed litho-structural model that identifies relevant lithologies, alterations, and structures allows for:

• Better representation of the mineralizing system.
• Proper incorporation of metals like copper, which are increasingly important.
• Definition of smarter drilling campaigns.
• More rigorous resource estimation and classification.
• Guiding future exploration with a robust predictive tool.

In times of energy transition and shifting prices, the value of the geological model as a pillar of any technical and economic strategy cannot be understated. In that sense, updating our way of modeling is also updating our way of creating value.