Pits and slopes under pressure: critical triggers and decision governance.

Three predictable vectors and the challenge of decision-making under uncertainty.

In open-pit mines, significant instabilities are almost never "geological surprises." In practice, they are the predictable result of three accumulating and reinforcing factors: water that alters the effective stress regime and degrades mobilizable resistances; structural anisotropy that defines the kinematics and shortens the path to failure; and operational interventions (especially blasting and mining sequences) that introduce damage, change boundary conditions, and compress response time. The central point is not recognizing these triggers—every technical team recognizes that. The point is to govern the decision under uncertainty, with explicit criteria, integration between disciplines, and traceability of "why" and "when" each operational constraint was applied. A safe slope is not a slope with a comfortable number in a report. It is a slope with declared uncertainties, tested hypotheses, and a decision system that operates at the pace of the operation.

The asymmetry between mass response and organizational rhythm

Operational reality imposes an asymmetry: the massive "responds" continuously, while the organization decides in discrete moments, in meetings, shifts, and production windows. It is in this friction that events are born. Stability models, however good they may be, lose value when the system does not translate signals into actions. Similarly, sophisticated monitoring becomes just an archive of evidence if there are no triggers, levels, and responsibilities previously agreed upon. In slope governance, the most costly mistake is not getting a parameter wrong. It is getting the mechanism wrong, underestimating the coupling between processes, and failing to create decision-making discipline.

Water as a dominant and often underestimated trigger

Water remains the most frequent and, paradoxically, the most underestimated trigger when a pit "appears dry." This is because many operations treat rainfall as a meteorological variable and not as a variable of the rock mass's hydraulic system. From a geotechnical point of view, what matters is the evolution of pore pressures and the dynamics of dissipation. Infiltration, fracture connectivity, the presence of hydraulic barriers, the level of damage, and the pumping regime determine whether a rainfall event will be merely an operational inconvenience or the beginning of a change of state. In deep pits and heterogeneous rock masses, hydraulic compartmentalization and perched aquifers are common, with distinct responses by sector. A "dry sector" does not validate its neighbor. Mature governance, therefore, asks not only "how much it rained," but "where the water entered, how it moved, and what was the rock mass's response."

Hydraulic mechanisms and implications for decision-making.

The action of water manifests itself in technically distinct but operationally convergent ways. Increased pore pressure reduces effective stresses and compromises available resistance, especially in altered zones, fine materials, clayey contacts, and discontinuities with infill. In parallel, rising water levels and localized seepages indicate preferential trajectories and loss of drainage efficiency, frequently associated with clogging of drains, degradation of channels, maintenance failures, or changes in the upstream recharge regime. There are also progressive mechanisms, such as internal erosion and leaching of fines at interfaces, which may not produce immediate rupture but accelerate deformations, open flow paths, and reduce apparent cohesion over time. For decision-making, this means one thing: rainfall is context, hydraulic response is evidence, and the combination of the two defines urgency.

Structural anisotropy and the definition of "where" and "how"

If water is the most common trigger, structural anisotropy is the trigger that most frequently explains "why it happened in that place and in that way." In open-pit mines, ruptures are rarely controlled by the intact rock mass in "homogeneous" shear. They are dominated by discontinuities, foliations, shear zones, lithological contacts, and structural arrangements that define the kinematics: planar rupture, wedge rupture, toppling, and compound mechanisms with shear and sliding components. The real difficulty lies less in recognizing that "there are structures" and more in quantifying what actually controls stability: persistence, connectivity, roughness, infill, spatial variation, and, above all, the degree of damage induced by excavation and blasting, which can transform a set of "discrete" fractures into an effectively continuous rupture surface.

Recurring failures in structural governance

This is where three governance failures emerge, recurring in operations of different sizes. The first is mapping that doesn't translate into decision-making: structural data remains confined to reports, while planning and operation proceed without incorporating kinematic zones and constraints. The second is modeling that "forces" parameters to calibrate a safety factor, masking mechanisms and creating false confidence. The third is treating structural uncertainty as a detail, when it should be an explicit part of the system: sectors with unfavorable geometry may be "stable on average" and unstable in subdomains, and it is in these subdomains that the operation usually works first. In terms of early signs, anisotropy appears as geometric coherence: cracks parallel to the crest, progressive opening in an alignment, block falls with repetitive geometries, aligned seepages, and displacements with a direction compatible with the expected mechanism. The correct question is not "is there a crack?"; it is "does the crack confirm which mechanism and which trajectory, and what does this change in the next mining and blasting decision?"

Disassembly as accelerators of existing mechanisms

The third vector, blasting, is often treated in a simplified manner, reduced to vibration limits and PPV reports. This solves only a fraction of the problem. Blasting modifies the system in multiple ways: it induces damage and relaxation of the rock mass (amplifying connectivity and reducing interface resistance), alters boundary conditions in banks and berms, can degrade surface drainage, and, depending on the hydrogeological context, can influence flow paths and localized pressures in fractures. Furthermore, the operational sequence often amplifies the effect of blasting: high banks, "delayed" berms, inoperative channels, temporary loadings near the crest, and multiple fronts in structurally unfavorable sectors create a scenario where blasting ceases to be an isolated event and becomes an accelerator of an already established mechanism.

Damage control versus vibration control

Robust governance differentiates between "vibration control" and "damage control." Vibration is an indicator; damage is a geometric and mechanical consequence. Without acceptance criteria for the excavated profile, overbreak/backbreak control, berm discipline, and clear rules restricting blasting in critical sectors, the system operates on autopilot until the wrong combination occurs: water increasing pore pressure, structures providing kinematics, and blasting providing damage and a temporal trigger. When this happens, the response time shrinks, and the decision becomes dependent on individual courage rather than process.

Decision governance as a security operating system.

Given this scenario, the key element that separates resilient operations from vulnerable ones is decision governance. Governance here isn't an organizational chart. It's a geotechnical safety operational system that can consistently, auditably, and rapidly convert evidence into action. This system begins with defined roles and authority: who declares a state of alert, who raises the response level, who restricts operations, who authorizes a return to normal operations, and based on what criteria. If this isn't formalized and practiced, the decision becomes an informal negotiation under pressure, and the slope becomes a hostage to production urgency and hierarchical asymmetry.

Trigger architecture with operational levels

The second component is a multi-level trigger architecture. A single “threshold” rarely works because it mixes weak and strong signals, producing two equally bad extremes: frequent alarms that wear down the system or excessive tolerance that postpones action until the response time becomes too short. A multi-level structure, with associated actions, allows for response calibration and preserves credibility. In practice, this means defining operational states, such as observation, alert, action, and emergency, and associating measurable variables, matching criteria, and explicit operational measures with each state. The goal is to avoid the most common pattern in the field: discussing “whether we will act” when the trigger is reached. What should be discussed beforehand is “what action will be taken” and “what changes to return to the normal state.”

Layered variable integration

For this to work, it is necessary to integrate variables in layers, because isolated triggers are ambiguous and generate noise. An effective architecture begins with the hydroclimatic context, using rainfall in moving windows, intensity, forecast, and response history by sector, to increase vigilance and adjust operational posture. Next, it incorporates the hydrogeological response, with pore pressures, water levels, flows in drains, seepages, and pumping performance, because here the water is observed "within the system," and not just in the sky. The third layer is the mechanical response, with displacements, rates, and accelerations, including spatial coherence and direction, as it is the kinematics that reveals whether the mechanism is mobilizing. The fourth layer is operational integrity, which verifies whether berms, surface drainage, crests, accesses, and mining sequences are compatible with the sector's risk status. Without this layer, the organization correctly diagnoses and continues executing its own trigger.

Recognizing changes in state, not just extremes.

The value of this architecture lies in the combination. A mature operation does not expect each variable to "exceed its historical limit" in isolation. It recognizes that serious events often begin as weak, but consistent, correlations between layers. Significant rainfall with a hydraulic response and incipient acceleration of displacements, even if below the "limit," can be more critical than a high, but stable, absolute displacement without a hydraulic response. Similarly, rising pore pressure without rain may indicate drainage failure, upstream recharge, interception of conducting structures, or changes in the pumping regime. In decision governance, the focus is not only on detecting the "worst-case scenario," but on recognizing changes in state.

Integration ritual in rhythm with the operation

The third component is the integration ritual. Without a disciplined routine, decision-making depends on ad hoc meetings and the availability of key personnel. Integration should occur at the pace of operations, with a fixed agenda and common language: water status by sector, structural status and dominant mechanisms, instrumentation status and data quality, decommissioning plan and restrictions, critical areas and mitigation measures. This ritual does not replace in-depth analyses, but it ensures that the system responds in the short term and that in-depth analyses are risk-oriented, not driven by technical curiosity.

Traceability as a technical and institutional protection.

The fourth component is traceability. Decisions without documentation become mere opinions. Robust records need to capture the observed condition, the mechanism hypothesis, the available evidence, the explicit uncertainties, the action taken, the person responsible, and the reassessment deadline. This is not bureaucracy. It is what sustains consistency across shifts, protects the organization during audits, and, above all, creates institutional memory to avoid repeating the same mistake in seasonal cycles and team changes.

Making decisions with imperfect data is part of maturity.

There is an inevitable discomfort: there will always be a lack of data. Those who use "lack of data" as a reason not to decide, in practice, transfer the decision to the physics of the problem, and physics decides without considering production, exposed personnel, and business continuity. Maturity lies in deciding with the best available data, explicitly stating what is missing, reducing uncertainty in a targeted way, and adjusting the instrumentation and investigation plan to answer the right questions: where is the mechanism, what is its sensitivity to water, what is the influence of operational damage, and what is the plausible response time of the system.

Evidence of mature governance

When governance is mature, the operation can present objective and consistent evidence: a living map of mechanisms by sector, with confidence levels and operational implications; a risk-oriented instrumentation program, in which each instrument has a clear purpose and data quality criteria; a trigger matrix with levels, variables, windows, and actions; a history of decisions and exceptions that shows coherence and learning; and integration with water planning and management, so that transient conditions are reduced and not just monitored. This set of factors transforms slope stability into a safety system, not just a collection of studies.

Stability as a system problem, not a tool problem.

Ultimately, the performance of slopes in pits is less a question of "having or not having" a model, a radar, or a report. It's a systems problem: can the organization recognize changes in state, integrate weak signals into consistent decisions, and act with discipline before the response time disappears? Water, anisotropy, and rockfalls will continue to exist. The differentiating factor that separates resilient operation from vulnerable operation is whether, under pressure, with incomplete data and multiple fronts, the decision remains the same, made with the same ritual, the same standard of evidence, and the same non-negotiable priority: protecting people and maintaining technical control over what truly governs risk.